Moving On, Even When You're Broken: Fail-Active Trajectory Generation via Diffusion Policies Conditioned on Embodiment and Task

TL;DR

This paper introduces DEFT, a diffusion-based trajectory generator that enables robots to perform tasks safely despite actuation failures, demonstrating high success rates and generalization in simulation and real-world scenarios.

Contribution

DEFT is a novel diffusion-based method that generalizes across failure types and supports constrained and unconstrained motions for fail-active robot operation.

Findings

Achieves 99.5% success rate in unconstrained motions in simulation.

Outperforms baselines like RRT and differential IK in failure conditions.

Successfully deployed on real robots for multi-step tasks.

Abstract

Robot failure is detrimental and disruptive, often requiring human intervention to recover. Our vision is 'fail-active' operation, allowing robots to safely complete their tasks even when damaged. Focusing on 'actuation failures', we introduce DEFT, a diffusion-based trajectory generator conditioned on the robot's current embodiment and task constraints. DEFT generalizes across failure types, supports constrained and unconstrained motions, and enables task completion under arbitrary failure. We evaluate DEFT in both simulation and real-world scenarios using a 7-DoF robotic arm. DEFT outperforms its baselines over thousands of failure conditions, achieving a 99.5% success rate for unconstrained motions versus RRT's 42.4%, and 46.4% for constrained motions versus differential IK's 30.9%. Furthermore, DEFT demonstrates robust zero-shot generalization by maintaining performance on failure conditions unseen during training. Finally, we perform real-world evaluations on two multi-step tasks, drawer manipulation and whiteboard erasing. These experiments demonstrate DEFT succeeding on tasks where classical methods fail. Our results show that DEFT achieves fail-active manipulation across arbitrary failure configurations and real-world deployments.