BACKGROUND

Artificial intelligence (AI) is rapidly transforming many industries, and the transportation sector is no exception. AI-based decision support systems (DSS) have the potential to revolutionize the way state departments of transportation (DOTs) operate and manage their transportation networks.

AI-based DSS can be used to collect, analyze, and interpret large amounts of data from a variety of sources, such as traffic cameras, sensors, mobile phones, and social media. This information can be used to generate insights and recommendations to help DOTs make informed decisions about a range of topics from traffic management and infrastructure planning to daily DOT agency operations. At the same time, there are numerous challenges and factors that need to be addressed including, but not limited to, developing AI-based DSS that are accurate and reliable in real-world transportation settings, AI-related processes that can be integrated into state DOT’s workflows, and AI tools that can be used by state DOT staff.

Research is needed to assist state DOTs in developing ground rules for the implementation of AI-based tools to make better decisions more quickly and efficiently.

The structure for these ground rules shall be aligned and support the American Association of State Highway and Transportation Officials (AASHTO) 2021-2026 Strategic Plan goals and objectives:

• Safety, mobility and access for everyone;
• National transportation policy leadership; and
• Organizational excellence with world-class services.

OBJECTIVE

The objective of this research is to develop a toolkit for the successful conceptualization, development, evaluation, and adoption of AI-based DSS solutions by state DOTs.

The toolkit shall leverage and build upon relevant NCHRP, federal government, state, and private research and information (see Special Note B). The toolkit shall be a user-friendly, step-by-step resource capable of addressing current and future AI-based DSS solutions. At a minimum, the research shall consider technical, and organizational and business process considerations important to state DOTs.

Technical considerations are:

• Needs assessment (e.g., technical and infrastructure requirements);
• Selection processes and success metrics (e.g., accuracy, precision, generalization capability, explainability, interpretability, safety, and security);
• Data governance and management;
• Reasonableness of results and human collaboration and feedback (mechanisms to update AI-based DSS solutions based on human feedback); and
• Benefits and limitations.

Organizational and business process considerations are:

• Workflow and systems integration;
• Workforce implications;
• Security and privacy;
• Risk management;
• Agency policy and alignment with strategic goals;
• Coordination with federal and local agencies, and other stakeholders;
• Transparency and results communication (within and outside state DOTs) to build trust and support;
• Alignment with regulations;
• Funding allocation and long-term financial impact;
• Change management and leadership role;
• Ethics; and
• Procurement strategies, including independent verification and validation.

These considerations are not exhaustive, and individual items may apply to more than one consideration. 

Accomplishment of the project objective(s) will require at least the following tasks.

TASKS

The research plan will describe the appropriate deliverables that include, but are not limited to, the following (which may also represent key project milestones):

1. An annotated outline of the toolkit when approximately 20 percent of the project is complete. Work on the subsequent tasks will not begin until NCHRP approval of the annotated outline.

2. An interim report and panel meeting. The interim report should include the analyses and results of completed tasks, an example of a draft chapter of the toolkit, an update of the remaining tasks, and detailed outlines of the final research product(s).  A panel meeting will take place after the panel review of the interim report. The interim report and panel meeting should occur after the expenditure of no more than 40 percent of the project budget. Work on the subsequent tasks will not begin until NCHRP approval of the interim report.

3. Final research product(s) intended mainly for DOT leadership, discipline experts, and information technology managers, including:

• A toolkit;
• A conduct of research report that documents all research efforts, including:
  • An executive summary;
  • Strategies for long-term management of AI-based DDS solutions;
  • A list of research topics identified but not addressed in this project; and
  • Appendices for milestone deliverables (e.g., literature review, white papers).

4. A technical memorandum titled “Implementation of Research Findings and Products”.

5. A webinar organized and coordinated by the proposer on how to use the toolkit (the webinar’s recording and presentation with the presenter’s notes will be part of the final deliverables). NCHRP will approve the final list of webinar attendees.

STATUS: Proposals have been received in response to the RFP. The project panel will meet to select a contractor to perform the work.