Research Projects

Framework and Requirements for Using Artificial Intelligence to Improve Airport Runway SafetyArtificial Intelligence is

Artificial Intelligence is a powerful system that can automatically predict runway safety issues and produce optimal decisions to solve the problem in real time. The purpose of this project is to develop a framework for an AI system and provide guidelines for integrating an AI system to airports in order to improve runway safety.

Background (Describe the current situation or problem in the industry, and how your idea would address it.)

Airport safety is critical to the success of the aviation industry and reducing runway safety risk remains a top priority for the Federal Aviation Administration (FAA). In recent years, many efforts have been devoted to understanding how incidents and accidents happen at airport runways. A common conclusion is that most incidents or accidents are caused by human error. However, accident investigations indicate that human error is only the last link in a chain that leads to an incident or accident. There are many factors that could be related to runway incidents or accidents. Understanding the holistic picture of the runway incidents or accidents would help prevent or mitigate the risks. Furthermore, a lot of runway safety data has been collected, yet the big data has been adequately and effectively analyzed and used in predicting the runway safety risk and finding solutions. Finally, most safety efforts are reactive and lack of proactive actions and real time decisions.
The rapid growth of artificial intelligence (AI) applications in many industries indicates that AI could be useful to effectively reduce runway safety issues. AI can be defined as a system that simulates human intelligence for either solving a problem or making a decision (Chouwdhury & Sadek, 2012). AI can learn, predict, improve, and solve. Through training, an AI application is able to generate optimal solutions and apply them to novel situations not encountered before. Faced with the challenges of growing air traffic, resource demands, increasing uncertainties, and operational complexity, AI can automatically predict runway safety issues and provide real time optimal solutions and decisions under uncertainties, hence reducing the risk of runway incidents and accidents. However, AI is not without limitations and challenges. Since AI is often regarded as a black box, there are concerns about AI's capabilities to generalize to new situations or how AI determines the best decisions, hence some skepticism on AI's capabilities. It is also a challenge to integrate AI to current airport systems. Furthermore, AI could also be a potential liability given its autonomous nature.

Objective (What is the desired product or result that will help the airport industry?)

This project has four major objectives as follows
• Developing a framework for an AI system in runway safety
• Examining comprehensive requirements for developing and integrating the AI system at airports
• Evaluating challenges and risks associated with an AI system at airports
• Developing risk management guidelines for the AI system integration

Approach (Describe in general terms the steps you think are needed to achieve the objective.)

The following tasks are recommended for this project

  1. Conduct a comprehensive review of the literature on AI system, AI components, framework, machine learning, neural networks, applications, big data, cloud computing, and search and optimization
  2. Review airport safety systems and runway safety issues
  3. Describe runway safety datasets and their details
  4. Survey airport safety investigators and runway safety experts
  5. Develop a framework for an AI system in runway safety
  6. Describe and evaluate requirements for integrating an AI system to airport safety
  7. Examine risks and challenges associated with the AI integration at airports
  8. Develop risk management guidelines for the AI system integration

In addition, periodic teleconferences with the project panel will be conducted. The interim report will be submitted to receive feedback from the panel. Finally, the final report will be submitted by the deadline.

Cost Estimate and Backup (Provide a cost estimate and support for how you arrived at the estimate.)

This project is estimated to require $500,000.00, which includes travel, data collection, software, and expenses.
The project is estimated to require 18 months to complete.

Related Research - List related ACRP and other industry research; describe gaps (see link to Research Roadmaps above), and describe how your idea would address these gaps. This is a critical element of a synthesis topic submission.

Research gap:

While the role of AI in transportation in general, and air transportation, in general, has been well received in the related research, the application of AI in airport runway safety is missing. In addition, given the complexity and challenges of integrating an AI system to airport safety systems, comprehensive guidelines on the framework, requirements, and challenges of an AI system in airport safety is critical to the success of AI application. Existing research projects do not provide such guidelines. This research project will fill the gaps in the literature by developing a framework for an AI system and providing guidelines for integrating an AI system to airports in order to improve runway safety. The associated risks with the establishment of that system will be evaluated and the risk management strategies will be proposed to handle the risks. While the role of AI in transportation in general, and air transportation, in general, has been well received in the related research, the application of AI in airport runway safety is missing. In addition, given the complexity and challenges of integrating an AI system to airport safety systems, comprehensive guidelines on the framework, requirements, and challenges of an AI system in airport safety is critical to the success of AI application. Existing research projects do not provide such guidelines. This research project will fill the gaps in the literature by developing a framework for an AI system and providing guidelines for integrating an AI system to airports in order to improve runway safety. The associated risks with the establishment of that system will be evaluated and the risk management strategies will be proposed to handle the risks.

References:

  1. Transportation Research Circular E-C168 (2012). Artificial Intelligence Applications to Critical
  2. Transportation Issues. Transportation Research Board of the National Academies, November 2012.
  3. Transportation Research Circular E-C113 (2007). Artificial Intelligence in Transportation: Information for Application. Transportation Research Board of the National Academies, January 2007.
  4. US DOT Volpe Center (2019). Transportation in the Age of Artificial Intelligence and Predictive Analytics. US DOT Volpe Center Thought Leadership Series, Final Report, January 2019.
  5. US Department of Transportation (2018). Exploratory Advanced Research Program: The Role of Artificial Intelligence and Machine Learning in Federally Supported Surface Transportation Initiatives. US Department of Transportation.
  6. TRB's Airport Cooperative Research Program (ACRP) Report 3 (2008). Analysis of Aircraft Overruns and Undershoots for Runway Safety Areas. Transportation Research Board of the National Academies.
  7. TRB's Airport Cooperative Research Program (ACRP) Report 50 (2011). Improved Models for Risk Assessment of Runway Safety Areas. Transportation Research Board of the National Academies.
  8. ICAO (2018). Potential of Artificial Intelligence (AI) In Air Traffic Management (ATM), Thirteenth Air Navigation Conference, International Civil Aviation Organization (ICAO), October 19. Montreal, Canada.
  9. FAA Order 5200.11, FAA Airports (ARP) Safety Management System, August 2010.
Annotations

Voting

0 votes
1 up votes
1 down votes
Idea No. 200