Autonomous Recovery: Agentic AI for End-to-End Flight Disruption Rebooking at Consumer Scale
DOI:
https://doi.org/10.15662/IJEETR.2025.0706052Keywords:
Agentic Artificial Intelligence, Autonomous Flight Rebooking, Flight Disruption Management, Multi-Agent Orchestration, Real-Time Itinerary Optimization, Human-in-the-Loop Governance, Passenger Experience OptimizationAbstract
Flight disruptions, caused by weather, air traffic control constraints, crew scheduling limits, and mechanical events, affect a substantial share of global air travelers every year and generate billions of dollars in direct recovery costs plus a larger multiple in lost customer goodwill. Traditional rebooking relies on a mix of rule based systems and call center labor that scales poorly during irregular operations, when thousands of passengers require individualized itinerary changes within a compressed time window. This paper presents an architecture and evaluation framework for agentic artificial intelligence systems that plan, negotiate, and execute end to end rebooking on behalf of consumers at scale. We describe a layered platform consisting of a consumer channel layer, an agentic orchestration layer with planning and memory components, a tool integration layer that interfaces with global distribution systems and passenger service systems, and a real time data and signal fabric. We introduce a four tier autonomy model ranging from agent assisted suggestions to fully autonomous rebooking, and we propose evaluation metrics spanning auto resolve rate, time to rebook, cost avoidance, fairness, and customer satisfaction. Using a simulated multi region dataset modeled on publicly reported industry benchmarks, we show that a well governed agentic system can resolve the majority of disruption events without human intervention while reducing median time to rebook from tens of minutes to under a minute, and we discuss the guardrails, escalation tiers, and regulatory considerations required to deploy such systems responsibly at consumer scale.
References
1. Alvarez, M. and Petrov, D. (2023). Operations research approaches to irregular airline operations recovery. Transportation Research Part A: Policy and Practice, 171, 103640.
2. Bhatt, R., Lindqvist, S., and Omondi, K. (2024). Passenger centric fairness in automated flight rebooking systems. Journal of Air Transport Management, 115, 102520.
3. Chen, Y. and Fairweather, J. (2022). Multi agent coordination for large scale service recovery. Proceedings of the International Conference on Autonomous Agents and Multiagent Systems, 44 to 52.
4. Delacroix, P. and Nakamura, H. (2024). Tool augmented language model agents for enterprise workflow automation. arXiv preprint arXiv:2403.08841.
5. Esposito, L. (2023). Trust calibration in algorithmic customer service: a field study. ACM Transactions on Computer Human Interaction, 30(4), 1 to 29.
6. Government Accountability Office (2023). Airline passenger protections: compensation and rebooking practices during widespread disruptions. GAO 23 105412.
7. Hassan, A., Oyelaran, F., and Kowalski, T. (2024). Confidence calibrated decision gating in autonomous service agents. Proceedings of the ACM Conference on Fairness, Accountability, and Transparency, 212 to 224.
8. International Air Transport Association (2024). Annual review of global air traffic punctuality and disruption trends.
9. Jørgensen, L. and Villanueva, C. (2023). Interline data exchange standards and their effect on recovery speed. Journal of Revenue and Pricing Management, 22(5), 401 to 414.
10. Kapoor, S. and Reyes, M. (2025). Evaluating large language model planners on long horizon travel itinerary tasks. Proceedings of the Annual Meeting of the Association for Computational Linguistics, 3310 to 3324.
11. Lindholm, E. (2022). Call center capacity planning under irregular operations: a queuing theory approach. Transportation Science, 56(6), 1489 to 1503.
12. McKinsey and Company (2024). The future of travel recovery: automation and the consumer experience.
13. Nowak, B. and Suresh, V. (2024). Memory augmented agents for persistent, multi session consumer support. Proceedings of the AAAI Conference on Artificial Intelligence, 38(10), 11204 to 11212.
14. Osei, D. and Marchetti, F. (2023). Regulatory divergence in passenger compensation regimes: a comparative study. Air and Space Law, 48(3), 221 to 243.
15. Patel, N., Lindberg, A., and Choudhury, R. (2025). Guardrails for autonomous decision systems in regulated consumer markets. Proceedings of the International Conference on Machine Learning, Workshop on Trustworthy AI.
16. Santos, I. and Weber, G. (2023). Measuring fairness in automated resource allocation during service disruptions. Management Science, 69(9), 5120 to 5138.
17. U.S. Department of Transportation (2024). Air travel consumer report: delay, cancellation, and complaint statistics.
18. Yamamoto, K. and Oduya, P. (2025). Evaluating multi agent orchestration patterns for high throughput consumer service automation. Proceedings of the International Conference on Autonomous Agents and Multiagent Systems, 1102 to 1114.
