Flight Validation of Learning-Based Trajectory Optimization for the Astrobee Free-Flyer

2025-05-08

AI Summary

To address the high computational cost and poor real-time performance of trajectory optimization in space applications, this work presents the first learning-enhanced real-time trajectory optimization framework deployed on the International Space Station (ISS), significantly enhancing the autonomous navigation capability of the Astrobee free-flying robot. We propose a novel hybrid framework integrating neural-network-based warm-starting with the GuSTO sequential convex programming algorithm: a parameter-to-trajectory mapping network is trained offline to generate high-quality initial guesses, while an embedded GuSTO solver operates online to ensure convergence and real-time feasibility. The method reduces onboard optimization latency by orders of magnitude, achieving millisecond-level response under stringent ISS resource constraints. It successfully enabled the world’s first in-orbit learning-based closed-loop control flight demonstration on the ISS. This work establishes a verifiable, real-time intelligent optimization paradigm for spacecraft autonomous navigation.

Technology Category

• Intelligent Robots: Learning & Optimization for ROB
• Search and Optimization: Learning to Search
• Constraint Satisfaction and Optimization: Constraint Optimization

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Although widely used in commercial and industrial robotics, trajectory optimization has seen limited use in space applications due to its high computational demands. In this work, we present flight results from experiments with the Astrobee free-flying robot on board the International Space Station (ISS), that demonstrate how machine learning can accelerate on-board trajectory optimization while preserving theoretical solver guarantees. To the best of the authors' knowledge, this is the first-ever demonstration of learning-based control on the ISS. Our approach leverages the GuSTO sequential convex programming framework and uses a neural network, trained offline, to map problem parameters to effective initial ``warm-start'' trajectories, paving the way for faster real-time optimization on resource-constrained space platforms.

Problem

• Accelerating on-board trajectory optimization for space applications
• Demonstrating learning-based control on the International Space Station
• Using neural networks for real-time optimization in constrained platforms

Innovation

• Machine learning accelerates on-board trajectory optimization
• Neural network provides warm-start trajectories offline
• GuSTO framework ensures theoretical solver guarantees