1. Home
2. Ideas
3. FRAMEWORK AND TOOLS FOR USING ARTIFICIAL...

FRAMEWORK AND TOOLS FOR USING ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING METHODS TO IDENTIFY OPERATIONS STRATEGIES FOR PROACTIVE SAFETY

• Jun 10, 2025

Research Description

The availability of connected vehicle (CV) data, vehicle probe and crowdsource data, high-resolution traffic signal data and data from other emerging sources provide new opportunities for the general use of artificial intelligence (AI) and machine learning (ML) methods to analyze safety-related measures. Some transportation agencies have started to use AI and ML to identify statistical underlying patterns in CV data that can be used for descriptive analysis and to identify and rank road segments based on safety scores and to prioritize improvements accordingly. However, most of the recent work stops at extracting patterns from the data. There is a need to develop frameworks and tools that can be used to incorporate predictive AI/ML methods in testing what-if scenarios that incorporate changes in human behavior that are not currently captured by any other tool and select optimal safety strategies based on their predicted impacts on the transportation network.

Objective The objective of this research is to develop a framework and tools to implement predictive AI/ML methods to evaluate and recommend alternative operations intervention strategies for safety improvements. The tools should be able to evaluate future outcomes based on each considered strategy and take that into account when making the recommended strategy selection.

Potential benefits The research is expected to result in the development of a framework and a set of tools that incorporate descriptive and predictive AI/ML methods to identify operations intervention strategies that can improve highway safety for all users. This can lead to a reduction in the number and severity of crashes, in addition to improving mobility metrics and providing safe mobility.

Additional Supporting Information

AI and ML are increasingly shaping the transportation sector, particularly in autonomous vehicles, traffic management, and safety applications. AI has been recently used to address challenges related to the transition toward autonomous and mixed traffic environments, fostering cooperative driving between human-driven and autonomous vehicles. AI and ML are reshaping transportation by improving efficiency, safety, and sustainability. From reinforcement learning in adaptive traffic control to AI-driven pedestrian safety policies, these technologies offer transformative solutions to longstanding transportation challenges. However, there is a critical research gap in using predictive AI/ML methods that incorporate collective intelligence methods to select optimal safety-based control strategies based on predicted responses of human behavior in multimodal traffic operation.

The research will build off the foundation of several NCHRP projects, including NCHRP 17-100 Leveraging Artificial Intelligence and Big Data to Enhance Safety Analysis and others.

The proposed research aligns with the 2021-2026 AASHTO Strategic Plan by addressing: • Safety, Mobility, and Access for Everyone: By focusing on enhancing VRU safety at intersections through AI technologies, the research advances a safe, multimodal transportation system. • National Transportation Policy Leadership: Evaluating emerging trends in AI applications positions AASHTO and state DOTs as proactive leaders in transportation innovation. The learnings and methods can support greater safety integration in the Transportation Operations Manual (TOM).

This research would also support the Safe System Approach, which emphasizes proactive applications.

---

Submitted By

Tracy Scriba

FHWA

202-366-0855

Attachments