American Association of State Highway and Transportation Officials

Special Committee on Research and Innovation

 

FY2023 NCHRP PROBLEM STATEMENT TEMPLATE

 

Problem Number:  2023-B-31

 

Problem Title

Characterizing economically disconnected areas in the era of Artificial Intelligence.

 

Background Information and Need For Research

Transportation equity has long been a persistent issue that transportation officials have been trying to address for decades.  While significant efforts have been made, achieving equity is yet a persistent challenge that may not always be easily realized. Equity in transportation could be related to mobility and accessibility, equal balance of cost and benefits, spatial distribution of costs and benefits, and/or creating opportunities for travelers to make travel decisions.  Problems related to these equity aspects often result in creation of economically disconnected areas in cities around the world and in the U.S. 

Recent advancements in transportation technologies have resulted in the ability to collect data with unprecedented variety and amounts (Big Data) which can be leveraged to improve mobility, safety, and the environment.  Yet, whether and how such data could be best utilized to address the equity problem remains an unanswered question. For example, lack of transportation service in many areas, especially the suburban and rural areas, may lead to unintended biases when analyzing the transportation system behavior. 

Advanced data analytics and artificial intelligence (AI) have offered ample opportunities to address data scarcity and associated equity issues.  AI techniques can help to effectively extract and understand structure and patterns underlying the data and make robust inferences for future conditions and trends.  With these capabilities, it is envisioned that AI can help analyze effects of previously implemented strategies to address a variety of equity issues related to transportation safety, mobility, accessibility, affordability, resiliency, and recreation.  In essence, this research aims to address the following key questions:

1-         What data is available that can be used to address the equity in transportation? What data gaps need to be addressed? What traditional and untraditional sources can help close these gaps? What methodologies and techniques should be implemented to collect these data?

2-         How can these data be used to identify economically disconnected districts? What are the characteristics (socioeconomic, transportation system related, etc.) that should be used to identify these districts?

3-         What system-wide strategies have been and should be implemented to close the gaps?

4-         What tools could be used to empower economically disadvantaged groups to close the equity gap (provide more “personalized” resources) at the individual level?

5-         What framework should be adopted to inform decision makers based on lessons learned from Question 3 above? How the public and private stakeholders could be effectively involved and collaborate within this framework? How could this framework be resilient and adaptable to mega-level events such as COVID-19 to ensure equitable access to diverse opportunities?

It is expected that the research team will leverage advanced analytical techniques and emerging AI technologies [data fusion of structured and unstructured data, predictive analytics, pattern recognition, etc.] to address the research questions above.  Addressing these questions will result in an effort that will align with the goals of the AASHTO Committee on Planning to “Promote a multi-disciplinary planning process that encourages State DOTs to define a long-term vision; establish a framework for system and project-level actions; integrate the perspectives of the many and varied stakeholders across modes, sectors, and jurisdictions; and promote an alignment between vision and program implementation”, and  “Support State DOTs in the understanding and use of models, tools, analysis techniques, methodologies and professional development that are important in implementing a multi-modal transportation planning process”.

 

Literature Search Summary

Transportation planning decisions often have significant equity impacts. Particularly due the emergence of new technologies and transportation services (such as transit, automated vehicles, electric vehicles and charging stations, ride-share services, and connected vehicle environments), transportation equity issues have brought a significant attention to policy makers.  In fact, recent studies noted that the benefits from such consumer-facing technologies are usually not fairly distributed among population groups, leading to equity issues.  A comprehensive review of literature on the TRID database and other databases showed that while significant efforts have been directed to address the equity issue, such efforts were mostly directed on the policy side.  Whether other approaches were investigated, including extracting indicative measures using advanced data analytics and AI from past experiences, was not part of the driving force for such policies.   The National Academies Press’s Report on the 2019 Critical Issues in Transportation identified equity as one of twelve areas with pressing needs.  Additionally, the report indicated that artificial intelligence may offer an opportunity to “guarantee significant future changes and benefits”.  Therefore, the targeted research herein aims to investigate how AI tools can help guide an effective decision-making process to help address the transportation equity challenge.

 

Research Objective

The objective of this research is to develop a comprehensive framework that can be used by departments of transportation, metropolitan planning organizations and municipalities to address transportation equity issues in their jurisdictions. The framework should (I) provide science-driven methodologies to identify and diagnose the causes of transportation equity issues and characterize economically-disconnected districts in urban areas; and (II) provide a tested process that can be used by the different stakeholders to develop comprehensive plans to address transportation equity issues in their jurisdictions.  Accomplishment of the project objective will require working on at least the following tasks.

Task 1. Provide a comprehensive literature review on transportation-related equity issues in urban areas and their adverse social and economic consequences.  The survey is expected to provide a categorization of the different equity issues with respect to for example, mobility, accessibility, safety, environmental quality, and recreation.   Additionally, the review should cover transportation system-based strategies (e.g., information systems, public transportation, toll subsiding, carpooling, shared-economy mechanisms, etc.) that were successfully adopted to address transportation equity in urban areas.

Task 2. Develop science-driven methodologies to identify equity issues and characterize economically disconnected districts in urban areas. The task should identify data requirements and describe the sources and methods to collect this data.

Task 3. Validate the methodologies developed in Task 1 by demonstrating how they can be used to diagnose transportation equity issues and characterize economically disconnected district in at least two different urban areas.

Task 4.  Develop an AI-based process that can be used by the stakeholders to construct a comprehensive plan that integrates a subset of the strategies identified in Task 1 to address transportation equity issues in their jurisdictions.  The process is expected to account for possible synergies and conflicts among these strategies, their cost of deployment and how they contribute to equity goals defined for each jurisdiction.     

Task 5.  Demonstrate the application of the process developed in Task 4 for at least two different urban areas (testbeds) and illustrate - using appropriate predictive modeling techniques - the effectiveness of the integrated plans developed by this process.

 

Urgency and Potential Benefits

State and local DOTs are facing increasingly challenging problems in transportation operation, management, and safety.  The recent literature shows that AI and ML tools present unique opportunities to address many of those complex problems and issues.  AI research in transportation has advanced recently at an unprecedented rate and will continue to shape the future of transportation research by offering more effective solutions to unique transportation problems in many areas.  Yet, as we brace for a new intelligence revolution wherein AI and ML will play a significant role in decision support systems, it is not clear to State DOTs what implications and opportunities AI will have on critical issues such as transportation equity, fairness, diversity, policies, workforce development, etc.  This research needs statement highlights the importance and potential of addressing these issues with AI tools.  This will ensure that State DOTs are well prepared for and have enough understanding of the potential implications of AI in the broad transportation domain.

 

Implementation Considerations

The planning divisions of state DOTs and MPOs will likely be the direct users of the results and products generated from this research.  The goal of the project is to help state DOTs and MPOs to adopt practical data-driven procedures to incorporate due equity considerations into their long-range transportation planning process.  The AI-based framework developed from this research is expected to be generic and flexible to capture equity aspects at different spatiotemporal scales to ensure transferability. This will help to facilitate adoption at the state level by considering the availability of data and computing resources, which may vary significantly across states.  To guarantee successful implementation of the anticipated products from this research, state DOTs will need to identify and leverage their existing data sources. Data exchange and fusion from related divisions within state DOTs and MPOs are anticipated. Data sources from the private sector may be required depending on the desired level of sensitivity.

 

Recommended Research Funding and Research Period

Consistent with other recent framework development efforts under the NCHRP program, the recommended research fund to perform this research is $650,000, including $635, 000 for accomplishing the research objectives and $15,000 for online surveys required to collect the data required for the research.  The recommended research period is 36 months.

 

Problem Statement Author(s): For each author, provide their name, affiliation, email address and phone.

           Southern Methodist University, Khaled Abdelghany, 214-768-4309, khaled@lyle.smu.edu [AED50 Member]

           University of Tennessee, Osama A. Osman, 423-425-4398, Osama-osman@utc.edu [AED50 Member]

           Old Dominion University, Sherif Ishak,  757-683-5572, sishak@odu.edu [AED50 Member]

           University of Georgia, Jidong Yang, Jidong.Yang@uga.edu [AED50 Member]

           University of Missouri, Yaw Adu-Gyamfi, 573-882-7546, adugyamfiy@missouri.edu [AED50 Member]

           University of South Alabama, Min-Wook Kang, 251-460-6174, mwkang@southalabama.edu [AED50 Friend]

           Virginia Transportation Research Council, Mo Zhao, 434-293-1938, Mo.Zhao@VDOT.Virginia.gov [AED50 Member]

 

Potential Panel Members: For each panel member, provide their name, affiliation, email address and phone.

 

Person Submitting The Problem Statement: Name, affiliation, email address and phone.

           Julius Codjoe

           225-767-9761

           Julius.codjoe@la.gov

           Louisiana Department of Transportation and Development/Louisiana Transportation Research Center