Oregon communities are using artificial intelligence to help public agencies understand, plan and maintain their transportation systems—saving time and money

 

By: Carl Springer

When you take your car in for maintenance, the first thing the mechanic does is hook the car up to an electronic diagnostic tool to examine how your car’s systems are working. That tool gives them a full readout on what’s happening: which systems are working, which are degraded, and where any urgent issues lie. Without that diagnostic baseline, any repair on your car is just educated guesswork.

The same principle is true for transportation agencies—those run by states and cities—that manage transportation systems. A comprehensive transportation system inventory is also a valuable diagnostic tool—it tells agencies what assets they have, where assets are located, what shape they are in, and how these assets changed over time. It is a critical first step in deciding where to invest funds, where to prioritize maintenance and where and when to apply for funding.

In other words, the transportation system inventory is the foundation for smart, cost-effective, and accountable infrastructure management.

The difficulty comes in the time and resources these agencies need to get to that system inventory. Designing and completing one is costly and until the recent past, usually took months or even years to complete. Conducted by a team of people on the ground armed with notepads or tablets, agency staff had to wait for the team’s observations and then synthesize them. A few years ago, the city of Vancouver, Washington undertook a curb and ramp assessment of its sidewalks, in order to evaluate their condition, slope, damage and tripping hazards. A team of interns armed with iPads walked every section of sidewalk in the city for this valuable analysis—yet the project took two years to complete.

Today, technology has advanced to speed both the collection and analysis of transportation system data. New mapping technologies, including high-resolution satellite cameras, as well as artificial intelligence and machine learning programs are providing data gathering and data analysis capabilities that can produce  more complete and accurate inventories in a fraction of the time. This is improving safety for users, helping agencies manage costs and ensuring that maintenance investments and decisions build a system that is responsive to future needs.

System-level data from the ground up

The Oregon Department of Transportation (ODOT) and the Department of Land Conservation and Development (DLCD) have long supported policies and projects to reduce emissions that are contributing to climate change. In 2022, they updated the transportation planning rules to require a more granular assessment of how walking, bicycling and transit can support these policies.

The challenge was that many local agencies lacked the on-the-ground data required to accurately make these assessments. To rectify this, the state agencies committed to provide the initial inventory to all cities and counties within Metropolitan Planning Organizations (MPO).

In 2024, DKS Associates began leading a project for ODOT to gather multimodal transportation system data for 48 cities and 11 counties in eight MPOs across Oregon that participate in regional transportation planning. The focus: compile location and attribute data for all streets and roads within these 59 local agencies, enabling them to have the complete datasets required for assessing how people use all modes of travel. The final product? An updated set of data files in geographic information system (GIS) format for each agency within the MPOs, and Long-Term Data Management and Maintenance Plan by the end of 2026. The agencies were grouped into three cohorts, and the first GIS datasets will be delivered beginning next month.

This project is the equivalent of multiple teams of people reviewing about 17,000 miles of road—something that would have taken months or even years for cities and the state to collect and analyze. In fact, some of the GIS information on the condition and use of the system will be new even to the state’s larger cities, including those with the most advanced active transportation networks.

The DKS team includes two companies with expertise in capturing imagery and converting the imagery into GIS data, Vexcel Data Program and Ecopia.

Vexcel captures aerial imagery using a combination of advanced aerial camera systems and specialized aircraft, deploying its own high-resolution aerial cameras into planes that fly predetermined routes over cities. Vexcel’s high resolution cameras are so powerful that one pixel of an image equates to three inches of road—providing an incredible amount of detail about the condition of small features in the roadway, even showing the condition of bike lanes and the location of things like handicap access ramps.

Analyzing these detailed images and translating them into valuable GIS layers is where Ecopia comes in. Ecopia uses advanced AI-powered image analysis (also known as machine learning) to convert high-resolution aerial imagery into high-precision GIS data. The company’s  deep learning models are trained to recognize and extract features from imagery, with specific detail revealing buildings, features of parks and greenspaces and of course, the condition of roads, bike lanes and sidewalks, giving transportation asset managers a level of sophisticated data that can drive efficiencies and meaningful system improvements.

DKS is responsible for managing these efforts, providing analysis and oversight of all work products delivered to the state. The entire team coordinates with ODOT and local local transportation agency staff members in all eight MPOs to develop the methods and guidelines for incorporating these new datasets into the existing transportation GIS datasets managed by cities and counties.

Filling information gaps across communities

This kind of GIS data is a huge opportunity for public agencies to have complete, multimodal system level analysis that fills current data gaps at the state and local level. It will be essential in helping Oregon’s public sector transportation managers understand exactly how their system is performing, and how it is aligning with updated state policies and planning rules. For example, cities can now evaluate how safe walking and cycling to key destinations is in given neighborhoods, like safe routes to schools, shopping or medical care. Prior to this project, many of these cities did not have access to this kind of granular data.

The benefits of this kind of data are numerous, and include:

Allowing a view into disparities in a city’s transportation system. This data will help smaller communities gather and use data they would otherwise not be able to get, giving them a holistic view in a single data layer of their own system conditions—something to which many cities have never had access. It will help uplevel a city’s understanding of their pedestrian and bicycle safety needs and provide a clear picture of transit accessibility, needs which are not tracked as often as road conditions.

Improves safety by revealing the true condition of transportation assets. This kind of GIS data can uplevel a city’s understanding of pedestrian and bicycle safety in different neighborhoods, needs which are often not tracked as often as road conditions. This better GIS data can facilitate a more robust analysis of bicycle and pedestrian facilities, similar to the kinds of analysis done of the road network. For example, the GIS data can reveal where safe travel sheds (quality sidewalks, safe pedestrian and bike crossings, adequate streetlighting) exist, and where they are lacking.

Guides future investment decisions. The GIS data created through this project will provide  current information on which cities can base transportation investment decisions. For example, if a particular area of the city is revealed to have a lack of safe travel sheds, policymakers and agency staff could guide more investment there, improving conditions for people on the ground. This intent is at the heart of the state’s recent update to transportation planning rules, developed to enable local communities to strategically and quantitatively leverage transportation investments towards greater safety, access and convenience of travel for all users.

Better maintenance and more efficient investments over time

Transportation agencies face growing demands to modernize infrastructure, which are occurring hand-in-hand today with difficult budget gaps. The detailed data provided by this project offers a transformative solution for these times—delivering accuracy and scalability in decision-making without a massive investment in data acquisition. Streamlining transportation system inventories provides a way to achieve policy goals across modes in ways that deliver on safety, performance and emissions reductions.

Oregon’s path to safer, more climate-friendly transportation is clearer than ever—and with the smart use of technology and artificial intelligence, cities and states across the country could be equipped to navigate this path with confidence.