A Guide to Using Artificial Intelligence to Enhance Transportation Asset-Management Plans

State departments of transportation are required to develop transportation asset-management plans (TAMPs). These data-driven plans are based on asset inventories, condition, and available resources and provide a plan for investing in and managing assets over 10 years. Advances in artificial intelligence (AI) provide opportunities to develop and enhance these plans. However, selecting appropriate tools and developing verifiable, transparent applications is challenging. This paper presents a guide for state departments of transportation to understand the concepts and to use these tools to enhance their TAMPs. The paper aims to raise awareness of opportunities and potential issues and help asset managers develop the vocabulary needed to interact with AI experts. An overview of machine learning covers the different paradigms and tools. Building on the review of AI methods, the paper presents opportunities for using AI within the TAMP building blocks (asset inventory, condition assessment, deterioration modeling, and decision making), as well as two cross-cutting functions: data quality and integration and communication. The paper then presents how these building blocks are used to develop the required sections of the plan (gap analysis, life-cycle planning, risk analysis and risk management, financial plan, and investment strategies) and consistency determination. Example applications are included. Finally, implementation guidance is provided.