Strategizing at Speed: A Learned Model Predictive Game for Multi-Agent Drone Racing

2026-02-06

AI Summary

This work addresses the challenge of balancing strategic depth and real-time performance in high-speed multi-agent drone racing under dynamic opponent behaviors and strict time constraints. The authors propose a Learnable Model Predictive Game (LMPG) framework, which introduces learning-based mechanisms into the traditional Model Predictive Game (MPG) paradigm for the first time. By amortizing strategic reasoning through learned components, LMPG substantially reduces decision latency while preserving strong interactive awareness. The approach synergistically integrates model predictive control, trajectory optimization, and multi-agent game-theoretic reasoning to achieve both computational efficiency and strategic sophistication. Extensive simulations and real-world hardware experiments demonstrate that LMPG significantly outperforms conventional MPC and MPG methods in one-on-one high-speed racing scenarios.

Technology Category

• Multiagent Systems: Multiagent Learning
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions

Abstract

Autonomous drone racing pushes the boundaries of high-speed motion planning and multi-agent strategic decision-making. Success in this domain requires drones not only to navigate at their limits but also to anticipate and counteract competitors'actions. In this paper, we study a fundamental question that arises in this domain: how deeply should an agent strategize before taking an action? To this end, we compare two planning paradigms: the Model Predictive Game (MPG), which finds interaction-aware strategies at the expense of longer computation times, and contouring Model Predictive Control (MPC), which computes strategies rapidly but does not reason about interactions. We perform extensive experiments to study this trade-off, revealing that MPG outperforms MPC at moderate velocities but loses its advantage at higher speeds due to latency. To address this shortcoming, we propose a Learned Model Predictive Game (LMPG) approach that amortizes model predictive gameplay to reduce latency. In both simulation and hardware experiments, we benchmark our approach against MPG and MPC in head-to-head races, finding that LMPG outperforms both baselines.

Problem

• multi-agent drone racing
• strategic decision-making
• model predictive game
• computation latency
• high-speed motion planning

Innovation

• Learned Model Predictive Game
• Multi-Agent Drone Racing
• Interaction-Aware Planning
• Latency Reduction
• Model Predictive Control