Revolutionizing Transportation Planning with AI-Powered Data

Abigail Coholic, Senior Director of Partnerships at Ecopia AI (Ecopia), recently participated in a webinar with geospatial experts from the Maricopa Association of Governments (MAG), Replica, and Bentley Systems. They discussed new opportunities for transportation planners to foster safer, more accessible, equitable, and sustainable communities. This blog highlights the key points from their insightful discussion, which is part of a larger series on GIS for state and local government asset management. A recording of the webinar is available below.

Using big data to understand granular transportation trends with Replica

The webinar begins with Arthur Getman, Senior Solutions Engineer at Replica, providing insights into how Replica leverages big data to understand transportation trends in detail.

Arthur explains that Replica combines various data sources, including census data, consumer and resident information, location data, economic activity, and ground truth data to provide a comprehensive understanding of who is living and traveling in cities and states, and why and how they are moving around.  

Arthur shares that Replica’s process involves creating a synthesis of the population that is then matched to real-world data down to the block group level, ensuring that socioeconomic and demographic splits align with census data. Replica then trains behavior models using granular data and site-specific movements to understand how different people interact with the built environment. This includes factors like age group, income bracket, residence location, and work industry. The final step involves calibrating the outputs by comparing them to things like transit counts, ridership numbers, traffic counts, and bike and pedestrian counts to help ensure the accuracy of their simulations. 

Arthur explains that this data is valuable for supporting transportation planning and decision-making across a wide range of scenarios. For instance, detailed data about drivers on an interstate is essential for planning improvements on an exit ramp. Planners need to understand not only the traffic volume but also the origins and destinations of these drivers. This insight helps gauge how changes will affect the overall transportation network. 

Arthur elaborates on use cases of Replica’s data, sharing an example from New Jersey. He explains how land use types generate different trip types, which can be analyzed to inform planning decisions. For example, Replica’s data can be filtered to compare trips originating from single-family residential areas versus multi-family residential areas. Arthur points out that data shows that trips from single-family residential areas lead to significantly higher vehicle miles traveled (VMT), and higher-density residential developments typically result in lower VMT and more sustainable transportation choices. This underscores the importance of considering land use in housing and development plans. By analyzing how different residential types affect VMT, planners can better understand how to reduce emissions and create more sustainable communities.

Arthur highlights additional examples of Replica’s data for planning applications including a bike route planning example from Denver, where Replica’s data was used to analyze users of the Cherry Creek Trail. This analysis provided insights into popular connections, trip distances, purposes, times of day, and trends over time, aiding in effective bike route planning.

Harnessing high-precision data for more resilient and equitable communities with Ecopia AI

In the next portion of the webinar, Ecopia’s Abigail Coholic explains how Ecopia is empowering communities with AI-powered mapping to help ensure safe, equitable, and sustainable communities. 

Transportation agencies at all government levels aim to create safer networks, increase active and accessible transportation, and reduce carbon footprints. To achieve these goals, a thorough understanding of the current transportation system and its assets is essential. This foundational knowledge allows for the identification of necessary improvements and informed decision-making. For instance, selecting appropriate safety measures, ensuring access to community resources for those without vehicles, and promoting zero-emission travel all require a detailed understanding of existing infrastructure. Abigail highlights that Ecopia delivers this essential geospatial data to cities and transportation agencies, providing a precise and current view of their infrastructure, which is crucial for effective planning and decision-making.

Overcoming challenges in geospatial data: from fragmented sources to AI-powered solutions

Abigail explains that transportation agencies have long struggled with accessing up-to-date, high-quality foundational map data, facing trade-offs between affordability, accuracy, speed, and scalability. For example, obtaining accurate data might require cities to manually digitize features, a process that is time-consuming and costly, making frequent updates impractical. Conversely, many automated software solutions that claim to produce features often capture only 60-70% of the necessary data and require significant manual corrections to address inaccuracies.

Abigail highlights how Ecopia’s technology is addressing these challenges. Ecopia’s AI-powered systems leverage high-resolution aerial imagery from a vast network of partners to convert it into high-precision vector data, eliminating the need for manual digitization but maintaining the accuracy of a GIS professional. Abigail provides an example of detailed transportation features that Ecopia extracts, including individual road lanes, turn lanes, bike lanes, shoulders, lane medians, crosswalks, driveways, sidewalks, and more. She explains that this comprehensive, accurate, and up-to-date data enables Departments of Transportation (DOTs), Metropolitan Planning Organizations (MPOs), Councils of Government (COGs), and cities across the US to make informed investment decisions, enhance infrastructure improvements, and positively impact communities. Abigail shares that Ecopia's AI also normalizes data, allowing for true updates rather than complete map recreations, which is crucial for monitoring the impact of decisions and adjusting strategies accordingly.

Understanding safety risk with shoulder presence and width

Abigail demonstrates how Ecopia’s data advances transportation planning by offering detailed insights into infrastructure features, such as shoulders, which are crucial for conducting thorough safety analyses. She notes that while urban areas experience a higher number of crashes, rural roads have a higher crash fatality rate. Infrastructure features, such as shoulders, can act as a buffer to reduce collision risks and fatalities in these areas. Accurate mapping of shoulder presence and width offers planners and engineers a real-world understanding of safety risks and helps identify areas for improvement.

Abigail explains that though studies indicate that shoulder infrastructure can mitigate crash severity by providing a safety buffer, maintaining accurate data on rural roads is challenging, as shoulders are often mapped infrequently or estimated based on guidelines, leading to potential inaccuracies. For example, varying shoulder widths, from 2 to 8 feet, can result in overestimations of safety protection, which may inadvertently increase crash risks. In contrast, Ecopia’s data, derived from high-resolution imagery, quickly provides precise and up-to-date mapping of shoulder locations and widths across the entire network. This detailed information enables agencies to assess risks accurately, develop effective safety strategies, and work towards their goal of reducing fatalities and ensuring safe conditions for all road users, including pedestrians and cyclists.

Optimizing safety with strategic median placement  

Next, Abigail explains that medians are another cost-effective safety measure widely used in transportation networks. While decisions on median placement are influenced by crash data and daily traffic volume, understanding the existing infrastructure—such as sidewalks, crosswalks, intersection controls, and lane types—is essential. Safety guides recommend evaluating these factors alongside the current presence and type of medians. By integrating comprehensive, accurate, and up-to-date data on physical infrastructure with information on road volumes, crashes, and land use, planners can systematically identify the optimal locations for median installation and other safety countermeasures. Abigail explains that this data-driven approach makes safety investment decisions proactive and evidence-based. She highlights that these data layers also enhance safety for vulnerable road users and support the adoption of zero-emission transportation alternatives. Abigail illustrates that without a clear understanding of the existing system, prioritizing these changes within capital constraints can be difficult. The detailed data that Ecopia offers helps decision-makers prioritize strategies effectively, ensuring that safety improvements are both impactful and feasible. 

Enhancing pedestrian safety with sidewalk and bike lane data

The webinar continues with Abigail explaining that Ecopia digitized San Bernardino County’s entire transportation network in just 3 months—a task that would have taken years with manual methods. This included over 17,000 miles of sidewalks and all bike lanes in the region, providing a comprehensive view of the system, identifying infrastructure gaps, and enabling the prioritization of improvements.

Abigail shares an image of Ecopia’s sidewalks categorized by width, created in San Bernardino. She highlights the importance of detailed sidewalk features for safety analysis, noting that overlaying sidewalk gaps with census data on car access can support equitable investment decisions. She also points out that sidewalks narrower than 3 feet do not meet Americans with Disabilities Act (ADA) or Public Right-of-Way Accessibility Guidelines (PROWAG) standards, making them uncomfortable for pedestrians. Abigail explains how Ecopia’s data helps identify these problem areas, enabling agencies to prioritize improvements that enhance accessibility, comfort, and safety, thus supporting broader community goals.

In a similar vein, Abigail explains that Ecopia’s data layers are vital for identifying and improving bike lanes by offering detailed information on both protected and unprotected lanes, and their widths. She explains that analyzing this data alongside details about lane configurations, average right turn lengths, and on-street parking frequency provides a comprehensive understanding that helps design bike lanes to be safer and more comfortable for cyclists. For instance, in San Bernardino, 0.63 miles of bike lanes were found to be under 2 feet wide, posing discomfort and collision risks, while 3.6 miles of 3-foot lanes, mostly protected, offer improved comfort and safety. By integrating this data with other infrastructure details, traffic conditions, and land use information, planners can make informed investment decisions to enhance connectivity and improve the overall network.

Powering the Arizona Sun Cloud Portal to support transportation solutions

Abigail shifts the focus to how Ecopia’s Advanced Transportation Features have supported the Arizona Sun Cloud portal. This portal offers high-quality transportation and socioeconomic data to assist communities across Arizona’s Sun Corridor megaregion in making informed transportation planning decisions. Abigail outlines Ecopia’s collaboration with the Maricopa Association of Governments (MAG), explaining that in just 5 weeks, Ecopia delivered detailed transportation features to support the Arizona Sun Cloud. The data included sidewalk polygons, bike lanes (both painted and unpainted with width attribution), various types of medians (raised, with vegetation, painted, and split), and additional infrastructure data spanning the 30,000-square-mile region. Abigail then introduces Ted Brown, the Deputy Transportation Director for Data Management and Application at MAG, to provide more insights into the project.

Ted explains that the Sun Cloud project stems from an accelerated innovation deployment grant awarded in 2017 to MAG, the Arizona Department of Transportation, and other partners. This grant aimed to develop a data portal that enhances the accessibility, usability, and quality of data and analytical tools for transportation planning across Arizona’s Sun Corridor megaregion.

Integrating Ecopia’s median data for predictive safety evaluations

Including the data provided by Ecopia, Ted explains that the portal features 22 distinct layers, encompassing traditional transportation datasets such as traffic counts, model data, and asset conditions. He emphasizes that safety is a key focus area, detailing how MAG used the Highway Safety Manual methodology to identify impactful solutions by integrating various datasets, such as crash data, travel demand models, and safety performance functions. Ted highlights that Ecopia’s median data - providing detailed information about the presence and width of medians - was crucial for safety analyses. He explains that this enabled a more advanced project evaluation, transitioning from a crash rate-based approach to a predictive methodology.

Examining bike lane stress metrics with detailed data

Ted also highlights the bike lane and sidewalk layer available in the portal. He highlights that Ecopia supplied detailed data on the presence, type, and width of bike lanes throughout the entire arterial network. This information has proven invaluable, aiding in the validation of their in-house bikeways map and playing a crucial role in establishing a level of traffic stress metric for all bike facilities.

Improving equity evaluations through detailed user analysis 

Ted then discusses the disadvantaged facility users layer, which represents a significant advancement in equity scoring. He shares that previously, equity assessments were based solely on proximity to priority communities, a method that did not account for actual roadway users. In contrast, the new approach involved partnering with Replica to conduct a select link analysis on each network segment, yielding detailed reports on user numbers and home locations. By integrating this data with information on disadvantaged communities, MAG was able to calculate the percentage of users from these communities, allowing for more accurate equity evaluations. Ted shares an example of "South Extension Road," which initially did not meet the criteria for disadvantaged communities. However, the new analysis method revealed that over 70% of its users come from such communities, providing a more nuanced and actionable understanding of equity in transportation planning.

Ted explains that the Sun Cloud platform is designed for flexibility and ease of use, offering tools for customizing symbolization, filtering by attributes or geography, and exporting data for further analysis. He emphasizes the platform's adaptability with a custom widget that allows users to set their own weights and rescore data quickly, catering to diverse agency needs. Ted then demonstrates how to switch between different data views, apply filters, and export maps, highlighting the platform's practical applications for grant applications and other data-intensive tasks. He also mentions the ongoing efforts to secure funding for platform updates, noting that while some data remains relevant for years, other layers may need more frequent updates. Ted concludes by providing links to additional resources and introducing the next speaker, Glen Franklin.

Integrating and analyzing reality data

In the final portion of the webinar, Glen Franklin, Application Engineer at Bentley Systems, explained Bentley's iTwin Capture platform, and methods for importing, managing, and integrating this data to support various applications that Bentley Systems offers. He explains that the process begins with capturing reality and GIS data using remote sensing units, which are then analyzed through manual, semi-automated, or AI-driven methods. The analysis results are federated and shared online, allowing for further integration and use across different stakeholders and applications. Glen emphasizes that while GIS data is a component of reality data, it plays a crucial role in the process, often serving as the starting point for analysis and subsequent integration into GIS formats.

Glen then highlights the practical applications of Bentley’s technology through various examples. For instance, by converting point cloud data into classified objects, the platform can analyze and extract useful information such as telephone pole attributes or tunnel components. He also discusses how AI detectors improve the classification of unstructured data and how the resulting GIS data can be shared with third-party applications like ArcGIS and QGIS. This integration ensures that updated and accurate data is available for further analysis and application. Glen concludes by noting the ongoing evolution of AI technologies that enhance the extraction and usability of reality data, stressing the importance of flexible integration options to support a range of GIS applications and workflows.

Driving a safe and sustainable future with accurate data

Our world is constantly evolving, and keeping up with these changes is essential. As cities grow and transform, outdated data can hinder effective policy-making and the monitoring of essential safety, equity, and sustainability initiatives. A comprehensive understanding of infrastructure enables more informed decisions that enhance road safety for vulnerable users, advance Vision Zero goals, support air quality and CO2 reduction strategies, and more. Ecopia provides high-precision geospatial data with the speed, affordability, and reliability necessary for effective decision-making. To learn more about Ecopia’s Advanced Transportation Features dataset, please get in touch or view the full webinar below.