Intelligent departure metering assistant tool (IDMAT) for airside congestion management

• Main Authors: Ali, Hasnain, Pham, Duc-Thinh, Tran, Thanh-Nam, Alam, Sameer
• Other Authors: School of Mechanical and Aerospace Engineering
• Format: Conference or Workshop Item
• Language: English
• Published: 2023
• Subjects: Engineering::Computer science and engineering::Information systems::Information interfaces and presentation; Airport Departure Metering; Deep Reinforcement Learning; Intelligent Pushback Control; Decision Making Under Uncertainty
• Online Access: https://hdl.handle.net/10356/162187
• Institution: Nanyang Technological University

Airport Departure Metering (DM) is an effective approach to contain taxi delays by controlling departure pushback timings. In this work, we demonstrate the potential of Deep Reinforcement Learning (DRL) based DM method to reduce taxi delays by effectively transferring delays from taxiways to gates. This work casts the DM problem in a markov decision process framework to train a DM policy over simulations generated using historical airport surface movement data. We further develop an Intelligent Departure Metering Assistant Tool (IDMAT) that employs the trained DM policy to recommend pushback advisories to Air Traffic Controller (ATCO). Furthermore, to assist pushback approval decisions, IDMAT displays additional traffic information like potential downstream conflicts, runway queues, and delay evolution on the taxiway to ATCOs. We intend to perform validation experiments with ATCOs to evaluate the efficacy and acceptability of the recommended pushback advisories. Similar scenarios---with and without pushback advisories and additional traffic information---shall be presented to ATCOs to evaluate the ATCO performance while managing congestion. ATCO actions (advisory accept/reject) shall also be fed-back to train the DM policy into recommending ATCO-like actions.