Airlift Challenge: A Competition for Optimizing Cargo Delivery

2024-04-26

arXiv.org

AI Summary

Real-time rescheduling of air cargo—particularly military/high-value and time-sensitive shipments—under sudden disruptions (e.g., weather, mechanical failures) remains a critical challenge. Method: This paper introduces a scalable OpenAI Gym-based simulation environment and pioneers the application of general-purpose temporal planning to air cargo scheduling. It proposes a temporal PDDL modeling framework tailored for pickup-and-delivery problems and designs a progressive difficulty evaluation mechanism. Furthermore, it integrates reinforcement learning with symbolic planning to enable hybrid decision-making in dynamic environments. Contribution/Results: The framework served as the official platform for an international algorithm competition (November 2023–April 2024). Empirical results demonstrate that the proposed approach significantly improves agent robustness and timeliness across diverse disruption scenarios. It establishes a verifiable, scalable new paradigm for dynamic air cargo optimization.

Technology Category

• Planning, Routing, and Scheduling: Temporal Planning
• Search and Optimization: Sampling/Simulation-based Search
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Search and Retrieval-Augmented AI: Agentic search

Abstract

Airlift operations require the timely distribution of various cargo, much of which is time sensitive and valuable. These operations, however, have to contend with sudden disruptions from weather and malfunctions, requiring immediate rescheduling. The Airlift Challenge competition seeks possible solutions via a simulator that provides a simplified abstraction of the airlift problem. The simulator uses an OpenAI gym interface that allows participants to create an algorithm for planning agent actions. The algorithm is scored using a remote evaluator against scenarios of ever-increasing difficulty. The second iteration of the competition was underway from November 2023 to April 2024. This paper describes the competition, simulation environment, and results. As a step towards applying generalized planning techniques to the problem, a temporal PDDL domain is presented for the Pickup and Delivery Problem, a model which lies at the core of the Airlift Challenge.

Problem

• Optimize cargo delivery in airlift operations under disruptions
• Develop algorithms for rescheduling during weather and malfunctions
• Apply generalized planning to solve Pickup and Delivery Problem

Innovation

• Simulator with OpenAI gym interface
• Algorithm scored by remote evaluator
• Temporal PDDL domain for planning