All Road Network of Linear Referenced Data (ARNOLD) Reference Manual

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page i DOT Contract #GS-35F-0001P

September 2014

21BFOREWORD

On August 7, 2012, FHWA announced that the HPMS is expanding the requirement for State

Departments of Transportation (DOTs) to submit their LRS to include all public roads. This requirement

will be referred to as the All Road Network of Linear Referenced Data (ARNOLD). Many States will be

challenged by this requirement, and as such, FHWA has contracted with Applied Geographics, Inc. under

DOT Contract #GS-35F-0001P to produce guidance materials to help State DOTs implement ARNOLD.

The project deliverables are listed below, and tasks 2-6 represent the specific guidance that is offered to

States:

22BPROJECT DELIVERABLES

• Task 1: Project Schedule, Workplan, Risk Assessment and TFTN crosswalk

• Task 2: Local Road Collection Systematic Approach Report

• Task 3: LRS Components and Best Practices Report

• Task 4: LRS Temporal Maintenance Plan Report

• Task 5: LRS Technical Instructions, Rules and Diagrams Report

• Task 6: Reference Manual summarizing information gathered from tasks 2-5

23BPROJECT TEAM

• US DOT

• Joe Hausman (Project Manager)

• Tom Roff (ARNOLD Lead)

• Justin Clarke (Team Lead)

• Contractors

• Applied Geographics, Inc. (Prime Contractor)

• David R. Fletcher (Subcontractor)

• Michael Baker Jr., Inc. (Subcontractor)

• Expert Panel

• Mark Sarmiento – FHWA Planning

• Mike Neathery – FHWA Planning

• Robert Pollack – FHWA Safety

• Stuart Thompson – FHWA Safety

• Maria Chau – FHWA NY Division

• Christopher Chang – FHWA Office of Infrastructure

• Dave Blackstone – Ohio DOT

• Frank DeSendi – Pennsylvania DOT

• Keith Dotson – Kentucky Transportation

• Sharon Hawkins – Arkansas Highway and Transportation Department

• James Meyer – Arizona DOT

• Michele Barnes – University of Michigan

• Paul O'Rourke – Florida DOT

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page ii DOT Contract #GS-35F-0001P

September 2014

Table of Contents

EXECUTIVE SUMMARY ........................................................................................................................................... 1

1 0BINTRODUCTION ............................................................................................................................................. 3

1.1 6BWHY IS AN ALL ROADS OUTLOOK IMPORTANT? ........................................................................................................ 3

1.2 7BWHY THE U.S. DOT AND FHWA NEED ALL ROADS ................................................................................................ 4

1.3 8BWHAT IS THE ALL-ROADS GEOSPATIAL REPRESENTATION STUDY? .............................................................................. 4

2 1BIMPLEMENTATION PLANNING ....................................................................................................................... 5

2.1 9BTHE OPPORTUNITY TO REVIEW THE AGENCY'S OVERALL NETWORK AND LRS DATA MANAGEMENT ...................................... 5

2.2 10BWHAT KIND OF PLANNING DO WE NEED? ............................................................................................................... 6

2.3 11BIMPLEMENTATION PLANNING BEST PRACTICES ........................................................................................................ 8

3 2BDATA COLLECTION ....................................................................................................................................... 10

3.1 12BHOW DO WE COLLECT ALL ROADS ACROSS THE STATE? ............................................................................................ 10

3.2 13BWHAT COMPONENTS WILL OUR BASELINE CENTERLINE NETWORK CONTAIN? ............................................................... 15

3.3 14BDATA COLLECTION RECOMMENDATIONS ............................................................................................................. 16

4 3BINTEGRATING ALL-ROADS AND CONSTRUCTING THE LRS ............................................................................ 18

4.1 HOW DO WE CREATE THE LRS THAT WE NEED? ..................................................................................................... 21

4.2 16BWHAT TOOLS DO WE NEED TO CONSTRUCT AND MAINTAIN LRS? .............................................................................. 26

4.3 17BRECOMMENDATIONS FOR BUILDING THE LRS ....................................................................................................... 29

5 4BONGOING DATA MAINTENANCE ................................................................................................................. 31

5.1 18BWE'VE SPENT ALL THIS EFFORT BUILDING IT; HOW DO WE KEEP IT CURRENT? ............................................................... 31

5.2 19BMANAGING TEMPORALITY WITHIN THE LRS ......................................................................................................... 34

5.3 20BLRS MAINTENANCE RECOMMENDATIONS ............................................................................................................ 36

6 5BCONCLUSIONS AND THE PATH FORWARD ................................................................................................... 38

TECHNICAL APPENDICES ...................................................................................................................................... 40

APPENDIX A 25BBASICS OF LRS ............................................................................................................................. 41

A.1 30BOVERVIEW OF LINEAR REFERENCING SYSTEMS (LRS) ............................................................................................. 41

A.2 31BLINEAR REFERENCING SYSTEM BUSINESS FUNCTIONS ............................................................................................. 43

A.3 32BTYPES OF LRMS .............................................................................................................................................. 46

A.4 33BCOMMON BASELINE NETWORK REQUIREMENTS & MATURITY ASSESSMENTS ............................................................. 47

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page iii DOT Contract #GS-35F-0001P

September 2014

APPENDIX B 26BROADWAY GEOMETRY ............................................................................................................... 51

B.1 34BROADWAY SEGMENTATION ............................................................................................................................... 51

B.2 35BDUAL CARRIAGEWAYS ...................................................................................................................................... 53

B.3 36BTRAFFIC CIRCLES ............................................................................................................................................. 58

B.4 37BRAMPS .......................................................................................................................................................... 61

B.5 38BCUL-DE-SACS AND LOOPS ................................................................................................................................. 63

APPENDIX C 27BROADWAY ATTRIBUTES .............................................................................................................. 66

C.1 39BROUTE EVENTS VS. SEGMENTED ATTRIBUTES ........................................................................................................ 66

C.2 40BARNOLD SCHEMA ......................................................................................................................................... 67

C.3 41BROUTE ID NUMBERING .................................................................................................................................... 68

C.4 42BROAD NAMING ............................................................................................................................................... 69

C.5 43BMULTIPLE LINEAR ROUTE MEASURES .................................................................................................................. 72

C.6 44BPUBLIC VS. PRIVATE ROADWAYS ........................................................................................................................ 73

C.7 45BINSTALL DATE AND INSPECTION DATE.................................................................................................................. 74

C.8 46BADDRESSING .................................................................................................................................................. 75

APPENDIX D 28BMAINTENANCE ........................................................................................................................... 77

D.1 47BMETADATA STANDARDS FOR GIS AND ROADWAY ASSET DATA ................................................................................ 77

D.2 48BPLANNED, DESTROYED AND DECOMMISSIONED ROADWAYS .................................................................................... 78

D.3 49BGEOARCHIVING ROADWAY SEGMENTS ................................................................................................................ 79

D.4 50BROADWAY DATA DISTRIBUTION AND CHANGE COMMUNICATION ............................................................................. 80

APPENDIX E 29BCREATING AN INTEGRATED ALL-ROADS NETWORK..................................................................... 82

E.1 51BDATA COLLECTION & CATALOGING ..................................................................................................................... 83

E.2 52BDATA EXTRACTION FROM INPUT SOURCES ........................................................................................................... 84

E.3 53BDATA PROFILING ............................................................................................................................................. 85

E.4 54BDATA TRANSFORMATION AND LOADING .............................................................................................................. 85

E.5 55BEDGE-MATCHING AND MATCH POINTS ............................................................................................................... 86

E.6 56BLRS AND NETWORK TOPOLOGY ......................................................................................................................... 89

E.7 57BOUTPUT DATASETS .......................................................................................................................................... 91

LIST OF ACRONYMS ............................................................................................................................................. 93

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 1 DOT Contract #GS-35F-0001P

September 2014

E X E C U T I V E S U M M A R Y

Although a rich body of work covering Linear Referencing Systems (LRS) and Geographic Information

Systems for Transportation (GIS-T) has been developed over the past 25 years, there is no national

consensus on LRS processes, data, or business rule standards. The Study team’s experience is that every

Department of Transportation (DOT) maintains a local, internal set of LRS rules, specific to their

organization and its business requirements. Moreover, those States that have begun to expand their

Geographic Information Systems (GIS) networks to encompass the all-roads requirements have, in many

cases, merely extended the LRS approach used on their State route network, which may or may not be

appropriate for local roads or multimodal applications. This approach is further complicated by the

functionality of various commercial off-the-shelf (COTS) packages, each of which provides a different

level of LRS support.

As a consequence of this evolutionary approach, no nationally endorsed or industry-wide LRS standard

practices or business rules have been officially and universally embraced. But certainly, there are many

existing local approaches to various LRS component-level issues that are satisfactory to meet specific

business needs. Therefore, the ARNOLD

Reference Manual is to be used as guidance, and is not

intended as a strict and enforceable standard. Its purpose is to report on the common conventions that

can be considered best practices, and to provide guidance for implementation.

This Reference Manual covers the four overarching steps for a statewide, all-roads LRS implementation

process, including:

 Implementation planning

 Data collection and integration

 Building the LRS

 Ongoing data maintenance

The content in this Reference Manual is based on interviews with several State DOTs and local/regional

agencies, as well as collaboration and discussion with the project expert panel, and is supplemented by

relevant subject matter research, all of which resulted in four individual reports that contain the findings

and recommendations of the All Public Roads Geospatial Representation Study.

While the four technical reports are comprehensive and detailed, the main body of this document is

synoptic and is aimed at walking a user through the overall process of planning and developing a

statewide, all roads network that includes LRS. This document highlights the most important content

All Road Network of Linear Referenced Data (ARNOLD)

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 2 DOT Contract #GS-35F-0001P

September 2014

from the other four technical reports in the context of an overall implementation process workflow,

while also providing Technical Appendices that comprise much of the more detailed material that was

developed for the individual stand-alone technical reports.

Most importantly, this document provides practical guidance and a handy Reference Manual to assist

state DOTs in moving forward to meet the new Highway Performance Monitoring System (HPMS)

requirements for the submittal of complete, all roads inventories and linear-referenced networks for

every State and territory. This requirement is known as ARNOLD – the All Road Network of Linear

Referenced Data. ARNOLD replaces the previous requirement of only collecting Federally Aided Route

networks from each State.

24BOVERVIEW OF RECOMMENDATIONS

Each section in the document contains specific recommendations pertaining to the topic covered in that

section (data collection, maintenance, etc.). The following list represents an overview of these

recommendations and represents themes that came up repeatedly throughout the Study:

 Collaborate with Stakeholders

 Other States, State agencies, local agencies, non-government entities, etc.

 Move Toward an Enterprise Approach

 Build it once, use for many

 Find Sustainable Practices

 For collection, maintenance, dissemination, etc.

 Expect and Manage Change

 Emphasize flexibility and scalability for data, linear referencing methods, software, etc.

 Build your LRS Incrementally

 Be realistic about current needs, and allow for the system to grow

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 3 DOT Contract #GS-35F-0001P

September 2014

1 0BI N T R O D U C T I O N

1 . 1 6BW H Y I S A N A L L R O A D S O U T L O O K I M P O R T A N T ?

Geospatial data for transportation is a key data theme within the National Spatial Data Infrastructure

(NSDI). The revision to the HPMS data submittal requirements that now require an "all roads network"

to be provided to U.S. DOT emanates from the simple fact that given today's technology and

transportation challenges, all roads datasets are needed by both the Federal government and the

States. Indeed, many States had developed and maintained all roads datasets long before this

requirement was formalized in 2012. Equally, and as documented in the U.S. DOT's 2011 Transportation

for the Nation0F

strategic plan, both the Federal government and States are already tracking and

managing infrastructure and activity that occur on all roads, such as bridges and accidents.

In addition, some of the most pressing transportation issues and concerns, such as safety, freight, aging

infrastructure and traffic management, demand nationwide data and an all roads outlook. The timing is

right for this evolution. This document aims to provide useful guidance on the planning, decisions and

approaches that will assist States in successfully meeting the new requirements.

Almost 100 years after the Federal Aid system was put in place through the Federal Aid Road Act of

1916, States and the Federal government are still working together to improve the transportation

infrastructure of the country1F

. In the early years, activity was focused on planning and constructing a

physical, national highway system based on the individual, yet coordinated, efforts of the States. In the

21st century, with modern technology and the increased use of data analysis to support planning and

management of the physical infrastructure, effort is focused on building a national road network dataset

that requires the same kind of coordinated work between the States and Federal government as

building the roads required. Indeed, this national road database will be an invaluable tool that will meet

current business needs while also paving the way for future advancements that range from Next

Generation 911 (NG911) and safety innovation to autonomous vehicles.

See: TFTN Strategic Plan

Earl Swift, The Big Roads: The Untold Story of the Engineers, Visionaries, and Trailblazers who Created the

American Super Highways, Houghton Mifflin Harcourt, Boston, 2011.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 4 DOT Contract #GS-35F-0001P

September 2014

1 . 2 7BW HY T H E U . S . D O T A N D F H W A N E E D A L L R O A D S

Requirements to meet the following business needs are driving the demand for all-roads LRS within the

U.S. DOT and FHWA:

 Certified Public Miles

 All public road centerlines

 Bureau of Indian Affairs (BIA) and Tribal delineations

 Fiscal Management Information System (FMIS)

 All public roads, including dual carriageways

 Highway project locations

 Bridge project locations

 Fatal and Serious Injury Crashes

 All public roads, including dual carriageways

 Link to Model Inventory of Roadway Elements (MIRE) and other safety data

 Freight

 Dual carriageways

 Truck network

 Traffic volumes and vehicle tracking

 Routing topology

 Performance Measures for Safety

 Crash locations by Urban Area and Metropolitan Planning Organization (MPO)

 Vehicle Miles Travelled (VMT) by Urban Area and MPO

 Performance Measures for Pavement

 Dual carriageways

 Pavement condition

1 . 3 8BW H A T I S T H E A LL- R O A D S G E O S P A T I A L R E P R E S E N T A T I O N S T U D Y ?

Developing and maintaining a statewide, all roads network that includes LRS is a complex, technical

endeavor. This Reference Manual represents the findings and guidance, both general and technical, of

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 5 DOT Contract #GS-35F-0001P

September 2014

the full All Roads Geospatial Representation Study. This study included four individual technical reports

that cover the activities necessary to realize the ARNOLD vision.

 Local Road Collection Systematic Approach Report

 LRS Components and Best Practices

 LRS Temporal Maintenance Plan Report

 LRS Technical Instructions, Rules and Diagrams Report

The Reference Manual is the culmination and compilation of the work done in these four interim

reports. In addition to the main body, it contains a set of Technical Appendices comprising details of the

topics covered throughout the main sections. It is assumed that the reader has a general understanding

of LRS, but if this is not the case, a basic introduction to LRS can be found in Appendix Section A.1.

This document is organized around the four key steps of a statewide, all roads LRS implementation

process, as follows:

Figure 1: All Roads LRS Implementation Process Diagram

2 1BI M P L E M E N T A T I O N P L A N N I N G

2 . 1 9BT H E O P P O R T U N I T Y T O R E V I E W T H E A G E N C Y ' S O V E R A L L N E T W O R K A N D L R S

D A T A M A N A G E M E N T

It is well understood that the development and maintenance of a statewide, all roads network

containing LRS is an involved and complex process. It is also understood that state DOTs may have a

variety of existing road networks and LRS that are in current use throughout the agency. In short, there

may be an existing and complicated data and LRS environment and adding yet another road network

and LRS can be viewed as a chore. At the same time, the new ARNOLD requirements provide an

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 6 DOT Contract #GS-35F-0001P

September 2014

opportunity for the DOT to review the existing data landscape and to have the ARNOLD requirements

catalyze a purposeful planning process that may go beyond simply building a new network, and may

involve a reconsideration of current practices. Options for approaching ARNOLD development include,

but are not limited to:

 Building a new network from scratch

 Adapting or enhancing an existing network

 Consolidating multiple existing networks into a single, multi-purpose enterprise resource

Ultimately, the new ARNOLD requirement can be viewed as an opportunity and reason for a State to

review its overall network and LRS data management approach and to make investments that address

what may be a backlog of known issues and challenges.

2 . 2 10BW H A T K I N D O F P L A N N I N G D O W E N E E D ?

Planning processes can take a variety of forms, and written plans can be built to cover various levels of

detail. For example, a plan to build a new single-purpose ARNOLD network would differ from a plan that

involved consolidating multiple existing LRS into a multi-purpose, enterprise dataset that may power a

variety of applications. As such, there is no single way that implementation planning should proceed.

Rather, the most important point is that planning needs to happen. It will then be up to the DOT to

determine the appropriate level of detail and the resources necessary to carry out the planning.

Regardless of the level of detail chosen, the following list presents the most important questions that

any planning process should answer:

What are the requirements? Datasets are not constructed for the sake of creating data; rather, the data

are created to support business requirements and to support planning and decision making. There are at

least two categories of requirements that the ARNOLD data should meet:

 FHWA HPMS submittal requirements: The HPMS program requires an annual data

submission of an all roads network that, among other things, can be used to validate a

State's road mileage figure.

 Additional business requirements: As documented in Appendix A.2, LRS are versatile and

can be used to support a wide variety of DOT activities and business functions (as seen in

Figure 2). These activities range from Transportation Improvement Planning (TIP) to safety

management and crash reporting to asset inventory and management. As DOTs plan

potential expansions or improvements to the LRS, it is critical to fully catalog and

understand all potential uses of the LRS.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 7 DOT Contract #GS-35F-0001P

September 2014

 What roles and responsibilities need to be

covered? Together, a statewide road network

and LRS are a complex database that changes

over time and requires human resources for

management. Additionally, as technology and

software continue to evolve there may be a

concomitant need for technical evolution of

the LRS. As such, planning for the LRS should

identify the human resource requirements,

the "organizational owners," and other

participants in managing and updating the LRS

on an ongoing basis.

Figure 2: DOT Business Functions

 Is there an established change management strategy? Constructing a statewide network and

LRS is not a one-time activity. Indeed, both the network characteristics (e.g., additions and

changes in road alignment) and the technologies available for managing, storing, and accessing

LRS-based data will change. As such, change management should be part of any implementation

planning exercise, with a focus on:

 Understanding and documenting the initial changes in current practices that are necessary

to develop the new, enhanced all roads network and LRS

 Designing with flexibility in mind so as to accommodate inevitable technological

advancement and change

 What are the desired outcomes of planning? The planning process will help the organization to

answer key questions and identify the resources that need to be marshaled to complete the

work of developing a statewide, all roads network. Several of the key issues that the planning

process will answer are highlighted in the succeeding sections of this report:

 Identify a data collection approach and process, including a repeatable updating process

(Section 3)

 Identify the data structure and underlying software for building and storing the network and

LRS (Section 4)

 Establish sustainable maintenance processes for keeping the data current and useful to all

stakeholders (Section 5)

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 8 DOT Contract #GS-35F-0001P

September 2014

2 . 3 11BI M P L E M E N T A T I O N P L A N N I N G B E S T P R A C T I C E S

Figure 3: Implementation Planning Key Recommendations3F

The recommendations below represent a synthesis and encapsulation of the best practices for

implementation planning gathered through research, interviews, and analysis.

 Work toward a shared, enterprise-wide LRS foundation within a State’s DOT. Rather than the

proliferation of different methods of LRS implementation within an agency, the all-roads

integration requirement is a rare opportunity to not only expand the roadway geometry under

consideration, but also move a DOT towards constructing and utilizing a single, multi-purpose

network and LRS across the network. This includes developing an improved institutional,

organizational, and procedural context surrounding the all-roads network – including a shared

LRS foundation. It should be noted that while moving to a single LRS may not be feasible in the

short term, minimizing the number of LRS in use is strongly recommended, and a single network

and LRS should remain a long-term goal.

 Assume that customer and business requirements and technology will change, so avoid over-

modeling the enterprise-wide LRS.

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 9 DOT Contract #GS-35F-0001P

September 2014

 Maintain a modern outlook  embrace change and facilitate adoption

 Monitor and control change to an appropriate degree to ensure the smooth operation of

interdependent systems (see next recommendation).

 Implement Change Management and communication processes for both organizational and

technical components of the LRS implementation and maintenance.

 Preparing for Change: Include activities to prepare the organization for the application of

change management strategies, to enable sponsors to support the change, and to help

architect a high-level change management strategy.

 Managing Change: Include the design of the change management plans and activities, and

the implementation of those plans throughout the organization. These plans will be

customized based on the characteristics of the change and the unique attributes of the LRS

and related organizations.

 Reinforcing Change: Include analysis of the results of the change management activities

and implementation of corrective actions. This phase also focuses on celebrating early

successes, conducting after-action reviews, and transferring ownership for change

management to the organization.

 Design flexibility and scalability into the core system so that temporal features can be added as

modular extensions of the core system.

 Employ a data structure that tracks inventory projects and roadway/route changes so that

questions regarding data changes can be answered.

 Recognize that many downstream users and business processes depend on LRS. Any

changes to the LRS will cascade down to them and may have unintended effects.

Understand these relationships during the design and development stage.

 Plan for education and training on LRS concepts, methods, tools and data objects, for both LRS

maintainers and end users.

 Proactively manage the ARNOLD deployment and manage predictable resistance with

education, training, and positive reinforcement.

 Adopt a customer orientation, with awareness and empathy for customer expectations.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 10 DOT Contract #GS-35F-0001P

September 2014

3 2BD A T A C O L L E C T I O N

The core difference between the previous HPMS road data submittal requirements and the new

ARNOLD requirements is that the State road network must now contain all roads within the State, not

just the Federally Aided routes. Thus, the core challenge for DOTs is identifying mechanisms and

repeatable processes for collecting the all roads data. State DOTs are not the only entities that map

roads within a State. Other local levels of government, such as counties and cities, are also involved in

road data collection and management. In addition, private sector companies collect and sell high-quality

road data. As such, there are significant opportunities for DOTs to partner with other entities to meet

the new requirements. The following sections lay out two key questions that State DOTs need to answer

as they embark on developing a statewide, all-roads network.

3 . 1 12BH O W D O W E C O L L E C T A L L R O A D S A C R O S S T H E S T A T E ?

There are four "local roads supply chain" patterns that can effectively deliver the information necessary

to build a statewide, all-road network. While each of these supply chains is feasible, they differ in how

important potential partnerships are, and also in the level of cash and direct DOT labor that may be

involved. The following information provides an overview of each of these supply chain patterns.

1. Local government supplies roads data to the State DOT: The DOT collects and assembles

centerline data from multiple governmental organizations, typically local and Federal

governments that have jurisdictional responsibility over some set of roads. Often, these

organizations have their own geospatial capacity and are already using geospatial technology to

manage their roads. At the local government level, these organizations typically include

municipalities and counties. At the Federal level, agencies such as the U.S. Forest Service,

National Park Service, Bureau of Land Management and the Bureau of Indian Affairs have

jurisdiction over the local roads in their geographic domains.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 11 DOT Contract #GS-35F-0001P

September 2014

Figure 4: Local Government Supply Chain Pattern4F

When this pattern is chosen, the core task is to establish outreach, communication and

collaboration with various partners. The communication is critical, and non-trivial amounts of

effort should be devoted to it so that a regular data exchange between partners occurs.

Nevertheless, collecting data on a regular basis is only the start of the process. This pattern also

requires that DOTs establish repeatable processes and workflows for assembling a cohesive

"whole" from the "parts" that are collected from local and Federal partners. Appendix E

provides detailed guidance on integrating local data into a statewide resource through

techniques such as: data profiling; data extraction, transformation and loading (ETL); edge-

matching; and the application of new LRS.

Pros:

 Highest data quality emanates from obtaining data from local sources that know the

landscape best

Cons:

 State DOT takes on the burden of data compilation and edge-matching

 Update and maintenance involves many stakeholders

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 12 DOT Contract #GS-35F-0001P

September 2014

 Communication and collaboration with local entities, particularly larger counties and

cities, can be difficult

2. Commercial and third-party road centerline data supporting a State DOT: The third-party

entity collects and aggregates road data from a variety of agencies and makes these data

available to the State DOT. This third party may be another government or quasi-government

agency (e.g., a regional Metropolitan Planning Organization (MPO), a State GIS clearinghouse,

State E911 program) or a commercial data supplier (e.g., HERE, TomTom, or Google). In

addition, this third party could be a publicly available data source such as OpenStreetMap5F

(OSM) or a Federal data source, such as the U.S. Census TIGER6F

files. In essence, the third party

takes on the role of gathering and assembling a statewide dataset from a variety of sources that

it chooses.

Figure 5: Geodata Supplier Supply Chain Pattern7F

See: Open Street Map website

TIGER stands for Topologically Integrated Geographic Encoding and Referencing. TIGER products are published by

the U.S. Census Bureau and contain features such as roads, railroads, and rivers, as well as legal and statistical

geographic areas. See U.S. Census Bureau TIGER Products webpage

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 13 DOT Contract #GS-35F-0001P

September 2014

Currently, several States, including Florida, Illinois, New York, and Massachusetts, have

developed relationships with commercial road centerline data suppliers. Others, such as

California, which uses TIGER, are using publicly available road data as a component of its

statewide, all roads networks. In addition, there is precedent for Federal agencies purchasing

commercially licensed street data, including the National Geospatial-Intelligence Agency (for the

Highway Safety Improvement Program) and the U.S. Geological Survey (for The National Map).

Pros:

 The State does not need to carry the full costs and business processes associated with

assembling the dataset, as the third party takes these on

Cons:

 State DOT does not have control over the data creation

 When a commercial supplier is involved, licensing restrictions can limit distribution

3. The State DOT does it all: As illustrated in Figure 6

, the DOT creates and manages the

statewide, all roads data layer on its own, irrespective of whether other agencies are also

managing centerline data. The DOT becomes responsible for identifying and accurately mapping

all new roads and other road changes (alignments,

names, etc.). Because the State is wholly

responsible, this method may require considerable

resources for original data collection and mapping

on top of just managing the technical aspects of the

dataset and LRS. In some States, such as Delaware,

there is not a choice, as the DOT is administratively

responsible for all public roads in the State.

Pros:

 The DOT is in complete control

Cons:

 Cost can be higher as the DOT takes on more

data collection and mapping

 Quality of data can suffer without proper local involvement to review and “ground

truth” the data

4. Hybrid approach: Given the three other patterns, a variety of hybrid approaches can be

pursued. Most typically, the DOT collects as much data as is available and useful from a geodata

Applied Geographics, Inc., 2014

Figure 6: State DOT creates and manages all data

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 14 DOT Contract #GS-35F-0001P

September 2014

supplier (e.g., a regional agency or State GIS clearinghouse) and then fills in gaps as needed

through its own efforts and by working directly with local and/or Federal government agencies.

In essence, the State DOT can choose one approach whereby it can collect the most data in the

best condition, and then uses additional tactics and efforts to fill in gaps or address

shortcomings. Other examples may include a State with a strong MPO that provides data for the

metropolitan area and then direct outreach to rural counties and Federal agencies for the less

developed parts of the State.

Figure 7: Hybrid Supply Chain Pattern8F

Pros:

 Blends the benefits of getting data from a strong third-party aggregator with having the

DOT remain directly involved in data collection from other partners

Cons:

 State DOT takes on the burden of data compilation and edge-matching

 Update and maintenance involves many stakeholders

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 15 DOT Contract #GS-35F-0001P

September 2014

3 . 2 13BW H A T C O M P O N E N T S W I L L O U R B A S E L I N E C E N T E R L I N E N E T W O R K

C O N T A I N ?

All road centerline datasets are not equal in their content. Indeed, part of the power of the road

centerline is its versatility and the ability for it to house a wide variety of related information. As

Appendix Section A.4 details, five key classes of information may be present in a statewide, all roads

network:

1. Road centerline geometry

2. Basic road attributes (e.g., road names)

3. Address ranges9F

4. LRS control

5. Network topology to allow routing

Figure 8 provides details on each of these key classes of information.

Figure 8: Common Baseline Network Requirements10F

Increasingly, address points are being collected for emergency dispatch and routing applications, since they

produce more accurate address-matching and geocoding results. If they are available, they are preferred to

address ranges.

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 16 DOT Contract #GS-35F-0001P

September 2014

Typically, more basic statewide networks will contain the first three components: geometry, basic

attributes and LRS. More advanced statewide networks will contain all five components. States that are

just embarking on their statewide, all roads networks may choose to start with a more basic set of three

components. Meanwhile, States that have had their own statewide, all roads networks for some time

and are contemplating the creation of more enterprise-oriented and multi-purpose networks may

choose to pursue all five components. (See Section A.4 for more on assessing network maturity.)

3 . 3 14BD A T A C O L L E C T I O N R E C O M M E N D A T I O N S

Figure 9: Data Collection Key Recommendations11F

The recommendations below represent a synthesis and encapsulation of the findings on best practices

gathered through research, interviews, and analysis. These recommendations provide an overall game

plan for effective approaches to collecting and integrating all-roads data into LRS that can be followed

by State DOTs and FHWA.

1. Create a conceptual framework based on supply-chain principles and best practices.

a. Define primary activities related to collecting and integrating all-roads data, and support

activities for a sustainable approach as part of the organizational approach.

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 17 DOT Contract #GS-35F-0001P

September 2014

b. Articulate the drivers, facilitators, components, and desired outcomes for the State, as well

as for other levels of government and other sectors that may be stakeholders or part of the

supply chain.

2. Reach out to non-DOT suppliers of all roads data, and treat them as true partners in meeting

requirements and creating bilateral benefits.

a. Make the effort to understand their capabilities and needs.

b. Identify mutually beneficial outcomes.

3. Jointly develop repeatable processes and/or systems for data exchange.

a. Consider updates more frequently than once per year.

b. Leverage the Internet and Web applications.

4. Be cognizant of the costs to local levels of government and the burden of, and resistance to,

unfunded mandates.

a. Unlike State DOTs, not all suppliers of road data are LRS-centric. This is especially true for

local governments, and many will not want to change their existing practices, especially if

new requirements are unfunded.

b. The key to a sustainable supply chain of local road data, flowing from local governments to

the State DOT, is to identify the mutually beneficial products of a partnership approach,

and to provide funding for activities that are uniquely required to meet HPMS reporting

requirements.

c. The State DOT also needs to be prepared to add the required value-added elements (edge-

matching, the addition of LRS, etc.) as a DOT function.

5. Understand related statewide initiatives for geospatial data sharing in general, and participate

as appropriate. For example:

a. A non-DOT government entity, such as the State GIS Office (or GIO

), may be coordinating

or partnering in the collection and distribution of all-roads data.

b. A non-government entity (e.g., commercial data provider) may be working in collaboration

with a non-DOT government entity, such as the Department of Public Safety, to collect and

maintain all-roads data (public and private).

c. Volunteered geographic information (VGI), such as Open Street Map (OSM), may be well

regarded in some States as a legitimate source of all-roads data.

Geographic Information Officer

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 18 DOT Contract #GS-35F-0001P

September 2014

4 3BI N T E G R A T I N G A L L - R O A D S A N D C O N S T R U C T I N G T H E L R S

Linear referencing systems are among the most important and complex datasets within a DOT. Thus,

great care needs to be taken in establishing new LRS or enhancing and extending the capabilities of

existing LRS.

This section highlights some of the key technical aspects of building LRS. The table below provides

summarized guidance for these technical details, along with the page number in the Technical

Appendices where additional background information, details, and diagrams can be found.

SUMMARIZED GUIDANCE FOR BUILDING LRS

ROADWAY GEOMETRY SUMMARIZED GUIDANCE

Roadway Segmentation

Implement an enterprise approach allowing multiple business

needs to be met. For example, maintain an intersection-based

network, and regularly generate the route-based network from it.

pg 51

Dual Carriageways

As defined, and in order to meet ARNOLD requirements, utilize a

dual-carriageway representation for divided roadways, ideally with

independent mileage calibration

pg 53

Traffic Circles

Model each traffic circle on a case-by-case basis, with the goal of

minimizing segment overlap and route segmentation.

pg 58

Ramps

Define the start and end of the ramp as the taper from and to the

mainline. Define deceleration and acceleration sections as LRS

events.

pg 61

Cul-de-Sacs and Loops

These roadway elements often have the same start and end point,

which can be problematic for at least one major vendor’s GIS

software to handle for LRS applications. The DOT will need to

establish standards for handling them consistently in the statewide

network, taking into account any software limitations.

pg 63

As described in the content in Appendix Section B.2, while the recommendation is for the mileage to be

independent, measures on both sides can be related. For example, as a road changes from divided to undivided

and back, a relationship between measures may be appropriate.

Tech. Appendix

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 19 DOT Contract #GS-35F-0001P

September 2014

SUMMARIZED GUIDANCE FOR BUILDING LRS

ROADWAY ATTRIBUTES SUMMARIZED GUIDANCE

Route Events vs.

Segmented Attributes

Store a minimum set of “base” attributes on the segment, and save

everything else as route events within the LRS.

pg 66

ARNOLD Schema

State DOTs should maintain or be able to generate the key ARNOLD

fields to meet submission requirements.

pg 67

Route ID Numbering

Define a standardized route identification convention as the

framework for aligning all DOT and local agency roadway asset

data.

pg 68

Road Naming

All roadways should include at least one standardized name.

Roadway naming should also include roadway aliases, historical

names, honorary names, etc.

pg 69

Multiple Linear Route

Measures

The GIS network should have the capability to support multiple

LRMs, while standardizing to a single LRM (such as driven mileage)

as the preferred measure.

pg 72

Public vs. Private

Roadways

Although the HPMS only requires the roads that correspond to

certified road mileage, State DOTs should include private roads in

their network to support emergency response and safety

considerations.

pg 73

Installation Date and

Inspection/Inventory

Date

For maximum data evaluation capability, capture and manage both

the construction date and inventory dates.

pg 74

Addressing

For emergency response purposes, discrete address point locations

linked to the LRS are preferable to give first responders an exact

location.

pg 75

LRS MAINTENANCE SUMMARIZED GUIDANCE

Metadata Standards for

GIS and Roadways Asset

Data

All published and distributed datasets should include standardized

metadata, ideally at both the layer level and the object level, but at

least at the dataset level.

pg 77

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 20 DOT Contract #GS-35F-0001P

September 2014

SUMMARIZED GUIDANCE FOR BUILDING LRS

LRS MAINTENANCE SUMMARIZED GUIDANCE (Cont.)

Planned, Destroyed and

Decommissioned

Roadways

Include planned, unbuilt facilities, as well as abandoned or

destroyed roadways, in the dataset.

pg 78

Geoarchiving Roadway

Segments

Always geoarchive data when significant updates and changes

occur.

pg 79

Roadway Data

Distribution and Change

Communication

Make data readily available to all users via web services, and

develop a consistent change communication mechanism.

pg 80

SUMMARIZED GUIDANCE FOR CREATING AN INTEGRATED ALL ROADS NETWORK

Data Collection and

Cataloging

Create a data inventory, including metadata, of all data sources to

be integrated.

pg 83

Data Extraction from

Input Sources

To streamline data loading and conflation, create a staging dataset

as needed for the ETL process, that contains the pertinent subset of

features from each source dataset.

pg 84

Data Profiling

Data should be evaluated for consistency and quality using a

combination of automated and manual procedures.

pg 85

Data Transformation and

When loading source data, only minor changes should be made

(e.g., re-projecting data, fixing obvious errors). Ideally, the source

data owner would take responsibility for needed data maintenance.

pg 85

Edge-Matching and

Match Points

Match points should be established to allow edge-matching and

data alignment between neighboring or overlapping transportation

agencies.

pg 86

LRS and Network

Topology

Topology rules and Open GIS Consortium (OGC) standards should

be applied to and enforced within the roadway network to ensure

data quality and stability, as well as to support routing and network

analysis.

pg 89

Output Datasets

The network should be built to meet the needs of routing, and then

be processed to support the needs of LRS.

pg 91

Tech. Appendix

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 21 DOT Contract #GS-35F-0001P

September 2014

The following section provides some focused and practical guidance for making the key decisions

necessary to build a statewide, all roads network of linear referenced data.

4 . 1 H O W D O W E C R E A T E T H E LRS T H A T W E N E E D ?

There are several key sets of issues, with attendant decisions that need to be made:

1. Managing both segmented and route-based road data

Traditionally, most GIS road networks are created and maintained in “segmented” form. That is,

if two lines intersect, each of those lines is broken at the intersection, or segmented. This is

useful since road characteristics can vary from segment to segment (e.g., the number of lanes

changes) and the intersection itself may have various characteristics to record (e.g., a “no left

turn” restriction). At the same time, most LRS are created and maintained in “route-based”

form. That is, each unique street name is stored as a route that has the complete geometry of

the entire street, from beginning to end and through all intersections. Typically, within LRS,

when two routes intersect, they are not broken into segments.

These two modes of storing road network data have evolved for good reason, based on

different use cases and capabilities. For example:

 Segment-based networks support details such as one-way streets and turn restrictions

at intersections, and these characteristics are critical in terms of vehicle routing and

emergency response.

Figure 10: Segment-Based Network Diagram12F

 Route-based networks are more traditional within a DOT’s LRS, as they enable roads to

be mileposted, from beginning to end, in a continuous fashion.

Michael Baker Jr., Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 22 DOT Contract #GS-35F-0001P

September 2014

Figure 11: Route-Based Network Diagram13F

Each type of network also uses a different approach for storing attributes. In a segment-based

network, attributes are stored as database fields associated with each segment. In a route-

based network, attributes are stored as events that are measured along the route (see Appendix

Section C.1).

Currently, most DOTs recognize that both types of networks are valuable and support different

use cases. For example:

 Segment-based networks support vehicle routing and are better for storing some types

of attributes, such as one-way streets

 Route-based networks support the storage of attributes such as pavement condition,

which may cover only a portion of a segment, and can be used to store point events

(e.g., an accident) that occur along a network

Understanding that DOTs need both types of networks, the challenge becomes developing a

data maintenance workflow that doesn’t involve the need to complete an edit twice (i.e., once

in the segment-based network, and again in the route-based network). Thus, the recommended

approach is to implement an enterprise road dataset that contains both segment geometry and

comprehensive LRS that can meet multiple business needs. One approach for achieving this

would involve the following (see Figure 15 as well as Appendix Sections B.1 and E.7):

 Maintain the segment-based network for the base geometry and enter all changes (e.g.,

new roads, realigned roads, etc.) into the segment based network

 Use geospatial software, ideally automated routines, to regularly generate the route-

based network as a derivative of the segment based network

Michael Baker Jr., Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 23 DOT Contract #GS-35F-0001P

September 2014

2. What linear referencing method(s) (LRM) will we use? Do we need more than one?

One of the key characteristics of LRS is the ability to store “measures” along the network. A

measure allows locations along the network to be described in a unique way. For example, a

culvert could be described as existing “4.62 miles from the beginning for Route 495”. This

example uses a specific “linear referencing method” (LRM) for identifying the location of the

culvert. In this case, the LRM is “the absolute distance from the start of the road”.

There are several different LRM besides “absolute distance” and (Appendix Section A.3 provides

details on the most common LRMs in use by DOTs):

 Absolute: Distance from the start of the route segment (e.g., 4.62 miles)

 Relative: Distance from a reference location (e.g., 292 feet from milepost 101 on Route

495)

 Interpolative: Proportional distance from start of segment (e.g., 68.2 percent)

 Addresses: Can generally be done in two ways (see Appendix Section C.8 for more

details):

o Address Range: estimated distance based address range of a segment

o Address Points: location of an actual, measured address location

 GPS route: Measured Global Positioning System (GPS) coordinates are projected onto a

segment/route in the network

Ideally, the statewide, all roads network should have the ability to support multiple LRMs, while

the DOT standardizes on a single LRM as the preferred, default measure (see Appendix Section

C.5). As such, identification of all of the LRMs in use by a DOT, as well as the most frequently

used ones should be an important aspect of planning the statewide, all roads network .

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 24 DOT Contract #GS-35F-0001P

September 2014

Figure 12: Multiple Linear Route Measures Diagram14F

3. How will the LRS handle the most challenging geometric roadway elements?

Roadway networks can be extremely complex, and as highway construction and traffic

management techniques continue to evolve they will continue to increase in complexity.

Initially, digital representations of roadway networks, particularly those designed to house LRS,

were simplified, schematic representations. That is, every road was represented as a single line,

and every intersection was depicted as a single point/node where two lines intersected.

However, as technology has advanced and as the uses of electronic roadway data have

broadened, it has become increasingly important to more accurately depict the layout and

alignment of roadway networks.

As more State DOTs perform work on their road networks to meet the new ARNOLD

requirements, it may be appropriate to improve and enhance the existing networks to not just

contain all roads, but also to include more accurate roadway configurations. As detailed in

Sections B.2 – B.5 of the Appendix, the following describes some of the more challenging road

configurations that need to be modeled to create the most accurate LRS possible. Properly

handling these situations will help DOTs develop the most

accurate possible statewide roadway mileage by use of

their all roads network.

 Dual carriageways: Store divided roadways as two

separate segments with each direction having its

own measurements

(see Appendix Section B.2).

Michael Baker Jr., Inc., 2014

Graphic by Michael Baker Jr., Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 25 DOT Contract #GS-35F-0001P

September 2014

 Traffic circles/rotaries: Model each traffic circle on a case-by-case basis, with the goal of

minimizing segment overlap and route segmentation (see Appendix Section B.3).

 Ramps: Model as a special form of

intersection containing unnamed

segments

(see Appendix Section B.4).

 Cul-de-Sacs and Loops: These features

often have the same start/end point, which can be problematic for LRS. The DOT will

need to establish standards for handling them consistently in the statewide network

(see Appendix Section B.5).

4. Creating a seamless network using edge-matching and match points

Match points (also known as integration points, touch points, smart points, demarcation points,

agreement points, snap-to points, join points, etc.) are point locations established within the GIS

to mark the connection point between two (or more) geospatial datasets. These points allow

datasets to be seamlessly joined together without any overlap or gaps (which is essential to

network topology, as described in the section below). In terms of a nationwide ARNOLD,

establishing these points between neighboring States will be critical in facilitating the edge-

matching of data and ultimately stitching together a nationwide roadway dataset.

Graphic by Michael Baker Jr., Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 26 DOT Contract #GS-35F-0001P

September 2014

Figure 13: Edge Matching Scenarios15F

If match points to facilitate data integration have been agreed upon at the State or local levels,

they should be used. If they do not exist, then a set of recommended points should be

presented to the affected jurisdictions for negotiation and agreement. Feedback and

adjustments should be allowed for, and incorporated into an agreed-upon Statewide Match

Point Layer (see Appendix E.5 for more detail).

4 . 2 16BW H A T T O O L S D O W E N E E D T O C O N S T R U C T A N D M A I N T A I N L R S ?

The geospatial software industry - both for computer-aided design (CAD) and geographic information

systems (GIS) - has consistently advanced the toolsets that are available for developing, managing and

maintaining both linear networks and LRS. In short, a variety of commercial, off-the-shelf (COTS)

software solutions can provide the tools a State DOT needs, and most of these can be extended with

customization for particular situations in a given State.

The following list provides an overview of the core software capabilities that are necessary for the

construction and maintenance of a statewide, all road network and LRS:

1. Constructing and Maintaining the Centerline

 Geometric editing of the centerline data: Having the core capabilities to create and edit

data and to maintain network topology ensures that new roads can be added, obsolete

See: Esri, ArcGIS Resources, About Edgematching

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 27 DOT Contract #GS-35F-0001P

September 2014

features can be removed (and archived), and network connectivity and attributes can be

properly maintained.

 Data import/export from/to common, standard formats: These tools are particularly

important when the chosen supply chain involves the collection and integration of data

from partners and other third parties.

 Extract, transform, and load (ETL): These tools are also particularly important in the

process of integrating data obtained from multiple sources into a single statewide

dataset. The ETL process may involve taking data from one format and running it

through conversion routines that prepare it for loading into another dataset, in another

format.

 Conflation: This feature involves the ability to transfer the geometry and/or attributes

from one dataset to another, including edge-matching functionality.

 Multi-user editing and versioning: Given the size of statewide networks, it is highly

desirable to have a software environment that enables multiple people to edit the same

network simultaneously. When this takes place, advanced features such as feature

locking and data versioning (i.e., the ability to track and manipulate multiple versions of

the same dataset) become increasingly important.

2. Applying and Maintaining the LRS

 LRM calibration: The baseline geometry of networks can change over time, or wholesale

improvements can occur in response to an event such as a new flyover. When the

underlying geometry changes, tools are necessary to re-calibrate the LRM to the new

geometry and to allow fixed assets, such as mileposts, to maintain their positions.

 Applying an LRM: This capability involves taking a baseline geometric network and

applying the LRM so that it can calculate, house, and maintain measure-based values.

 Storage of, and access to, measure-based information: Once the LRM is applied, the

software needs to be able to house derivative datasets/features that are based on

measurements. Typically, these additional features are stored as "events" that

reference the LRS. Thus, a user can access and manipulate datasets of "accidents" or

"culverts" or "pavement conditions" based on their measured values.

3. Publication and Sharing of LRS Data

 Ability to publish web services: Increasingly, routine end user access to data of all

types, including LRS and derivative measures, is via web browser-based applications,

including access on mobile devices. As such, it is important that the chosen software

environment is able to publish the data as web services that can be consumed by

browser-based applications, mobile applications, and by many desktop geospatial

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 28 DOT Contract #GS-35F-0001P

September 2014

environments. For greatest flexibility, the publication environments should support

open geospatial standards such as the Web Map Service16F

(WMS) from the Open

Geospatial Consortium17F

 Programmatic access to LRS via APIs: Like web services, Application Programming

Interfaces (APIs) are an important tool for making LRS and measure-based data

accessible through web browser and mobile applications. Unlike web services, which

provide access to raw data, an API can provide tools to manipulate and query the data,

thus providing expanded capabilities to application developers.

 Download of LRS information: Public availability of road network and LRS data is

important, and DOTs should anticipate creating a capability for public download, or

adding road centerline and LRS data to existing download capabilities. Broadly speaking,

the download capability can be considered an extension of the process of providing the

final data products to the HPMS program.

Ultimately, building and maintaining a statewide, all roads network is an involved process. As described

above, a variety of tools are required to perform the three core functions of centerline creation and

maintenance, application and management of the LRS, and the publication and use of LRS and measure

data. While some toolsets may be able to meet all of the requirements of State DOTs, it is feasible and

can be beneficial to combine tools to create “best of breed” solutions. For example, some tools are

highly specialized for activities such as ETL or high-performance web publication, and other tools are

tightly focused on the maintenance and management of LRS and measure data.

See Open Geospatial Consortium, Web Map Service

See Open Geospatial Consortium Standards

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 29 DOT Contract #GS-35F-0001P

September 2014

4 . 3 17BR E C O M M E N D A T I O N S F O R B U I L D I N G T H E L R S

Figure 14: Building the LRS Key Recommendations18F

The recommendations below represent a synthesis and encapsulation of the LRS best practices gathered

through research, interviews, and analysis.

 Build LRS incrementally. Due to its foundational nature, the all-roads LRS must be developed

with greater care and accuracy than almost any other data within a DOT. Practically speaking,

the magnitude of this effort may be somewhat mitigated using an incremental approach.

Ideally, the initial design would outline the ultimate LRS configuration, which would then be

incrementally achieved using a series of intermediate projects. Given the all-roads HPMS

reporting deadline, an incremental strategy may be a practical necessity.

 Give proper consideration to specialized roadway elements, such as dual carriageways, traffic

circles, and ramps. For example:

 Dual carriageways necessitate two or more sets of linework to adequately represent the

roadway geometry. As defined, and in order to meet ARNOLD requirements, utilize a dual-

carriageway representation for divided roadways, ideally with independent mileage

calibration.

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 30 DOT Contract #GS-35F-0001P

September 2014

 Traffic circles should be represented in a way that matches their use. The smaller, local road

traffic circles are best modeled in a simple way. Larger and more complex traffic circles may

require a more detailed linework representation.

 Defining ramps can be a challenge due to their ambiguous nature. Define the start and end

of the ramp as the taper from and to the mainline. Define deceleration and acceleration

sections as LRS events.

 Focus on interoperability when implementing LRS and LRM. It is not advisable that State all-

roads LRS efforts perpetuate the non-interoperable silos of the past.

 One way to achieve improved interoperability is to have a smaller number of permissible

LRMs.

 Interoperability is key, in terms of both the LRMs and the software tools.

 Current business rules are a key driver of LRS software. LRS software choices within an agency

are typically driven by existing practices and workflows.

 Whenever possible, pursue software and technology choices that match the existing

practices of the organization.

 It can be easier to implement a new technology than to alter an established business

practice within a large agency.

 When measuring mileage, actual driven measures that account for elevation and other

variability in roadways are more accurate than calculated measures. Since mileage is certified

for HPMS reporting purposes, this is an important consideration in terms of verification.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 31 DOT Contract #GS-35F-0001P

September 2014

5 4BO N G O I N G D A T A M A I N T E N A N C E

As described above, statewide, all road networks are inherently complex to create and are vital to State

DOTs for a wide variety of business purposes. This innate complexity carries over to the maintenance

activity, especially since physical roads are in a constant state of change based on new construction and

development. Thus, it is critically important that building the statewide, all roads network and LRS not

be considered a one-time task. Rather, regular maintenance and updates need to be considered a

fundamental part of an overall statewide, all roads data program.

5 . 1 18BW E ' V E S P E N T A L L T H I S E F F O R T B U I L D I N G I T ; H O W D O W E K E E P I T

C U R R E N T ?

There are at least three components to a statewide, all road network and LRS, and each of these may

change; thus, some level of updating attention is required for each component, including:

 The baseline centerline geometry

 Route system topologies that may be derived from the segmented centerline

 Multiple LRS/LRM that are applied to the route system, and measured features derived from the

LRS

And, there are four key considerations when planning for or developing a program for LRS maintenance:

1. Identify actions/activities that trigger a need for maintenance

First, external events emanating from the DOT or from other road-building authorities in the

State may prompt a need for LRS maintenance. These events include:

 New road construction by the DOT or a local authority

 Construction that impacts alignment/roadway geometry

 Roadway name changes

 Other attribute changes (speed limit, number of lanes, etc.)

Second, internal DOT events may prompt LRS maintenance, including:

 Improved base map accuracy (e.g., through a new flyover that allows a more accurate

representation of the linear geometry)

 Improved geometry (e.g., adding dual-carriageway representation)

 Routine error identification and corrections based on user reports

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 32 DOT Contract #GS-35F-0001P

September 2014

The external events typically involve "feature by feature" maintenance to make sure individual

changes are represented in the network. The internal events may involve wholesale changes

that impact the entire dataset or large pieces of it, such as improving the geometry for all

divided highways.

2. Establish LRS maintenance best practices

It is strongly recommended that DOTs pursue an enterprise approach to their centerline and

LRS data development and management. To the extent practical, DOTs are well served by

moving to a single (or reduced number of) multi-purpose, enterprise road centerline and LRS.

Indeed, it is in the maintenance process where the largest payoff to this approach is realized. If

done properly, when roads change, that change will only need to be recorded once in the

enterprise road dataset. Otherwise, that change would need to be repeated in each of multiple

road centerlines and LRS.

Figure 15: Enterprise LRS Maintenance (base geometry supporting multiple business cases)19F

As discussed in Section 4.1 of this document as well as in Sections B.1 and E.7 of the Technical

Appendices, the ultimate goal is to maintain a single geometry that supports multiple business

cases (i.e. navigation/routing, as well as DOT/LRS). As depicted in Figure 15, node and segment

geometry are needed for the creation and maintenance of a roadway network (e.g., adding new

routes or new alignments). This geometry, along with its network topology, can be combined

with turn and flow restrictions and address points to satisfy routing and navigation use cases.

Similarly, LRS routes can be derived from the same updated roadway geometry and network

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 33 DOT Contract #GS-35F-0001P

September 2014

topology. These newly derived routes can satisfy linear referencing use cases when combined

with point and line events along the LRS.

Under all maintenance scenarios, it is a best practice to record and maintain metadata that

describes the origins and maintenance history of the road network. As described in Appendix

D.1, it is optimal if Metadata Standards are followed. Best practices imply that all

published/distributed datasets should include standardized metadata, ideally at both the layer

level and the object level (e.g., individual road features within the dataset).

3. Emphasize collaboration with stakeholders and data suppliers

As described throughout this report, there are two key kinds of collaborators:

1. Collaborators who contribute data to the statewide, all roads network as part of the

supply chain

2. End-users, both inside and outside of the DOT, that utilize the LRS but may not be

directly involved in its development, management and update

It is critical for the first group of "supply chain collaborators," to continue to remain involved in

the updating process as part of the supply chain, by providing data on the new and newly

aligned roads within their jurisdiction. Achieving this goal will require clear communication and

ongoing outreach for data exchanges.

For the second group of "business user collaborators," it needs to be recognized that many

downstream users and business processes are dependent on LRS. Many types of changes to the

LRS cascade down to them and may have unintended impacts (e.g., calculated measures may

need to be re-calculated if alignments are changed). These kinds of relationships need to be well

understood, and once again, regular and active communication to the user community must

occur when updates are made.

4. Data distribution and change communication

As detailed in Appendix D.4, it is important to make data readily available to all users via web

services, and to develop a consistent change communication mechanism. Ultimately, one of the

major benefits of web services is that changes are automatically pushed to all users of the

service. In other words, the end user does not need to do anything special to access the latest

data. While it remains important to support a download capability, one shortcoming is that

users need to remember to periodically download the latest data that reflects changes.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 34 DOT Contract #GS-35F-0001P

September 2014

There are three important best practices for change communication:

1. Establish a readily accessible change log to allow users to review and understand the

changes that have been made. Users who require download would review the change

log to determine when downloading a new copy of the data is beneficial.

2. Establish a means to collect and track change requests from users. Ultimately, the

regular users of the data are in the best position to detect errors or inaccuracies, and

they should be encouraged to report what they find so that those issues can be

addressed in future update cycles.

3. Proactively notify users when changes occur to enhance awareness.

5 . 2 19BM A N A G I N G T E M P O R A L I T Y W I T H I N T H E L R S

Temporality involves notions of time. In the LRS context, this means storing information about roadway

characteristics over time as part of the database. State DOTs routinely face questions about roads that

involve a time element. Examples of these questions include the following:

 Where are all the accidents within this construction boundary that occurred during the

construction period from June 2012 through October 2013?

 Where are the locations of all the accidents that occurred after the construction project was

completed in February 2014?

 Where are all the current road closures and temporary detours? What roads were closed on

December 15, 2012?

 What was the Annual Average Daily Traffic (AADT) for this route in 2010?

 What was the total statewide road mileage in 2012? In 2013?

Unlike Section 5.1, which describes techniques and activities for managing change within the LRS itself,

such as data updates and accuracy improvements, temporal LRS involves techniques for tracking and

archiving changes within the physical road systems as depicted by the centerline network and LRS. For

example, roadways may be planned, under construction, in use, or demolished at different points in

time. Routes may be renamed, reclassified, or transferred to other jurisdictions over time. Pavement

and bridges may have different condition indices as they wear-down over time and are refurbished or

replaced.

As such, it is key that the planning and development of a statewide, all road network consider how

temporal changes can be stored and managed. The following kinds of DOT programs require temporal

information:

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 35 DOT Contract #GS-35F-0001P

September 2014

 Travel demand forecasting

 Highway planning

 Asset tracking and management

 Construction project management

 Right-of-way (ROW) and property acquisition and disposal

 Crash reporting and safety analysis

When planning the LRS, it is important to design flexibility and scalability into the system so that

complex data, such as temporally based information, can be added over time and as the LRS matures. As

detailed in Appendix Section C.7, the most basic storage of temporal data can involve adding

appropriate attribute tables and fields to the segment-based road network; such fields could include:

 Construction date

 Inspection date(s)

 Maintenance date(s)

As detailed in Appendix Sections D.2 and D.3, more advanced incarnations of temporal data storage

include:

 Geometric features for planned (i.e., “paper streets”), destroyed, and decommissioned

roadways in the statewide, all road network. These features should be readily identifiable

through their attributes and could be either included or filtered out, depending on use.

 Development of a “geoarchive” of the road network and LRS that would enable users to go back

in time to view the entire dataset as it existed previously. Typically, geoarchives are created by

taking snapshots of the road geometry and associated LRS on a regular basis (monthly,

quarterly, annually, etc.) and then storing and providing access to them. A comprehensive

geoarchive over an extended period of time would enable the DOT to build an animation that

shows the development and evolution of the entire road network.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 36 DOT Contract #GS-35F-0001P

September 2014

5 . 3 20BLRS M A I N T E N A N C E R E C O M M E N D A T I O N S

Figure 16: LRS Maintenance Key Recommendations20F

The recommendations below represent a synthesis and encapsulation of the LRS maintenance best

practices gathered through research, interviews, and analysis.

 Recognize that many downstream users and business processes depend on LRS. Any changes

to the LRS will cascade down to them and may have unintended effects. Understand these

relationships during the design and development stage.

 Determine maintenance responsibilities internal to the enterprise. Often, although not always,

enterprise LRS maintenance responsibilities are assigned to the group responsible for base

mapping maintenance.

 Consider data sharing and inter-governmental collaboration on LRS maintenance activities.

Although the road system changes every year, the workload for LRS maintenance is not uniform

statewide.

 Almost all LRS maintenance (new or realigned roads) is driven by changes to local roads and

minor collectors in the system.

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 37 DOT Contract #GS-35F-0001P

September 2014

 Since local roads are usually the exclusive responsibility of local governments and are not

eligible for Federal aid (i.e., the DOT is not involved with their design or construction), inter-

governmental data-sharing relationships may be the most efficient way to perform much of

the required LRS maintenance.

 Track and maintain key dates at the individual asset level, including the date of installation (or

construction) and the date of inventory.

 The date of installation (or construction) is when the asset was built; this date may become

less critical if the data maintenance strategy is a periodic full asset inventory, but it can be

important for deterioration modeling and predicting asset lifespans.

 The date of the inventory is the date of the field observation of the asset or object; when

using manual feature extraction from collected roadway photos, the stored date should be

the date of the photo capture.

 Classify updates in terms of the nature of the data changes (e.g., roadway changes or improving

spatial accuracy) and the type of update (e.g., geometry or attributes), so that they can be

handled accordingly. In some cases, field verification might be required.

 Maintain historical (i.e., archived), current (i.e., production) and proposed (i.e., development)

versions of the LRS. The proposed alignments can go ”live” – a real-time update – as soon as the

road is open for traffic.

 Establish procedures for geoarchiving (i.e., storing snapshots of LRS data) and geopublishing

(i.e., disseminating LRS data to support various business needs).

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 38 DOT Contract #GS-35F-0001P

September 2014

6 5BC O N C L U S I O N S A N D T H E P A T H F O R W A R D

The All-Roads Geospatial Data Representation Study (the Study) was focused on the challenges faced by

State DOTs to gather and integrate all roads data. The Study resulted in a set of individual technical

reports and a final Reference Manual (this document) that provide guidance and best practices (not

strict rules or formal standards) to State DOTs for implementing and maintaining an All Road Network of

Linear Referenced Data (ARNOLD).

In terms of timing, the requirement for ARNOLD data submittals as part of the HPMS reporting process

preceded the Study. Initial data submittals toward meeting this new requirement began in June 2014,

before the ARNOLD Reference Manual was published. Thus, the guidance in this Reference Manual is

relevant for the June 2015 HPMS submittal cycle.

While these submittals are for individual States, there is still the national challenge of creating an all-

roads network for the nation from these data as the basis for the Transportation data theme for roads

as part of the National Spatial Data Infrastructure (NSDI) – i.e., Transportation for the Nation (TFTN). The

Study focused on the creation and maintenance of all-roads networks within a State, and not on the

challenge of conflating the data from the 50 States and the Territories into a national network;

therefore, work is still required in this regard.

Based on feedback from the Expert Panel assembled to provide advice during the Study, the following

suggestions for “next steps” were made:

For States:

 Leverage the AASHTO

GIS-T Survey to self-assess, to determine progress toward meeting the

ARNOLD findings and recommendations from the Study, such as:

 What data collection type (i.e., supply-chain pattern) do you use, and who are your

partners?

 Have you sufficiently collected data for all roads, including local roads?

 Have you built repeatable processes to update and maintain the statewide all-roads LRS?

 Have you adequately addressed the technical challenges and requirements of ARNOLD?

 Have you gone beyond the base requirements to build a robust and sustainable statewide

LRS?

American Association of State Highway and Transportation Officials

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Page 39 DOT Contract #GS-35F-0001P

September 2014

 Based on the State’s self-assessment, begin a planning process for long-term success, for

example:

 Understand what resources (e.g., funding sources), besides the technical guidance, are

available to support all-roads implementation effort (e.g., Highway Safety Improvement

Program funds)

For FHWA:

 Profile the June 2014 HPMS all-roads submittals to understand the data issues for moving

forward with conflating multiple States into a national network

 Provide clear feedback to State DOTs on enhancing their data where needed

 Create a collaborative forum for State DOTs to share their experiences and ideas on ARNOLD

implementation

 For example, build a website where each State has a log-in and can share ideas, input “self-

assessment” information, and connect with other States that fall into the same categories

(collection pattern, maturity level, etc.).

 Publish information on available resources (including funding sources) for ARNOLD

implementation

 Analyze findings from the individual State DOT self-assessments on the ARNOLD requirements,

and perform overall maturity assessments

 Publish aggregated information based on overall findings

 Tighten ARNOLD specifications where appropriate and needed

 Update the ARNOLD guidance based on “day forward” implementation discoveries

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 40 DOT Contract #GS-35F-0001P

September 2014

T E C H N I C A L A P P E N D I C E S

APPENDIX A 25BBASICS OF LRS ............................................................................................................................. 41

APPENDIX B 26BROADWAY GEOMETRY ............................................................................................................... 51

APPENDIX C 27BROADWAY ATTRIBUTES .............................................................................................................. 66

APPENDIX D 28BMAINTENANCE ........................................................................................................................... 77

APPENDIX E 29BCREATING AN INTEGRATED ALL-ROADS NETWORK..................................................................... 82

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 41 DOT Contract #GS-35F-0001P

September 2014

Appendix A 25BB A S I C S O F L R S

A.1 30BO V E R V I E W O F L I N E A R R E F E R E N C I N G S Y S T E M S ( L R S )

Linear referencing is a method for storing and managing geospatial information along a linear feature,

with positional location defined by a distance measure along that linear feature. LRS is most frequently

implemented for roads and highways. In LRS, the locations of both data and events are determined

according to their distance along a road from some known point (e.g., the beginning of the road, a mile

marker, or an intersection). Linear reference measures can also be applied to other types of linear

features, such as bus routes, railways, waterways, pipelines, or power lines.

LRS is crucial for managing the vast and varied data collected and maintained by a DOT. As stated by the

AASHTO Technology Implementation Group (TIG), “The Linear Reference System (LRS) aligns the linear

reference points in all databases so information from crash statistics, pavement management, and other

business data can be accurately mapped and data more easily analyzed.”21F

The figures below depict first the base centerline geometry, segmented at intersections (see Section B.1

for more on this), and then this same geometry with LRS event data overlaid on it. In Figure A.2,

pavement condition data and accident locations are stored as LRS events in tables (also shown), which

are displayed on top of the base geometry.

Figure A.1: Example of Base Centerline Geometry22F

See AASHTO Innovation Initiative - Linear Referencing System

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 42 DOT Contract #GS-35F-0001P

September 2014

Figure A.2: Example of Base Geometry with LRS Events23F

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 43 DOT Contract #GS-35F-0001P

September 2014

A.2 31BL I N E A R R E F E R E N C I N G S Y S T E M B U S I N E S S F U N C T I O N S

Long before the widespread use of GIS technology, transportation organizations used linear referencing

methods to measure distances, describe routes, and locate objects along transportation routes such as

roadways, railways, and waterways. The approaches have been as diverse as using stationing and

staking methods during the design and construction of new facilities, to using mile marker signage and

straight line diagrams24F

(SLDs) as a framework for road inventory purposes. The large number of linear

referencing methods that have been developed over time, such as those described above, are a good

indicator of the ubiquity of LRS throughout the transportation industry, particularly in the DOT sector.

Although many classifying taxonomies for LRS use have been developed over time, the simplest and

most prevalent can be derived from the commonly understood transportation system life cycle.

Straight Line Diagrams (SLD) are linear depictions of roads and intersecting features (e.g., intersections, fixtures,

structures) where the line does not include alignment geometry (i.e., curves). The distances between features may

or may not be to scale. The route-based segments discussed in the text generally include scaled alignment

geometry.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 44 DOT Contract #GS-35F-0001P

September 2014

Figure A.3: Transportation System Life Cycle25F

Each of these major life-cycle functions is supported by multiple business processes, which use and/or

share LRS. The following uses provide a sample of the wide variety and diversity of LRS used in State

DOTs today. Several of these, such as crash reporting, bridge inventory, and HPMS, already embody the

Federal all-roads requirements. Others, such as public information, over-legal permitting, incident

detection, and dispatch systems, commonly encompass an all-roads perspective as well.

58BBUSINESS FUNCTIONS THAT UTILIZE LRS

Planning

 Safety Management, including crash location and reporting

 Traffic Counting

 Travel Demand Modeling

 Corridor and System Planning

This diagram is based on several industry sources, including: Wisconsin DOT Enterprise Information Strategy Plan

(1991); Utah (1995) and Idaho (2000) Information Architectures; AASHTO Pooled Study business process model

(1993); and the National Cooperative Highway Research Program (NCHRP) 20-27 project.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 45 DOT Contract #GS-35F-0001P

September 2014

Program Development

 Environmental Impact Analysis

 Transportation Improvement Program (TIP)

Program Delivery

 Surveys

 Roadway Design (CAD)

Operations

 Winter Maintenance

 Over-Legal Permitting

 Public Information (511) Systems

 Incident Detection and Emergency Dispatch

Asset Management

 Pavement Inventory, including pavement condition surveys

 Bridge Management Systems

 Intermodal Management Systems

 Local Road Aids

 HPMS and other Highway Inventory Systems

 Videolog/Photolog

 Signing/Marking Inventory

In addition to supporting these and other business functions, transportation organizations are using LRS

(and related GIS-T26F

technology, such as dynamic segmentation) to support enterprise-level data

management, data integration, and data fusion initiatives27F

– all of which make an all-roads LRS network

more valuable.

GIS-T is the term used coined by Fletcher and Lewis in a TRB workshop circa 1986 and adopted by the American

Association of State Highway and Transportation Officials (AASHTO) circa 1987 to distinguish general-purpose

Geographic Information Systems (GIS) from those devoted to Transportation (T).

The term “data fusion” simply refers to the merger of different data, often from different sources, into a unified

dataset, using a variety of methods, depending on the data.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 46 DOT Contract #GS-35F-0001P

September 2014

A.3 32BT Y P E S O F LRMS

A Linear Reference Method28F

(LRM) defines a specific way in which locations are described (i.e.,

measured) along linear geographic features such as roads, railroads, and bus routes. While the features

themselves do not need to be abstracted as linear geometry, an LRM must support measurements in a

one-dimensional linear sense. Over time, the term LRM has come to refer to real-world measurements

using various kinds of instruments (e.g., Electronic Distance Measuring, odometers, 5th-wheel sensors)

and also refers to measurements taken along cartographic linear features (e.g., polylines, curves, and

directed edges) using specialized geospatial software algorithms.

As documented by the International Organization for Standardization (ISO) Standard on Linear

Referencing,29F

All LRMs can be characterized as belonging to one of three types: absolute, relative, and

interpolative.

Absolute methods measure the total distance from the start of the segment to the event. Absolute

methods include Mile Point and Project stationing where the start point is location 0.0 and the end

value is equal to the total route or project distance.

Figure A.4: Absolute LRM30F

Relative methods locate events according to their distance from a known reference location. Examples

include Mile Post, Reference Post, and Feature-based (i.e., literal description, such as an intersection).

The concepts embedded in Linear Reference Methods are not synonymous with Linear Reference Systems, and

the terms are not synonymous with the much broader topic of Location Reference Systems. A Linear Reference

System typically encompasses multiple methods, plus the office and field procedures necessary to establish,

maintain, and use each method, and also includes the knowledge, skills, experience, and technology involved with

linear referencing. Location Reference has a still larger scope, encompassing all location issues in all dimensions

used within an agency.

ISO 19148:2012 - Geographic information -- Linear referencing

Adopted from a 2012 GIS-T presentation by Paul Scarponcini, “NCHRP 20-27 to ISO 19148: 18 Years of Progress

in Linear Referencing.”

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 47 DOT Contract #GS-35F-0001P

September 2014

Figure A.5: Relative LRM31F

Interpolative (i.e., proportional) methods measure distance as a fraction of the entire section distance.

Examples include Percentage, Normalized, and M Values.32F

Figure A.6: Interpolative LRM33F

A.4 33BC O M M O N B A S E L I N E N E T W O R K R E Q U I R E M E N T S & M A T U R I T Y A S S E S S M E N T S

From a DOT perspective, when aligning supplied data with business needs, the level of maturity can be

assessed by evaluating the following components of local road data:

 Geometry (completeness, dual carriagemway representation, update cycle, scale, etc)

 Existence of basic attributes

 Existence/type of address information

Adopted from a 2012 GIS-T presentation by Paul Scarponcini, “NCHRP 20-27 to ISO 19148: 18 Years of Progress

in Linear Referencing.”

M values (or local interpolative methods) are a way of moving between map (i.e., page) distances and scale (i.e.,

real world) distances by interpolating linear measures in either map units or page units along a single polyline.

Note that because the scale factor for each polyline may be different, this method is considered a local

interpolation.

Adopted from a 2012 GIS-T presentation by Paul Scarponcini, “NCHRP 20-27 to ISO 19148: 18 Years of Progress

in Linear Referencing.”

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 48 DOT Contract #GS-35F-0001P

September 2014

 LRS accuracy/precision

 Network/linear topology

Figure A.7 details the baseline network requirements used to assess level of maturity. These baseline

requirements were derived from the FHWA network specification guidance, with some additions and

modifications based on the current recommended best practices. For example, the project team is

recommending that the baseline scale requirement be changed from 1:24,000 to 1:5,000 which is in line

with the linear precision and accuracy recommendation of 0.001 miles. Moreover, the degree of

cartographic generalization that occurs at 1:24,000-scale mapping makes dual-carriageway

discrimination problematic and obscures other roadside features.

Figure A.7: Common Baseline Network Requirements34F

59BMATURITY ASSESSMENTS

Common baseline network requirements can be characterized in one of three ways:

 Needs investment to meet common baseline network requirements – Local road network is

incomplete and will require additional effort before submission to the State DOT

 Satisfies common baseline network requirements – Local road network substantially meets the

minimal requirements established for the ARNOLD/HPMS national network

Applied Geographics, Inc., 2014. Regarding Addresses, address points are preferred when available.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 49 DOT Contract #GS-35F-0001P

September 2014

 Exceeds common baseline network requirements – Local road network meets all of the

minimal requirements for the ARNOLD/HPMS national network plus contains value-added

features supporting State-level business processes (e.g., address ranges)

Using the common baseline network requirements, the table below summarizes how different road data

components can be classified by their maturity and ability to meet these requirements.

Needs Investment to meet

common baseline network

requirements

Satisfies common baseline

network requirements

Exceeds common baseline

network requirements

Road Centerline Geometry

Data Inclusion

Some or all local roads, including

alleys, NOT included

Some or all private roads NOT

included

Some or all ramps, roundabouts,

frontage roads or other highway

geometry NOT included

All public and private

highways, roads and streets,

including ramps and

frontage roads

All public and private

highways, roads and streets,

including ramps and

frontage roads

+ Includes temporary,

emergency, construction

and/or evacuation roads

+ Contains historical and

future alignments

Dual

Carriageway

Single-carriageway representation

Dual-carriageway

representation where

positive barrier median or

median width > 4’

Dual-carriageway

representation where

positive barrier median or

median width > 4’

Scale

Scale smaller than 1:5,000

1:5,000 scale

Scale larger than 1:5,000

Update Cycle

> 1 year

Updated/certified annually

Updated more frequently

than annually

Coordinate

System

WGS 84

Road Attributes

Basic

One or more basic road attributes

missing

All basic road attribute

values exist:

• Persistent road ID

number

• Road/street name

• Functional class

• Year

• State

All basic road attribute

values exist

+ HPMS Section data

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 50 DOT Contract #GS-35F-0001P

September 2014

Needs Investment to meet

common baseline network

requirements

Satisfies common baseline

network requirements

Exceeds common baseline

network requirements

Advanced

No baseline network requirements for advanced road attributes have been established

Examples Include:

• Active vs. Planned; Public or Private; Improved vs. Unimproved

• X-section, pavement surface, signing & marking, traffic characteristics

• Certified mileage

• Political, administrative or census geographies

Street addresses35F

Right/Left

Address

ranges

No right side/left side address

ranges

Right side/left side address

ranges

Right side/left side address

ranges

E911

No rural addresses for E911

Urban and rural addresses

used for E911

Urban and rural addresses

used for E911

Site Address

N/A

Address points (i.e., actual

property locations)

Linear reference control

Linear

precision

> 0.001 mile (e.g.,, 0.01 mile)

0.001 mile

< 0.001 mile

End-end

centerline

mileage

accuracy

unknown or > 0.001 mile

0.001 mile

< 0.001 mile

Network/Linear Topology

Topology

Simple segments with no

topology

Common topology for road

network models (e.g., spatial

analysis, buffering)

Local road networks with

enhanced network/linear

topology for modeling

features such as grade

separations (i.e.,

over/underpasses), tunnels,

ferry routes, one-way traffic

flow, turn restrictions.

Most often, addresses are expressed as an "address range" (lowest address number to highest address number)

for each roadway segment. This allows any distinct address to be interpolated to a specific location along the

street. However, due to the inherent uncertainty of interpolation and the fact that addresses are often not

evenly distributed along a roadway -- particularly for longer, rural roads -- the interpolated location may not

accurately match the true location. This can adversely impact the routing of emergency vehicles to the correct

address location during dispatching. Due to these challenges, it is increasingly common for addresses to be

mapped as discrete address points for each location. When an address point dataset is used as a reference, the

addresses of facilities such as hospitals, daycare centers, police and fire stations, etc. can be matched precisely

to these discrete, accurate point locations, thereby facilitating routing and improving response time.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 51 DOT Contract #GS-35F-0001P

September 2014

A P P E N D I X B 26BR O A D W A Y G E O M E T R Y

B.1 34BR O A D W A Y S EGM E N T A T I O N

Roadway segmentation (i.e., depicting the physical start, end, and length of each roadway segment) can

be accommodated by one or more methods: 1) intersection-based (segmented); or 2) route-based.

Ideally, the chosen solution would combine the two methods to meet multiple business needs.

It is important to select and maintain a consistent approach that aligns with the internal business

systems of the DOT, as well as with those of external systems that rely on the transportation data.

However, it also must be acknowledged that there are many different uses of centerline data, with

various business drivers for both segmented and route-based networks (e.g., emergency response and

roadway inventory, respectively). Thus, multiple sets of centerline geometry (often at different map

scales) are frequently created and maintained to meet various business needs defined by State DOTs. In

order to achieve optimal benefits from the LRS, a shared enterprise approach should be considered such

that the same solution can be used to support several business needs. For example, the production data

maintenance network might be intersection based,

where nodes are located at at-grade intersections and

ramps (to support segment-level attributes such as

address ranges, and to support routing and turn

restrictions); and then a derivative product or export

from this could be a route-based network to support

DOT needs.

60BINTERSECTION-BASED SEGMENTATION

Intersection-based or variable-length segmentation approaches are defined by segment endpoints

occurring at geometric or physical intersections (i.e. “transportation opportunities”) of two or more road

centerlines. Routes are defined by associating segments in a logical from/to node order (see Figure B.1).

The main advantage of using a segmented network is that updates to segment geometry are local and

limited in their cascading effects. Additionally, intersection-based networks can support details such as

one-way streets and turn restrictions at intersections- both of

which are critical in terms of vehicle routing and emergency

response. A disadvantage of this structure is that segmented

networks can become very large and challenging to maintain

at the statewide level, especially if one or more LRMs are

defined for the network.

Implement an enterprise approach

allowing multiple business needs to be

met. For example, maintain the

intersection-based network, and regularly

generate the route-based network from it.

Intersection-based networks should

be used when details such as one-

way streets and turn restrictions at

intersections are needed.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 52 DOT Contract #GS-35F-0001P

September 2014

Figure B.1: Segment-Based Network36F

61BROUTE-BASED SEGMENTATION

A common alternative to variable-length segments is to create long polylines representing entire route

lengths and indexing various point locations along this line using one or more linear reference methods

(see Figure B.1). This approach is adopted from the Straight Line Diagram37F

(SLD) highway inventory

methods used by many DOTs. The advantage of using a route-based network is that it usually aligns with

legacy statewide systems and requires far fewer roadway segments in the GIS system to manage.

Disadvantages include increased complexity with routing, address ranging, and requiring a well-defined

LRM to assign roadway attributes. This approach also requires the underlying database or application

software to support some kind of “dynamic segmentation” functionality.

Figure B.2: Route-Based Network38F

Michael Baker Jr., Inc., 2014

Straight Line Diagrams (SLD) are linear depictions of roads and intersecting features (e.g., intersections, fixtures,

structures) where the line does not include alignment geometry (i.e., curves). The distances between features may

or may not be to scale. The route-based segments discussed in the text generally include scaled alignment

geometry.

Michael Baker Jr., Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 53 DOT Contract #GS-35F-0001P

September 2014

As defined, and to meet ARNOLD

requirements, utilize a dual-

carriageway representation for

divided roadways, ideally with

independent mileage calibration.

62BA COMBINED ENTERPRISE APPROACH

Recognizing that each alternative has strengths and weaknesses relative to certain business needs,

many States have implemented a solution that includes both types. For production data maintenance

needs, a segmented centerline model is often used for flexibility, incorporating roadway aliasing,

shielding, and multiple Linear Route Measures. Understanding that this large and complex model may

not work for many supported agencies and systems, data collaborators can then develop workflows that

export the GIS network into different formats (e.g., route-based) using Extract, Transform, and Load

(ETL) processes. Using this dual approach for flexible reporting, the needs of various end-users can be

met. For example, if the DOT gets a route-based network, and Emergency Response gets a segment-

based network for NG911, and the State GIS office gets a network to support the need for

cartographically accurate labels – all derived from the same fundamental road geometry – then all use

case business needs can be met.39F

B.2 35BD U A L C A R R I A G E W A Y S

Dual Carriageways for a roadway typically involve a physically divided roadway that necessitates two or

more lines to adequately model the road when it has

become too complex to be represented by a single line. The

multi-line representation can also help meet more rigorous

business needs and application requirements for real-world

fidelity. Increasingly, States are collecting road

characteristics on both sides of a divided highway, which is

another consideration. A multi-centerline representation of

the roadway provides high accuracy in representing the actual roadway elements, but requires

additional complexity in processing and managing the roadway data. Figure B.3 depicts the necessity for

dual carriageway to accurately reflect the road geometry.

For an example on how this has been implemented in New Jersey, please see the New Jersey Geographic

Information Network https://njgin.state.nj.us/NJ_NJGINExplorer/jviewer.jsp?pg=ROADS

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 54 DOT Contract #GS-35F-0001P

September 2014

Figure B.3: Single-Centerline vs. Dual-Carriageway representation40F

63BDUAL CARRIAGEWAY MILEAGE CALIBRATION

Keeping in mind that roads represented by dual carriageways are still technically the same road, defining

LRMs on dual-carriageway centerlines can be done in different ways. The simplest option is to assign the

same linear measure to both roadways using the “primary” measure on both roadways. When the

lengths of the two roadways are equal (or nearly so), this simple approach may be adequate. However,

in many cases, the differences between the primary and

secondary roadway’s geometry result in significant differences

in the lengths of the segments. In this case, the preferred

method is to maintain a separate and independent linear

reference for each segment (see Figure B.4). This allows for the

use of true measured mileage when performing roadway

Michael Baker Jr., Inc., 2014

For dual carriageways, maintain a

separate and independent mileage

for each segment, while providing a

mileage conversion factor.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 55 DOT Contract #GS-35F-0001P

September 2014

inventory, which is preferred. Be aware that a potential disadvantage of this approach occurs when

using small display or print scales. Distinguishing between the primary and secondary roadways can be

visually confusing.

Figure B.4: Secondary Independent Mileage41F

There are different options for handling measures on the non-cardinal (“secondary”) side of the

roadway (i.e., the blue line in Figure B.4). As shown, measures can be assigned in the same direction as

the cardinal side, using a similar (but separate) measure structure. Under this scenario, mile point values

increase in the direction of travel on primary roadways and decrease in the direction of travel on the

secondary side. This approach is used by all Interstate mile posts, and allows for tracking true mileage

on both sides of a dual carriageway, but with similar measures on both sides. Alternatively, measures

can be assigned in the direction of travel on the non-cardinal side, such that they would run opposite to

those of the cardinal side. This approach can help to simplify field inventory (e.g., using a Distance

Measuring Instrument, or DMI).

Ultimately, because most dual-roadway routes are on the Federal-Aid System, they will already be

included in a State's LRS, and the best approach would be for States to extend whatever business rule

that is in place for handling the non-cardinal measures on the Federal-Aid routes to any local dual-

carriageway roadways, in order to have a consistent, enterprise approach across the State. As described

below, mileage transformations can be used to convert secondary roadway "as-driven" mileages to

official mile points in either real-time or post-processing transactions.

Michael Baker Jr., Inc., 2014.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 56 DOT Contract #GS-35F-0001P

September 2014

In many cases, roadway attributes are referenced to the primary (or cardinal) direction of travel for a

roadway. As such, it is advisable to create and maintain a mileage conversion process, for translation to

and from the LRMs. For this situation, translation from the primary (or cardinal) direction mileage to the

secondary (or non-cardinal) direction mileage can be performed as needed. See Figure B.5 for details on

the provisions required for the LRM conversion.

Figure B.5: LRM Conversion Diagram42F

64BDETERMINING WHEN TO REPRESENT A ROADWAY AS DUAL CARRIAGEWAY

When digitizing a roadway into appropriate carriageways

(divided, undivided, express/local lanes, etc.) the

challenge is determining when it is most appropriate to

model a roadway as divided or undivided. Within the

HPMS guidelines, FHWA defines a divided roadway as “A

multi-lane facility with a curbed or positive barrier median or a median that is 1.2 meters (4 feet) or

Michael Baker Jr., Inc., 2014

FHWA defines a divided facility as having

a median with a positive barrier or a

width of 4 feet or greater.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 57 DOT Contract #GS-35F-0001P

September 2014

wider.” However, the State DOT may have other business needs (e.g., safety, NG 911, bridge inventory)

for representing facilities as dual carriageway differently than the HPMS guidance; the DOT should take

into account the predominant function of the roadway.

Situations with complex roadway geometry

or areas that may require a more simplified

model may warrant alternative models

beyond the basic HPMS guidance. For

example, a roadway island barrier to

delineate left turn lanes for an intersection

may not truly require the added complexity

of a dual carriageway (see Figure B.6 43F

Conversely, a roadway model strictly

following HPMS guidance should include

separate carriageways for roadways with

center striping. However, doing so would imply

a discontinuity between each travel-way, and would not accurately depict the ability to cross over to the

opposite direction of travel, particularly for emergency vehicle purposes). Ultimately, defining the

overall goal of the roadway data model and predominant function of the section of roadway can help

determine the most appropriate dual carriageway solution. If the goal is to create a topologically

detailed and accurate representation of the roadway, regardless of the additional complexities this

would introduce, then modeling divided roadways that strictly adhere to the guidelines may make the

most sense. In many cases, though, it is recommended

that States define a balance between an overly complex

model and an overly simplified model. For example,

implement a guideline that defines a divided roadway as a

continuous 500-foot divided physical barrier or median

(greater than 4 feet). This guidance also excludes a physical barrier that exists only for turning

channelization. This definition provides a good balance between overly simplified and overly complex

models.

Image from http://nacto.org/usdg/neighborhood-main-street/

Model a roadway as dual carriageway

when there is a divided physical

barrier or median greater than 4 feet

that continues for at least 500 feet.

Figure B.6: Roadway Island Barrier

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 58 DOT Contract #GS-35F-0001P

September 2014

Model each traffic circle on a

case-by-case basis, with the goal

of minimizing segment overlap

and route segmentation.

65BSPECIALIZED LANES

Specialized travel lanes, such as reversible

Lanes, HOV lanes, express lanes, through

lanes, etc., can also be a complicating factor

when it comes to dual-carriageway

representation. These types of lanes are

best handled tabularly, as LRS events.

Figure B.7: HOV Lane44F

B.3 36BT R A F F I C C I R C L E S

Traffic circles (also known as rotaries and roundabouts)45F

have proven to be a GIS challenge for

transportation agencies. Traffic circles are the intersection of two or more roadways in an uncontrolled

at-grade interchange, intended to keep traffic moving through the intersection. Defining this in GIS

terms and meeting the requirements of the LRS becomes much more challenging, and smaller traffic

circles might best be modeled differently than larger, more complex traffic circles. For small traffic

circles, even if the map geometry is generalized, the measured travel distance can be as driven. For large

traffic circles, involving higher-order roadways, the fundamental modeling debate becomes centered on

how to minimize roadway segmentation, while also minimizing segment overlap.

Due to the complex nature of how traffic circles interweave

different roadways, there are alternative ways of handling

various situations, rather than a single “one size fits all”

solution. Depending on the individual needs of the system,

options for modeling traffic circles include case-by-case

segment merging or more simplified configurations.

Image from http://transportationblog.dallasnews.com/2012/09/as-officials-mull-hov-lane-changes-north-texas-

toll-filled-future-comes-into-focus.html/

The difference between roundabouts and traffic circles is recognized in terms of traffic flow and safety concerns.

However, from an LRS modeling perspective, they are handled in a similar fashion.

Store specialized lanes

(HOV lanes, reversible lanes,

express lanes) as LRS events.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 59 DOT Contract #GS-35F-0001P

September 2014

CASE-BY-CASE SEGMENT MERGING

Since many traffic circles are not perfectly symmetrical in design, each circle can be modeled on a case-

by-case basis, with the goals of minimizing segment overlap and route segmentation , minimizing breaks

for the highest order route entering/exiting the circle, and defining ramps where needed. Figure B.8

shows an example of this type of complex ramp configuration.

Figure B.8: Complex Traffic Circle Ramp Configuration46F

Additionally, larger circles can be configured as concurrent routes, where the overlapping roadways

share centerline geometry. In these cases, the lower-order route would show a gap in mileage where

coincident with the higher-order roadway. Figure B.9 provides an example of this alignment:

Michael Baker Jr., Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 60 DOT Contract #GS-35F-0001P

September 2014

Figure B.9: Coincident Roadway Circle Configuration47F

66BSIMPLIFIED TRAFFIC CIRCLE CONFIGURATION

Some States have chosen non-traditional methods for representing smaller traffic circles. For example,

New Hampshire treats the circle as its own route, separate from the other participating routes. The

other routes show a segment break, but do not have a mileage gap in their LRM. See Figure B.10 for an

example of this method.

Michael Baker Jr., Inc., 2014. In this diagram, the small connecting pieces of Route 623 on either side of the

circle would best be modeled as connector ramps between the two divided I-95 segments. This approach would

minimize segmentation on the broken route (in this case, 623).

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 61 DOT Contract #GS-35F-0001P

September 2014

Define the start and end of the ramp

as the taper from and to the mainline.

Define deceleration and acceleration

sections as LRS events.

Figure B.10: Simple Traffic Circle Configuration48F

B.4 37BR A M P S

Ramps can be defined as connecting roadways that permit traffic to flow from one mainline highway to

another without crossing any other traffic stream49F

. The interchanges that use ramps may be grade-

separated (i.e., elevated) or at-grade (i.e., same elevation)

50F

. At-grade intersections can have a physical

barrier to separate the turn lane from the mainline. Depending on the ramp length, these can be

considered ramps. Ramps are generally not assigned mainline road names or address ranges, and they

typically have a single direction of travel. Defining ramp

start and end locations poses a special challenge to

transportation GIS networks. When building a GIS network,

ramps typically start and end at the point that the ramp

tapers from (or to) the mainline route (see Figure B.11).

Michael Baker Jr., Inc., 2014

Features such as “jug handles” are also handled as ramps within the network.

HPMS only requires that the ramp participates in a grade-separated interchange. However, this discussion of

ramps covers general guidance and best practices for both types of ramps, regardless of HPMS requirements.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 62 DOT Contract #GS-35F-0001P

September 2014

Figure B.11: Ramp Alignment Diagram51F

To further clarify the physical sections of a ramp (e.g., if needed for inventory and field work), the

deceleration and acceleration lanes should be defined through LRS events on the roadway. The

deceleration area of the ramp is between the mainline taper and the start of a physical barrier when

entering the ramp. The acceleration area of the ramp is between the end of the physical barrier and the

mainline taper when exiting the ramp. See Figure B.12 for details. The ramp mileage would start and

end at the mainline taper locations.

Figure B.12: LRS Events on the Ramp Segment52F

Michael Baker Jr., Inc., 2014

MAINLINE

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 63 DOT Contract #GS-35F-0001P

September 2014

Minimum ramp length

recommendation: 25 feet

These roadway elements often have the

same start/end point, which can be

problematic for LRS. The DOT will need to

establish standards for handling them

consistently in the statewide network, taking

into account any software limitations.

Additionally, it is recommended that DOTs determine a minimum ramp length for modeling ramps

within their roadway network. Without a defined standard on when to create a ramp features (and

when not to), a statewide network can become overly complex, with

many small and relatively insignificant segments. It is recommended

that small connecting ramps of less than 25 feet (excluding

deceleration and acceleration sections) do not need to be modeled in the network. These situations can

instead be modeled as simple intersections.

B.5 38BC UL- DE- S A C S A N D L O O P S

As State DOTs begin to incorporate local road data into their network, they will likely encounter certain

roadway features that do not occur at the State level. Some prime examples of these are Cul-de-sacs

and loops. Cul-de-sacs and loops are generally

described as highway segments where one

continuous arc starts and ends at the same

coordinate (see Figure B.13 for examples). In

terms of LRS, having the same start and end point

can create issues (e.g., with topology, measures,

and mileage calibration) in certain vendor

products. As these features are incorporated into

a statewide all-roads network, the DOT will need to establish standards of handling them to ensure

consistency and data quality. A key aspect of this issue depends on how COTS software products handle

these types of situations, and DOTs should understand the limitations of any particular vendor software

as they develop standards.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 64 DOT Contract #GS-35F-0001P

September 2014

Figure B.13: Cul-de-sacs and Loops53F

For certain looped roads (like the one shown in Figure B.13 on the right), the start/end point issue may

be unavoidable. However, cul-de-sacs can be modeled in a number of ways, depending on their nature

and the specific needs of the agency. A cul-de-sac without an island can terminate straight through to

the edge of the pavement, or terminate in the middle of the pavement. Cul-de-sacs with a physical

island may include a “lollipop” end (as is shown in Figure B.13 on the left). While “lollipop” ends may be

the cartographic preference, they are not recommended for use in LRS because self-connecting

segments can cause complications with topology checks, mileage calibration, and physical inventory.

Some DOTs have chosen to draw the “lollipop” for aesthetic purposes, but not connect it to the segment

(i.e., stop the loop 1 foot before it self-intersects). However, this approach can cause issues with

connectivity and network topology, which can inhibit routing and network analysis (see Appendix

Section E.6 for further detail). An alternative to the “lollipop” end is to create a separate loop segment

around the physical island, which will accommodate applications or uses that can’t handle the self-

intersecting segments. If address ranges are being stored, the circle can be made up of two segments,

one for each side (as seen in the lower diagram in Figure B.14). Ultimately, the agency will need to

determine the appropriate representation to match its specific needs, taking into account any software

or application limitations.

The State of Tennessee’s TIPS GIS Modeling Specifications document54F

contains some additional useful

guidance about modeling cul-de-sacs. Figure B.14 summarizes its recommendations, and distinguishes

between cul-de-sacs with and without a physical island.

VTrans Mapping Unit, 2014

TIPS GIS Modeling Specifications, State of Tennessee, Next Generation 9-1-1

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 65 DOT Contract #GS-35F-0001P

September 2014

Figure B.14: Cul-de-sac guidance from the State of Tennessee55F

TIPS GIS Modeling Specifications, State of Tennessee, Next Generation 9-1-1

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 66 DOT Contract #GS-35F-0001P

September 2014

A P P E N D I X C 27BR O A D W A Y A T T R I B U T E S

C.1 39BR O U T E E V E N T S V S . S E G M E N T E D A T T R I B U T E S

When developing a roadway GIS network, creating the network data schema (i.e., how attributes are

stored) can be a complex and involved task. One of the most important decisions that must be made

early in the design is whether roadway attributes will be related to the segments as route events, or

stored within the roadway segments. Route events are related to the segments through Linear

Referencing, allowing the flexibility to change extents without changing route segmentation. Segment

attributes are related directly to the GIS centerline segment database record. Figure C.1 depicts the

distinction between roadway events and segment attributes.

Figure C.1: Storing attributes as Roadway Events vs. Segments56F

Below is a brief description of each option, and some basic considerations.

Route Events are stored as separate roadway data, referenced to the road segments through Linear

Referencing. Any changes that affect the segment’s LRM would also need to be evaluated and changed

in the route event tables. Route events require additional considerations and necessitate well-

established, defined, and standardized Linear Reference Systems (LRS).

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 67 DOT Contract #GS-35F-0001P

September 2014

Local roadway attributes are often stored

within the segments; DOTs should keep

this in mind when integrating local data

into an ARNOLD dataset.

Store a minimum set of “base”

attributes on the segment, and

everything else as route events

within the LRS.

In Segment Attribution, roadway attributes are stored at the segment level of the database model. This

means that changes to these attributes would require additional roadway segmentation, which would in

turn increase the centerline data maintenance. As more attributes are included in the roadway network,

more line-work segmentation and maintenance would be required.

When deciding how to store attributes within a roadway

network, and how many to store, the DOT should choose a

minimum set of “base” attributes to store within the segment,

and then store everything else as route events in the LRS. At a

minimum, the base attributes should include the route ID,

beginning and end mileage, and data source for the segment.

The DOT may also choose to store other key attributes in the segments, such as route number, road

name, facility type, functional classification, ownership, etc. These “base” attributes should be

determined in collaboration with key stakeholders (both inside and outside the DOT, such as emergency

response), and take into account the ARNOLD schema (see next section). However, the goal should be to

minimize the overall number of “base” segment attributes and store the majority of the roadway

information (number of lanes, pavement condition, speed limit, etc.) as events. Automated routines can

then be created to export and publish centerline datasets that meet the needs of non-LRS users, such as

a GIS layer with speed limit information stored within the segments. Storing fewer attributes also results

in less costly data collection and maintenance.

Additionally, as DOTs are collecting much of the ARNOLD data from local agencies, the format of the

local data will need to be taken into consideration. Local transportation centerline data is typically

segment-based and stores attributes directly within the segments. Local streets and roads are, in many

cases, non-continuous and non-contiguous, which makes route-based schemes quite complicated.

Moreover, routes are often concurrent or coincident, adding to the complexity of route-based

approaches. Thus, it is anticipated that most local

attributes will be stored within segments, and the

DOT will need to accommodate this when creating

the ARNOLD dataset. Either way, attributes should be

evaluated to determine whether they are more

suitable for route-event storage or segment storage.

C.2 40BA R N O L D S C H E M A

As discussed in Section C.1, each State DOT likely stores similar information in its roadway dataset,

although in different attribute schemas (i.e., as part of the base, or as route events). This report is not

aimed at standardizing attributes across DOTs. Instead, the goal is to communicate the key, common

attributes needed for the ARNOLD requirements. As such, the following fields are critical for ARNOLD

and will need to be maintained and generated in some manner by the DOT:

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 68 DOT Contract #GS-35F-0001P

September 2014

State DOTs should maintain or be able

to generate the key ARNOLD fields to

meet submission requirements.

Define a standardized route

identification convention as the

framework for aligning all DOT and

local agency roadway asset data.

 Route_ID – unique road ID number

 Road Name

 Functional Classification

 Ownership

 Facility Type

 State Code

 Year_Record

 Source – the entity providing the data

 Geometry- Well Known Binary (WKB) using (x,y,m) geometry. This should adhere to the OGC

specification for Simple Features57F

. The measures will be stored in Miles (with a recommended

minimum precision is to the nearest thousandth of a mile.)

C.3 41BR O U T E ID N U M B E R I N G

One of the most vital components of an enterprise GIS Roadway Network is a standardized and common

route numbering schema. DOTs should create a common route identification method as the framework

with which all units within a State transportation department (safety, traffic, pavement, etc.) and local

agencies (counties, municipalities, etc.) can align their respective roadway asset data.

Ideally, DOTs will implement a standardized route ID

convention58F

that can be utilized by all jurisdictions

(counties, municipalities, etc.) in the State. However, the

DOT will need to understand that local government

agencies (LGAs) are unlikely to adopt State-mandated

standards unless they have had meaningful input. The

benefit for LGAs that choose to adopt such standards is reduced duplicative overlap with other local

roadway names and numbers. Alternatively, the State DOT may want to create unique system IDs to

share with LGAs.

FHWA requires that HPMS submissions meet the OGC standard for simple features. See Open Geospatial

Consortium Standards, Simple Feature Access

DOTs may also want to consider implementing a multi-modal route ID convention that includes other

transportation methods (such as Rail and Waterways). Multi-modal data is an important consideration for the

future but outside the scope of this Study.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 69 DOT Contract #GS-35F-0001P

September 2014

All roadways should include at least one

standardized name. Roadway naming

should also include roadway aliases,

historical names, honorary names, etc.

Modern best practices in database

management are to use IDs such as

GUIDs (Globally Unique Identifiers).

Traditionally, a route ID convention might include information

about the route type, county, municipality, route number, and

directionality. However, modern best practices in database

management are to use identifiers such as GUIDs (Globally

Unique Identifiers) rather than IDs comprising descriptive information. Regardless of the ID system

chosen, field values should be kept to a reasonable length (e.g., less than 30 characters).

C.4 42BR O A D N A M I N G

The roadway name is a core attribute for roadway systems. Excluding ramps, all roadways should

include at least one standardized name. Roadway

naming should also include roadway aliases,

historical names, and honorary names (if applicable),

ideally following a national standard (such as the

FGDC standard59F

). This level of detail is required for

proper geo-locating for addressing and emergency

services. In addition, route shielding (e.g., I-95) and coincidence (e.g., Route 1 & 9) should also be

accommodated for higher functional class roadways. In many cases, roadway names will not align with

route designations (or with each other). These situations typically arise from legacy route definitions

associated with older State and county route alignments. As shown in Figure C.2, County Route 612

covers multiple road names. Since road names may differ from the route designation, they should not

be part of the route definition as a rule, although they may be used as a surrogate when there is no

other route designation.

The Federal Geographic Data Committee (FGDC) standard is one example of an address standard. (See FGDC

project standards.) The National Emergency Number Association (NENA) also has a GIS Data Collection and

Maintenance standard with useful information relevant to emergency response data, including addresses and road

names. (See NENA GIS Collection & Maintenance.)

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 70 DOT Contract #GS-35F-0001P

September 2014

Figure C.2: Route Name Changes Diagram60F

For segmented roadway networks that are broken at physical or geometric intersections, road names

can be stored as attributes of the segments (although a separate database table with a one-to-many

relationship could be utilized to accommodate multiple names). For route-based networks, names

should be stored as a linear-referenced event.

The LRS database developer needs to be aware of several situations regarding multiple route names

associated with the same route segment.

1. Route aliases occur when two or more names are used for the same route segment in the same

route system). Route aliases are the simplest case of multiple names. In many cases, aliases are

not posted but are still recognized and used by travelers, local governments, or the U.S. Postal

Service.

2. Route concurrencies occur when two or more routes in the same route system share the same

road segment, as seen in Figure C.3.

Michael Baker Jr., Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 71 DOT Contract #GS-35F-0001P

September 2014

Figure C.3: Route Concurrency61F

3. Route coincidence occurs when two or more routes from different route systems share the

same route segment:

Figure C.4: Route Coincidence62F

Photo from AAROADS Interstate Guide website

Photo from The New I-26 Virtual Tour Tennessee and North Carolina

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 72 DOT Contract #GS-35F-0001P

September 2014

The GIS network should have the

capability to support multiple

LRMs, while standardizing on a

single LRM as the preferred

measure (such as driven mileage).

C.5 43BM U L T I P L E L I N E A R R O U T E M E A S U R E S

As isolated data systems begin to adopt a standardized roadway network, they do not all follow the

same standards for route measures (see Figure C.5). For example, it is common for HPMS supporting

systems to use driven mileage measures63F

through a physical Distance Measuring Instrument (DMI),

while accident records and State police use mile-markers or intersection offsets.

Figure C.5: Multiple Linear Route Measures 64F

Forcing a standard convention for all dependent systems and

business functions can be counterproductive in fostering

widespread adoption. Alternatively, multi-linear route

measures should be allowable and be utilized to solve the

needs of multiple dependent systems, while encouraging a

single LRM as the preferred, standardized measure. To

support field data collection efforts, it is recommended that

driven mileage be used as the standardized measure.

Supporting multi-linear route measures requires calibration points or segments for each linear route

measure, as well as functionality to translate between reference systems (e.g., conversion of a driven

mile reference to a mile marker reference). Several types of transportation GIS software will

automatically support the management and dissemination of data on multi-linear route measures

(Intergraph GeoMedia, Esri Roads & Highways, etc.).

In modern LRS, an as-driven system is a necessity.

Michael Baker Jr., Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 73 DOT Contract #GS-35F-0001P

September 2014

Figure C.6: LRM Transformation65F

As illustrated in Figure C.6, the location of an event can be described using multiple LRM types with no

loss of positional accuracy. A common example occurs when transferring roadway data from

construction files indexed by stationing (e.g., format STA XX+yy.yyy’) to a GIS-based asset inventory

database indexed by a statewide LRM such as Route Milepoint. The common linear element is the new

alignment, and the common reference object is usually the project terminus, generally located at a road

intersection, administrative boundary, or other geographic feature. The transformation aligns the

00+00.00 station with the corresponding route milepoint value and uses simple arithmetic to calculate

the station-to-milepoint conversion of each roadway object of interest.

C.6 44BP U B L I C V S . P R I V A T E R O A D W A Y S

Statewide roadway networks do not typically include private roadways; thus, State DOTs have not had

to manage or track private roadway data. But as the GIS network expands to include all local roads,

State DOTs will need to begin to manage the distinction66F

between public and private roads and store

this information in the roadway network. Although private roads are not required in the HPMS

submission67F

, Emergency Response is a key driver for including private roads in the network. Therefore,

Dave Fletcher, 2014

This distinction expands beyond simply public vs. private road, and can also include other non-public roads such

as military roads, Bureau of Land Management (BLM) roads, Forest Service roads, etc.

HPMS requires that the road geometry submitted correspond to the State-certified mileage, which is signed by

the Governor. In other words, if a State decides to include private roads in its certified mileage number, then the

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 74 DOT Contract #GS-35F-0001P

September 2014

Although HPMS only requires the

roads that correspond to certified

road mileage, State DOTs may include

private roads in their network to

support emergency response and

safety considerations.

For maximum data evaluation

capability, capture and

manage both the construction

date and inventory dates.

private roads should be properly attributed so that they can be removed from the HPMS submission, if

applicable. That said, if a State wants to submit private roads in its HPMS submission, they may be

accepted by FHWA.

The distinction between public and private roads is

typically defined at the local level, based on who is

responsible for maintenance and record keeping. For

example, a town/city/county-maintained road would

be “public,” while a road maintained by a

homeowners association within a gated community

would be “private.” Making this distinction between

public and private can be very difficult, particularly if

the only source is orthophotography. It is critical that this information come from an authoritative

source (e.g., the local jurisdiction).

C.7 45BI N S T A L L D A T E A N D I N S P E C T I O N D A T E

Implementing the ability to analyze transportation data over time (i.e., temporality) starts with the

inclusion of effective dates. However, an effective date can be more complex than a simple, single date.

The most obvious date to capture would be the installation or

creation date for the roadway. In reality, this information may not

be available to the group maintaining the GIS datasets. As such, it

may be more realistic to capture the inspection or field inventory

date of the roadway.

 Installation date (or “Open for Traffic” date) is the date

that construction on the new or realigned roadway section

was completed.

 Inspection or Inventory date is the date that the roadway

was inventoried or inspected to collect the newly

constructed alignment and attribution.

roads should be included in the HPMS geometry. If a State wants to include a private road in its submittal, the

facility type must be > 4.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 75 DOT Contract #GS-35F-0001P

September 2014

C.8 46BA D D R E S S I N G

While addressing is not typically a mainstream DOT business function and is not required for HPMS

reporting, it is very practical and rational to include addresses with roads, when feasible. There are two

main methods of storing address information:

 Address Range: Address locations are estimated by proportional relative distances determined

by the lowest address on the block (i.e., the “from address” attribute on the segment) to the

highest address on the block (i.e., the “to address” attribute on the segment). For example, if

the address range on one block of Main Street runs from 100 to 200, then the address of 150

Main Street would be located 50 percent of the distance along the street segment.

Figure C.7: Address Range LRM68F

 Address Points: In this method, address information is stored as distinct locations, represented

by a point feature stored at or near the physical address (building centroid, driveway entrance,

property centroid, etc.), or as an LRS event.

Figure C.8: Address Points LRM69F

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 76 DOT Contract #GS-35F-0001P

September 2014

For emergency response purposes,

discrete address point locations linked

to the LRS are preferable to give first

responders an exact location.

Coordination between DOT and local

E911 agencies is critical in ensuring

that emergency responders have

access to accurate data.

Either method is workable, but the following issues should be considered:

 For emergency response purposes, address point locations are preferable to give first

responders an exact location70F

, rather than an

approximated location based on the distance

along the segment. Particularly in rural areas,

approximated location based on address

ranges can be very far off from the exact

address due to long stretches of roadway and

irregular addresses. Whether using address ranges or points, field verification of house numbers

is strongly recommended for address-matching and geocoding applications.

 When address points are used, the location of the placement of these points may vary,

depending on local needs and standards. This can cause data conflation issues, depending on

where the point is placed (e.g., at the centroid of the parcel, centroid of the structure, front

door of the structure, the end of the driveway  all examples used by different localities for

placing address points).

 If using an intersection-based network, implementing address ranges is possible as the segments

are broken at actual intersections.

 If using a route-based network, it is recommended that address point locations be used;

implementing accurate address ranges along the entire length of a route segment would be very

difficult without segmentation to accommodate the assignment of address ranges.

Although address management is typically not a core DOT

function, it is needed for tracking crash and safety

information, as well for as several uses outside of the DOT

(e.g., emergency response). Address ownership and

origination are often outside of the DOT and usually at the

local (e.g., city or county) E911 agency. Thus, it is important

for the DOT and local E911 agencies to coordinate to ensure the most accurate data is available to

emergency first responders.

The National Emergency Number Association (NENA) has a GIS Data Collection and Maintenance standard with

useful information relevant to emergency response data, including addresses and road names. (NENA GIS Data

Collection)

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 77 DOT Contract #GS-35F-0001P

September 2014

All published or distributed

datasets should include

standardized metadata, ideally at

both the layer level and the object

level, but at least at the dataset

level.

Metadata should be embedded within

the GIS layer whenever possible. At a

minimum, complete the ISO metadata

sections containing Identification and

Metadata Reference.

A P P E N D I X D 28BM A I N T E N A N C E

D.1 47BM E T A D A T A S T A N D A R D S F O R G I S A N D R O A D W A Y A S S E T D A T A

A statewide GIS roadway network and associated data are typically distributed to stakeholders, both

internally to State agencies and externally to the public; therefore, having consistent and well-defined

metadata (i.e., information that describes the dataset and

dataset attributes) for all production data layers is critical

to data usability and widespread adoption. While metadata

creation is emphasized for GIS data layers, it is important to

include some form of metadata with any published

datasets, whether or not they are geospatial. End users

need the ability to reference important data facts, such as

data accuracy, collection methodologies, and data owners.

Data aggregators also need to know about such information.

Ideally, metadata is maintained at both the layer level and at the object level (i.e., for each geometric

object within the dataset). At a minimum, metadata is needed at the dataset level, which might contain

multiple data layers or themes. Whenever possible, the metadata should adhere to existing standards,

such as the Federal Geographic Data Committee (FGDC) Content Standard for Digital Geospatial

Metadata, or the Dublin Core Metadata Element Set71F

. States may want to work with their GIS data

clearinghouse on metadata standards as well. The metadata should be built into the GIS data layer,

without the need for additional data files. Additionally, the ability to export metadata information into a

report format is useful in communicating this information when sharing data, and for meeting HPMS

metadata reporting requirements.

It is recommended that all published data follow

International Organization for Standardization (ISO)

metadata standards, which should be embedded within

datasets using standard GIS software. The ISO metadata

sections containing Identification and Metadata

Reference elements are suggested to be included at a

minimum for all centerline data layers.

GIS metadata standardization is supported by the FGDC and International Organization for Standardization (ISO).

FGDC-endorsed ISO metadata standards can be found at: FGDC Geospatial Metadata Standards. The Dublin Core

Metadata Element Set can be found at: Dublin Core Metadata Element Set, Version 1.1.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 78 DOT Contract #GS-35F-0001P

September 2014

Include planned, unbuilt

facilities as well as

abandoned or destroyed

roadways in the dataset.

To document published data where embedding metadata within the dataset is not available (e.g., Excel

files, text files), define and mandate metadata standards.

D.2 48BP L A N N E D , D E S T R O Y E D A N D D E C O M M I S S I O N E D R O A D W A Y S

When building and maintaining a roadway GIS network, it is

natural to consider only existing facilities and roadways. But in

many cases, it is important to be able to manage and reference

unbuilt, planned roadways (sometimes called “paper streets”), as

well as alignments that have been abandoned or destroyed.

Doing this allows DOTs to track and manage infrastructure and

assets throughout the entire roadway data lifecycle (see Figure D.1).

Figure D.1: Roadway Lifecycle Diagram

It is recommended that State agencies include planned, unbuilt facilities in their networks, assigning

route identifiers and approximate mileage. Likewise, abandoned or destroyed roadways should be

properly archived. These features should either be designated as such, or kept in a separate database, to

minimize confusion.

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 79 DOT Contract #GS-35F-0001P

September 2014

Always geoarchive data

when significant updates

and changes occur.

D.3 49BG E O A R C H I V I N G R O A D W A Y S E G M E N T S

Archiving the road geometry and associated LRS (e.g., a weekly, monthly, quarterly, or annual snapshot

of the data) is a process known as “geoarchiving.” The lower the amount of data that is included in the

LRS, the easier it is to store snapshots; with more data, different updates might be more appropriate.

For incorporating temporal functionality (i.e., the ability to do

time-based analysis) into a GIS network, it becomes important to

archive data when significant72F

changes occur to the system.

Geoarchiving allows stakeholders and supporting systems to track

changes, details on the changes, and potential impacts to other

data. Geoarchiving is also an important consideration for compliance with record retention rules and

regulations. Further, because most changes happen at the local level, proper geoarchiving procedures

will become even more critical as State DOTs begin incorporating local data.

The recommended means to perform geoarchiving is to have a separate archive data structure

reflecting the road geometry table(s). For geoarchiving, a new data column (ex. ARCHIVE_YEAR) should

be added to the archive table to track the archive date. See Figure D.2 for an example.

Figure D.2: Geoarchiving data structure73F

In this context, the term significant can be defined as changes that can impact dependent systems directly. For

example, a change to an attribute that only affects labeling may not be deemed significant. The meaning of this

term needs to be clearly defined and communicated to stakeholders.

Michael Baker Jr., Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 80 DOT Contract #GS-35F-0001P

September 2014

Make data readily available to all

users via web services, and

develop a consistent change

communication mechanism.

Define a data retention policy

to dictate how long archived

data should be kept.

Data should be copied from the main production road geometry table to the archive table in its entirety

on a scheduled basis. Additionally, a data retention policy will need to be defined, which dictates when

data should be removed from the archive table. To group

changes together, an additional table can be linked to show a

logical data transaction. This structure includes the inherit ability

to “roll back” a grouped transaction if there was a problem with

the update.

D.4 50BR O A D W A Y D A T A D I S T R I B U T I O N A N D C H A N G E C O M M U N I C A T I O N

As external systems gain dependence on an enterprise GIS Roadway dataset, the distribution and

communication of data with dependent stakeholders becomes vital to a system’s success. As data

changes occur within the roadway network, these changes need to be communicated so that

dependencies can be reviewed and changed accordingly. This communication should include the

updated GIS dataset, as well as a change log that describes all

the changes that have been made to the network. Additionally,

dependent stakeholders need a means to communicate change

requests to the GIS roadway network, to maintain data

alignment between the systems.

The drivers and solutions of the three key components of data distribution and change communication

are further described below:

 Data Publishing:

o Driver:

 Dependent systems and users should have a way to directly connect to the most current

published version of the GIS network. It should be in a known, fixed location, easily

accessible to all users.

o Solutions:

 Utilize a feature service or map service (i.e., a specialized web service) to distribute the GIS

network to end users.

 Additionally, the agency may choose to distribute the network through a downloadable

geospatial data format. This would allow for off-line data use; however, it would put the

responsibility on the user to get the “most recent” version as needed. It is advisable that this

file format should not be vendor dependent but allow users to view the data using a variety

of different GIS tools.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 81 DOT Contract #GS-35F-0001P

September 2014

 Change Log:

o Driver:

 Stakeholders need a way to review a descriptive change log detailing the date and reasons

for roadway changes. This allows users to understand roadway changes and determine how

the changes affect their particular business needs. For example, a roadway realignment may

require additional field inspection of asset installations, pavement conditions, and roadway

attribution. In contrast, the addition of a simple traffic signal to augment a non-signalized

intersection may not require a full field inspection.

o Solution:

 A change log and complete metadata should be linked to the map service so that users can

view a detailed description of each change.

 Change Requests:

o Driver:

 Users should have a means to communicate change requests to the GIS network.

o Solution:

 This can be accomplished through a variety of means. For example a “red-lining” map

interface would allow users to identify the area of change, mark a spatial location, and

provide details on the requested change to the GIS maintenance group.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 82 DOT Contract #GS-35F-0001P

September 2014

A P P E N D I X E 29BC R E A T I N G A N I N T E G R A T E D A L L - R O A D S N E T W O RK

When creating an integrated all-roads network, whether at the State or national level, many of the same

processes will need to be completed in order to ensure data alignment and network continuity. This

integration of geospatial data is known as conflation. Conflation is the process used to reconcile datasets

from multiple sources, optimizing their quality and usability. For road data, the likely sources and

partners are described in Report 2 of this Study, on data collection.

Figure E.1 provides an overview of the entire process of integrating road data. The process is described

in detail in Sections E.1 through E.7.

Figure E.1. Overview of Roadway Data Integration74F

Once data is collected, conflation involves the unification of multiple datasets into a single dataset with

combined spatial features and attributes. The datasets may have common and uncommon features and

attributes, which require evaluation and reconciliation. There are decisions to be made about feature

matching (i.e., identifying corresponding features); feature additions and deletions (i.e., deciding what

to include or not include, such as removing obvious duplicates); spatial adjustments (i.e.,

rubbersheeting, adjusting gaps and overshoots); and attribute transfer (i.e., combining non-spatial

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 83 DOT Contract #GS-35F-0001P

September 2014

Create a data inventory,

including metadata, for all

data sources to be integrated.

descriptive data). The process is a combination of automated and manual steps for transforming and

aligning features and includes evaluating and validating changes and adjustments, which are based on

business rules about what is allowed. The following sections address the key issues and steps for

performing conflation and achieving data integration, which are also laid out in Figure E.2.

Figure E.2: Steps and Interim Outputs in Data Conflation and Integration Process75F

E.1 51BD A T A C O L L E C T I O N & C A T A L O G I N G

Different sources may have different map projections, levels of completeness,

currentness, and consistency. A cataloging or data inventory step to document the

contents of the datasets, including metadata generation, is important when

conflating data from different sources. The source data may be stored as

transactional data in a schema optimized for

performance during data collection operations

(e.g., insert, update, delete), but not necessarily

optimized for queries. These datasets are typically

part of an On-Line Transaction Processing (OLTP)

environment. Part of the cataloging step is to document the data schema as well as the actual populated

contents of the source datasets. This becomes useful in optimizing any aggregated datasets to support

indexing and a faster query response, which is typical of On-Line Analytical Processing (OLAP)

environments, such as data warehouses.

Applied Geographics, Inc., 2014

E.1

E.2

E.3

E.4

E.5

E.6

E.7

LRS & Routing

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 84 DOT Contract #GS-35F-0001P

September 2014

To streamline data loading and

conflation, create a staging dataset

that contains the pertinent subset of

features from each source dataset.

It is also important during data cataloging to understand and document any known issues or limitations

with certain datasets. The inventory and metadata that result from this step will inform the rest of the

processes described below.

E.2 52BD A T A E X T R A C T I O N F R O M I N P U T S O U R C E S

As described in task 2 of this report, source data can come from a wide variety of

agencies including:

 Local Government (e.g., city, town, county)

 Regional Entity (e.g., Regional Planning Agency (RPA) or Metropolitan Planning

Organization (MPO)

 State GIS Clearinghouse

 U.S. Census Bureau (TIGER Fdata)

 Federal Agency (e.g., BIA, BLM, USFS, NPS, etc.77F

)

 Commercial Data Provider (e.g., HERE, TomTom, Google)

 Volunteered Geographic Information (e.g., Open Street Map (OSM))

These datasets will all be in varied formats and locations,

and they may contain extraneous data and geospatial layers

that are not pertinent to the integration project. Thus, it is

recommended that a “staging” dataset be created for each

source dataset that contains only the pertinent features to

be conflated. For example, if the data extends beyond the

geographic extent of a State (e.g., TIGER, OSM), it is recommended that the staging dataset be a subset

of only the features in the pertinent geographic extent. Or, if the plan is to conflate local roads from the

local government data with county roads from county data and State roads from the DOT, each staging

dataset might contain only the subset of features that will ultimately be integrated.

BIA = Bureau of Indian Affairs; BLM = Bureau of Land Management; USFS = United States Forest Service; NPS =

National Park Service.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 85 DOT Contract #GS-35F-0001P

September 2014

Data should be evaluated for

consistency and quality using a

combination of automated and

manual procedures.

E.3 53BD A T A P R O F I L I N G

A data profiling step is recommended to evaluate the quality of each dataset, in advance

of data transformation and other pre-processing

steps. Incoming datasets may not be correct,

consistent, or complete. They should be reviewed

and evaluated using a combination of automated

and interactive procedures.

The following is a sample list of typical items to check for during profiling:

 Database values

 Dummy default values, such as 99999 for a zip code

 Missing values, such as no projection parameters

 Contradicting data values, such as when the road construction date is before the project

plan date

 Violation of business rules

 Multi-purpose fields, such as when the same field is used to store information with different

meanings

 Cryptic data, such as when ad hoc abbreviations or truncated road names are used

 Inappropriate use of an address line, such as splitting addresses across multiple lines

 Reused primary keys, such as when an old road and a new road in different places have the

same key

 No unique identifiers, such as when the same road element has several primary keys

 Interoperability flaws, such as when data is inadvertently related but should not be

E.4 54BD A T A T R A N S F O R M A T I O N A N D L O A D I N G

All datasets are not the same. Loading data into the data integrator’s review and edit

environment very often requires transformation. For example, if datasets from local

sources are in a different projection than the State DOT data, a common projection

needs to be chosen, and the data re-projected as needed. Depending on a State DOT’s

workflow procedures and tools, data transformation may be automated as part of the

extraction process from the source. Also, depending on the technology being used, this

can be an intermediate staging database for performing the edge-matching of

features. Edge-matching is explained in a subsequent section.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 86 DOT Contract #GS-35F-0001P

September 2014

When loading source data, only minor

changes should be made (e.g., re-projecting

data, fixing obvious errors). Ideally, the

source data owner would take responsibility

for needed data maintenance.

Match points should be established

to allow edge-matching between

neighboring or overlapping

transportation agencies.

Match points are not indicative of jurisdictional

boundaries or asset management and

maintenance responsibilities. They merely help to

delineate GIS data management responsibilities.

A data cleansing step may also be needed, in accordance with business rules, to assure the integrity of

the data. Such data cleansing may include de-duplication, for example, whereby common features and

attributes that are obvious duplicates are removed. Generally, only obvious discrepancies should be

changed during data cleansing, to assure fidelity

to the source data. Ideally, the supplier of the

data should make the changes, assuming that they

are also the party responsible for data

maintenance. Any changes made by someone

other than the data-provider should be tracked in

a change layer or change report for auditing

purposes.

E.5 55BE D G E - M A T C H I N G A N D M A T C H P O I N T S

Making sure that all road segments are properly edge-matched is a key step

in creating an integrated network. Many DOTs have already edge-matched

most major roadways, but this data enhancement needs to be extended to

local roads as part of the ARNOLD all-roads integration effort, to densify the

road network.

Match points (also known as

integration points, touch points,

demarcation points, smart points,

agreement points, snap-to points,

join points, etc.) are point locations

established within the GIS to mark the connection point

between two (or more) geospatial datasets. These points allow datasets to be seamlessly joined

together without any overlap or gaps (which is essential to proper topology, as described in Section E.6).

In terms of a nationwide ARNOLD, establishing these points between States will be critical in facilitating

the edge-matching of data between neighboring States, and ultimately stitching together a nationwide

roadway dataset.

It is important to note that in this context,

match points are not necessarily indicative of

actual jurisdictional boundaries or asset

management and maintenance responsibilities.

They merely help to delineate where the GIS

data management responsibilities start and stop for a particular agency. The essential idea is to locate

such points at unambiguous, verifiable locations that bordering jurisdictions can agree upon. The

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 87 DOT Contract #GS-35F-0001P

September 2014

agreement can be between the GIS data managers for the adjacent jurisdictions, and ideally would be

endorsed by their respective senior executives.

If match points to facilitate data integration have been agreed upon at the State or local levels, they

should be used. If they do not exist, then a set of recommended points should be presented to the

affected jurisdictions for negotiation and agreement. Feedback and adjustments should be allowed for,

and the points should be incorporated into an agreed-upon Statewide Match Point Layer.

Jurisdictions that have agreed upon match points should be identified during all-roads data collection.

Where they don’t exist, the State DOT can create these points based on evident criteria, and provide

them to Local Government Agencies (LGAs) for review; the LGAs should be given a reasonable amount

of time to review and to provide feedback to the DOT. Most DOTs do not have the resources to mediate

disputes between counties or cities, so the local jurisdictions will need to work this out between

themselves. If the LGAs cannot reach agreement, then the State DOT match points should be used.78F

While the FHWA has a business rule not to make spatial adjustments to the data received from State

DOTs, there are cases where DOTs make spatial adjustments within their State when combining datasets

from different sources. Such adjustments can be based on proximity, topology, and continuity analyses

and rules, as well as attribute comparisons. One method, commonly referred to as rubbersheeting,

adjusts input features to a target, based on rules. Another method involves the automatic correction of

gaps and overshoots, based on agreed-upon geometric tolerances. This method may be combined with

queued edit scenarios, where gaps and overshoots are presented to a human operator for a decision on

whether to make a change. Types of change in this context include: spatial change; attribute change;

spatial and attribute change; no change; new update features; and delete feature.

See: “State of Colorado Unified Road Layer Assessment,” September 30, 2012, p. 46.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 88 DOT Contract #GS-35F-0001P

September 2014

Figure E.3: Edge Matching Scenarios79F

A number of States and jurisdictions (Arizona, the District of Columbia, Maryland, Washington, and

others) are working on using match points to help streamline data integration from multiple sources

(State, county, local, tribal, etc.).

Based on the One Maryland One Centerline initiative80F

, the following list provides an overview of the

Integration Points Workflow81F

 Determine the road centerline segment recognized to be the authoritative source for the given

length of road (e.g., State, county, local)

 Create initial integration points at any intersection of road centerline where the data provider

attribute changes

 Intersect with a bridge polygon layer to identify and remove integration points where roads are

not intersecting

 Provide integration points to local jurisdictions for review, comment and acceptance

 State and locals edit their respective centerline data to coincide at integration points

See: ArcGIS Resources: About Edgematching

The Maryland State Highway Administration (SHA) is responsible for maintaining a statewide road centerline

dataset containing linear referencing route and measure information and an inventory of roadway characteristics.

More information can be found at Maryland iMAP.

From the One Maryland One Centerline initiative presentation at GIS-T 2014 by Erin Lesh and Marshall

Stevenson, Maryland SHA.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 89 DOT Contract #GS-35F-0001P

September 2014

Topology rules and OGC standards should be

applied to and enforced within the roadway

network in order to ensure data quality and

stability, as well as to support routing and

network analysis.

Regardless of the datasets being joined (i.e., those of neighboring States or local and State data), the

same basic workflow can be applied.

E.6 56BLRS A N D N E T W O R K T O P O L O G Y

This step is where important decisions need to be made regarding whether the ARNOLD

baseline layer will support routing (path) topologies for routing and navigational

purposes (e.g., emergency management and traffic modeling) as well as traditional LRS.

To support the decision-making process, this section provides a discussion of topology.

Transportation systems have structure and flow that are commonly represented

graphically as a network. Geometrically, a network comprises an arrangement of nodes

(a.k.a. vertices or points) and links (a.k.a. edges or lines) between the nodes, which

typically represent roads (but also railways, waterways, flight paths, bicycle trails, pedestrian walkways,

etc.). A sequence of links that are traveled in the same direction is a path, which is a fundamental

attribute for measuring traffic flows and finding navigational routes.

The structural arrangement and connectivity of a network represent its topology. Topology is sometimes

referred to as “coordinate-free geometry,” since a topological network can be stretched as if it were a

rubber sheet without changing the “from/to” and “left/right” relationships of the nodes and links. More

appropriately, it should be described as the study of those geometric concepts that can be defined

independent of any coordinate system. Nonetheless, in a transportation context, location, direction, and

connectivity are all important.

Topology rules define key spatial relationships

between corresponding datasets. Examples of

topology rules include:

 Prohibiting disconnected segments

 Disallowing segments missing associated

measures

 Not allowing segments to overlap

 Not allowing segments to have gaps

The roadway network should allow for the creation and enforcement of topology rules, and should

follow the Open Geospatial Consortium (OGC) Standard for Simple Features82F

. Violations of topology

FHWA requires that HPMS submissions meet the OGC standard for simple features (See Open Geospatial

Consortium Standards, Simple Feature Access).

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 90 DOT Contract #GS-35F-0001P

September 2014

Routing and navigation applications

typically require a node at every

legitimate intersection in order to model

items such as vehicle turn restrictions

and directionality, and for geocoding.

rules are stored as errors, but exceptions can be marked accordingly. Network topology can be encoded

and its properties measured using graph theory, which is incorporated at a fundamental level in most

GIS products. Network analysis based on graph theory is commonly supported in GIS data models. LRS

are an extension to GIS, adding dynamic segmentation and allowing State DOTs to measure the location

of features and events on the road network, thus handling linear measurements along a road segment.

In this GIS-T case, the node geometry of a road network might be dissolved along logically continuous

routes. Conversely, in a non-GIS-T case, the road network might be subdivided into a multitude of

segments for graphic purposes to support cartographic representation, or geocoding. These approaches,

while valid for their own purposes, can adversely impact the topology needed for traffic flow and

routing applications. In other words, a dataset that is optimized for linear measurements, or

cartographic representation, or geocoding, is less than optimal for traffic flow modeling and routing.

Topology, particularly for connectivity, forms the basis for network analysis and routing. A routable

network is key to many GIS applications, such as fleet

management, drive-time analysis, and most notably

emergency response. Routing allows first responders to

determine the quickest route from their location to the

dispatch destination. Typically, routing applications

require a node at every legitimate intersection where a

turn can be made, which may differ from the data

structure of the LRS. For example, to achieve processing efficiencies for non-routing applications, State

DOTs have traditionally dissolved segmented networks into non-segmented networks, to reduce the

number of nodes and to define numbered routes with the minimum number of essential segments.

However, for routing purposes, nodes and related intersections are very important for modeling vehicle

restrictions (turns, heights, weights, speeds, direction, stop signs, traffic signals, congestion,

construction, detours, etc.); and for geocoding purposes for matching addresses with a location along a

street.

Routing applications require a network topology in which the relationships of each node and link with

other nodes and links are modeled in terms of constraints and possibilities. Fundamentally, two tables

are required to geospatially represent a network data model: 1) a node table, including unique IDs and X

and Y coordinates for each node; and 2) a link table, also with unique IDs, as well as the “from” node

(the origin or starting point), and the “to” node (the destination or ending point). When these tables are

relationally linked, the topological structure of the network can be built, and network analysis based on

graph theory can be applied.83F

Rodrigue, Comtois, and Slack. The Geography of Transportation Systems. Taylor & Francis Group: New York,

2009, pp. 76-77.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 91 DOT Contract #GS-35F-0001P

September 2014

The network should be built to meet

the needs of routing, and then be

processed to support LRS needs.

E.7 57BO U T P U T D A T A S E T S

To create a road network that supports both routing and linear referencing use cases,

as described in Section 2.1, the network should combine roadway geometry with

network topology in order to meet the needs of routing and navigation, and then be

processed to “thin” the data to suffice for

the needs of LRS. If the network is built

only to support LRS (without topology), it

will be inadequate for routing. The goal

to “build once, use many times” can be achieved through this approach: the base

geometry and topology support multiple business needs.

Node and segment geometry is needed for the creation and maintenance (e.g., adding new routes or

new alignments) of a roadway network. As shown in the diagram below, this geometry, along with its

network topology, can be combined with turn and flow restrictions and address points to satisfy routing

use cases84F

. Similarly, the roadway geometry and network topology can be used to create LRS routes.

These routes, combined with point and line events, can satisfy linear referencing use cases. Frequently,

DOTs are focused on the LRS portion of the diagram (i.e., the bottom output), but through collaboration,

both use cases can be met.

There are different levels of complexity when it comes to routing and navigation. There are general/casual uses

for everyday navigation (e.g., generating directions given “from” and “to” locations), and more complex routing

scenarios, such as routing for heavy load purposes. The network should be designed to handle the full spectrum of

routing and navigation use cases.

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 92 DOT Contract #GS-35F-0001P

September 2014

Figure E.4: Routing and LRS Output Datasets85F

Applied Geographics, Inc., 2014

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 93 DOT Contract #GS-35F-0001P

September 2014

L I S T O F A C R O N Y M S

AADT – Annual Average Daily Traffic

AASHTO – American Association of State Highway and Transportation Officials

API – Application Programming Interface

ARNOLD – All Road Network of Linear Referenced Data

BIA – Bureau of Indian Affairs

BLM – Bureau of Land Management

CAD – Computer-aided design

COTS – Commercial-off-the-shelf

DOT – Department of Transportation

ETL – Extract, Transform and Load

FGDC – Federal Geographic Data Committee

FHWA – Federal Highway Administration

FMIS – Fiscal Management Information System

GIO – Geographic Information Officer

GIS – Geographic Information System

GIS-T – Geographic Information Systems for Transportation

GPS – Global Positioning System

HPMS – Highway Performance Monitoring System

ISO – International Organization for Standardization

LGA – Local Government Agency

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 94 DOT Contract #GS-35F-0001P

September 2014

LRM – Linear Referencing Method

LRS – Linear Referencing Systems

MIRE – Model Inventory of Roadway Elements

MPO – Metropolitan Planning Organization

NCHRP – National Cooperative Highway Research Program

NENA – National Emergency Number Association

NG911 – Next Generation 911

NPS – National Park Service

NSDI – National Spatial Data Infrastructure

OGC – Open GIS Consortium

OLAP –On-Line Analytical Processing

OLTA – On-Line Transaction Processing

OSM – Open Street Map

ROW – Right-of-way

RPA – Regional Planning Agency

SHA – State Highway Administration

SLD – Straight Line Diagram

TFTN – Transportation for the Nation

TIGER – Topologically Integrated Geographic Encoding and Referencing

TIP – Transportation Improvement Program

USFS – United States Forest Service

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 95 DOT Contract #GS-35F-0001P

September 2014

VGI – Volunteer Geographic Data

VMT – Vehicle Miles Travelled

WKB – Well Known Binary

WMS – Web Map Service

_____________________________________________________________________________________________

All Public Road Geospatial Representation Study ARNOLD Reference Manual

Technical Appendix Page 96 DOT Contract #GS-35F-0001P

September 2014