ACRP Graduate Student Paper - Planning Maintenance and Rehabilitation Actions for Airport Pavements: A Combined Supervised Machine Learning and Reinforcement Learning Approach

Maintenance and Rehabilitation (M&R) of airport pavement assets involves considerable financial resources. As such, even modest improvement in M&R action planning could lead to non-trivial savings. The state-of-the-practice for planning M&R actions mostly relies on condition thresholds and prioritization rules, while the state-of-the-art often requires unduly assumptions, and the computational challenge can present an important issue when characterizing pavement conditions and M&C actions involves large dimensions. This study proposes a machine learning (ML) approach that integrates pavement condition prediction using supervised ML with M&R action planning empowered by reinforcement learning (RL). The Q-learning method is used to train the RL model. The use of the integrated model is demonstrated using real-world data from the Chicago O’Hare International Airport. The results show the effectiveness of the proposed approach and potential to reduce M&R cost compared to the existing practice.

Language

  • English

Media Info

  • Media Type: Web
  • Features: Figures; References; Tables;
  • Pagination: 17p

Subject/Index Terms

Filing Info

  • Accession Number: 01763563
  • Record Type: Publication
  • Report/Paper Numbers: TRBAM-21-04080
  • Files: TRIS, TRB, ATRI
  • Created Date: Dec 23 2020 11:06AM