Over the past two decades, machine learning—the leading branch of artificial intelligence—has risen rapidly on the strength of exploding data volumes, cheaper storage and computation, and continual algorithmic breakthroughs. Deep-learning techniques and generative AI systems such as ChatGPT are now transforming day-to-day business practices. Yet state and local transportation agencies are drowning in new data streams faster than they can translate them into actionable insight. TRB’s NCHRP Research Report 1122, *Implementing Machine Learning at State Departments of Transportation: A Guide*, equips DOTs with a primer on promising ML use cases, tools for weighing costs, benefits, risks, and constraints, and a roadmap for fostering the data-driven culture needed to scale these capabilities across transportation programs.

This is a companion document to NCHRP Report 1122: Implementing Machine Learning at State Departments of Transportation.

Artificial intelligence (AI) has become a powerful force in transportation departments, particularly for managing and streamlining traffic flow. By using real-time data and predictive analytics, AI solutions can ease congestion, decrease travel times, and enhance safety through early hazard detection. In addition, AI-driven simulations offer a cost-effective method for evaluating and refining transportation networks, reducing the need for extensive physical testing.