NCHRP 23-16 [Active]
Implementing and Leveraging Machine Learning at State Departments of Transportation
| Project Data
||Old Dominion University|
With rapid advances in computing, data science, and big data, there has been a corresponding massive expansion of datasets that are acquired, recorded, and stored, and thus available for individuals and organizations to turn into actionable information. Traditional statistical methods, data mining, simulation techniques (system dynamics and agent-based modeling), and machine learning (ML) are all being used to process these datasets and produce information for decision-making. For the purposes of this RFP, ML is defined as a branch of artificial intelligence (AI) focused on building applications that learn and extract knowledge from data. Categories of ML include supervised learning, unsupervised learning, reinforcement learning, natural language processing, computer vision, and deep learning. Transportation agencies that wish to leverage ML tools and techniques to extract information for decision-making and system operation will need to understand and implement these tools as part of their business processes.
Deployment of ML tools and techniques—whether acquired or developed in-house—by state departments of transportation (DOTs) is somewhat limited. Research is therefore needed to help state DOTs (1) understand and leverage ML to extract information from data more eﬃciently, eﬀectively, and in a timely manner; (2) identify business needs and challenges that lend themselves to eﬀective application of ML techniques; (3) identify ML skills, training, and infrastructure needed to add value to the management of transportation systems and assets; and (4) share ML tools, models, and case studies.
The objective of this research is to advance the understanding and use of ML tools and techniques at state DOTs and other transportation agencies. The proposed research will aid state DOTs in transitioning to a more advanced state of practice by:
- Demonstrating the feasibility and practical value of ML in the context of transportation systems, to better understand its application opportunities, implementation processes, and data requirements.
- Identifying skills, capabilities, resource, and organizational capacities necessary to leverage ML.
- Identifying and learning from existing applications at transportation agencies.
- Providing insight into costs, benefits, and performance and limitations considerations.
- Identifying and sharing ML frameworks, tools, guidance, and ML code for common use cases.
NCHRP has identified a contractor to undertake the research activities required to meet the objective of this research project, that includes:
- A review on the state of art and state of practice of ML in the context of state DOTs and other transportation agencies.
- Case studies applicable to state DOTs that are ready for near term deployment, for a variety of transportation agency data environments – including examples where ML algorithms and code are open (shareable) as well as those using commercial and proprietary technology.
- A guidebook on how to select and implement appropriate ML techniques. The guidebook should help state DOTs identify promising applications, assess costs, benefits, risks and limitations, and develop a roadmap for an agency ML program that includes implementation best practices.
- The guidance should outline the basic steps of data processing, model selection, training/prediction, and interpretation of results.
- A compilation of illustrative available code examples, with appropriate citations, for major use cases of ML.
- A compelling presentation and a briefing document for senior executives and decision makers that present a rationale for adoption of ML techniques at state DOTs.
- A final report that documents the entire research effort and presents a clear plan for further development and deployment of the guidebook. The report should be accompanied by appropriate presentation materials and an implementation plan that identifies opportunities for dissemination and moving research into practice.