Performance-guided Task-specific Optimization for Multirotor Design

2025-10-06

AI Summary

Designing task-specific multirotor micro-air-vehicles (MAVs) remains challenging due to rigid, fixed-configuration paradigms that neglect mission-critical performance and manufacturability constraints. Method: We propose a performance-driven, end-to-end optimization framework that jointly optimizes airframe geometry and control policy toward closed-loop task performance (e.g., agile waypoint navigation), while explicitly modeling aerodynamic interference and manufacturability. The framework integrates reinforcement learning for controller co-optimization, Bayesian optimization and CMA-ES for efficient exploration of motor placement design space, and high-fidelity closed-loop control simulation with physics-based aerodynamic modeling. Results: The optimized configuration significantly outperforms classical quadrotor/hexrotor baselines and state-of-the-art fully-actuated designs in both simulation and real-world flight tests. Crucially, it demonstrates strong sim-to-real transfer capability. This work establishes a reproducible, scalable paradigm for task-oriented physical design of MAVs.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Intelligent Robots: Learning & Optimization for ROB
• Humans and AI: Planning and Decision Support for Human-Machine Teams

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

This paper introduces a methodology for task-specific design optimization of multirotor Micro Aerial Vehicles. By leveraging reinforcement learning, Bayesian optimization, and covariance matrix adaptation evolution strategy, we optimize aerial robot designs guided exclusively by their closed-loop performance in a considered task. Our approach systematically explores the design space of motor pose configurations while ensuring manufacturability constraints and minimal aerodynamic interference. Results demonstrate that optimized designs achieve superior performance compared to conventional multirotor configurations in agile waypoint navigation tasks, including against fully actuated designs from the literature. We build and test one of the optimized designs in the real world to validate the sim2real transferability of our approach.

Problem

• Optimizing multirotor designs for specific task performance
• Exploring motor configurations under manufacturability and aerodynamic constraints
• Validating optimized designs in real-world agile navigation tasks

Innovation

• Using reinforcement learning for robot design optimization
• Optimizing motor poses under manufacturability constraints
• Validating sim2real transfer with physical prototype testing