Raising Awareness of Artificial Intelligence for Transportation Systems Management and Operations

Executive Summary

This report introduces the fundamentals of artificial intelligence (AI) technologies and the potential applications in transportation management system (TMS) and transportation management center (TMC) operations, showcases successful AI applications for TMSs, and provides a list of important issues to consider in developing AI applications. This information is intended to raise the awareness of transportation agencies of the potential benefits, implications, and impacts of using AI for a TMS, TMC operations, or a transportation systems management and operations (TSMO) program.

AI and machine learning are elements of business intelligence (BI) strategies and technologies, which are used by enterprises for data analysis and information extraction. Traditional problems, functions, or actions that AI techniques can address include reasoning, knowledge representation, planning, learning, natural language processing (and understanding), perception, and the ability to move and manipulate objects.¹ In each problem area, AI technologies are proving to have significant performance benefits versus other traditional mathematical modeling approaches. For example, the capabilities of Alexa and Google Assistant to understand human speech significantly outperforms the interactive voice response (IVR) technologies used in 511 systems over the past 20 years.

Colloquially, "artificial intelligence" typically describes the ability of a machine to mimic human actions or cognitive functions, such as problem solving or maintaining a conversation. This type of artificial intelligence is typically referred to as "strong" AI. There are no strong AI systems in existence. Other specialized applications of AI are termed "weak" AI or machine learning applications. Machine learning applications offer the potential to supplant human work in a variety of TSMO areas, including traffic imagery analysis, incident detection, traffic control and traffic signal timing, TMC function automation, and data analysis.

Chatbots and question-answering (QA) systems may enable new ways to obtain insights in data. Neural networks can analyze imagery from a variety of sources for incident detection, incident management, and traffic data collection. Fuzzy logic is already used by a variety of departments of transportation (DOT) for ramp metering and fuzzy logic may find additional applications by simplifying if…then rule bases for decision-support systems. Unsupervised AI systems may learn new ways to control traffic and coordinate integrated corridor management actions across a variety of control and advisory technologies. Driverless vehicles and airborne and ground-based unmanned aerial systems (UAS) may improve TSMO staff safety and productivity. Additional applications may arise as agencies gain experience with AI tools and technologies.

There are thousands of companies competing for dollars in the AI space across essentially every consumer and Government market as the technology continues to mature. Technologies from Google, Amazon, Microsoft, and Facebook and open-source tools they have either developed or adopted tend to underpin most software and hardware AI products. As with Big Data a few years ago, hype in the capabilities of AI is also at a peak. As time moves on, these technologies are likely to come closer and closer to "plug and play," but currently there is still a reasonably large barrier between the dreams of AI-enabled TSMO applications and the need for significant expertise and investment to make those dreams a reality. As fast as the pace of development of AI tools and technologies is progressing, AI applications should find their way from research experiments and pilot demonstrations to fully scalable applications in the near term.

Common trends in AI development over the next five years, according to Forbes, may be:²

• Development of AI-specific hardware chips for embedding machine learning and training in consumer products, industrial processes, and vehicles.
• Movement of machine learning models from centralized Cloud systems to edge Internet of Things (IoT) devices.
• Interoperability among neural network modeling systems and frameworks via Open Neural Network Exchange (ONNX).³
• Automated machine learning with AutoML—speeding the process of building and deploying neural networks.
• Application of AI analysis to information technology operations.
• Continued evolution of chatbots and virtual assistants into more comprehensive, context-sensitive question and answer functions.
• Deployment of consumer-ready automated vehicle services.
• Democratization of machine learning services and software to professionals without deep software development and database management skills.
• Improvement of AI responsibility, transparency, and morality; the removal of systematic biases against minorities.

In the field of digital video processing, since existing products have already emerged for TSMO, the pace may be faster. Driverless vehicles are likely to be available to TSMO agencies in the near term and beyond visual line of sight (BVLOS)-automated unmanned aerial systems (UAS) operations in the medium term.

A variety of AI technologies have been deployed for TSMO applications. Several State and local DOTs (Nevada, Florida, and Iowa) have begun deploying neural network technologies for incident detection using video image analysis and traffic prediction. Fuzzy logic has been used by Washington State DOT for more than 20 years and California DOT (Caltrans) has begun deployment of fuzzy logic metering in a pilot corridor. Delaware DOT has piloted several AI applications for traffic congestion and incident prediction. The Metropolitan Transportation Commission of the Bay Area and several other agencies have light integration of 511 with Alexa. Several arterial management agencies are piloting use of Google Assistant. More than 20 State DOTs have active UAS programs that may be enhanced with AI in the near to medium term. Several DOTs are piloting use of automated vehicles for crash abatement. As AI technology continues to mature, the applications for AI in TSMO are likely to continue to expand.

Determining how to start in AI applications for TSMO will be unique to your organization. As is true with any TSMO activity, the basis for improvement of any activity with AI has three basic components:

1. A supporting institutional framework, policies, and appetite.
2. Processes, staff, and technology that support the program.
3. The implementation of the system itself.

The foundation of any successful program is first the institutional framework to support the activity. In the context of AI applications, developing the necessary organizational structure and functions for TSMO is an important element. After these enabling actions, the business processes for using AI technologies for TSMO practices should follow more readily and be more effective due to a strong foundation in business processes. These processes should enable the AI programs to function at a high level initially and continue to adapt and improve as AI technology advances.

Developing these foundational elements is important and answering the questions in chapter 5 will help to identify where the strengths and weaknesses lie. It is important to keep in mind in any technology deployment that some of the dimensions are inherently more difficult to deal with than others, yet they all should be addressed to move forward. Failing to consider issues related to staffing and organization, for example, may result in your AI project being a pilot that is never integrated into the main TSMO operation.

A holistic program plan may be developed considering many potential applications and pared back to consider what might be accomplished with more realistic budgets and resources. Agencies in the early planning stages may take the following steps:

1. Convene an interdepartmental workshop to educate stakeholders, partners, and potential partners on AI and brainstorm potential applications and synergies.
2. Discuss priorities, opportunities, and barriers to AI applications in each of the TSMO areas.
3. Determine a short list of high-priority applications and a longer list of secondary-priority functions that address regional issues, challenges, and goals. While many goals are generic, tailoring the AI strategy to regional hot-button issues is typically helpful in gaining broader buy-in from decisionmakers and associated departments of State DOTs and local infrastructure owner-operators (IOO).
4. Review the list of general and detailed questions in this chapter and consider the responses of your organization to each.
5. Develop a project plan to implement the actions.