EgoRecovery: Acquiring Failure Recovery Ability Through Human Recovery Demonstration

2026-07-22

AI Summary

This work addresses the challenge of enabling robots to autonomously recover from failure states in real-world environments, where teleoperated data collection for recovery is costly and poorly scalable. The authors propose EgoRecovery, a novel co-training framework that efficiently leverages first-person human demonstrations of failure recovery by constructing a shared corrective intent space between humans and robots. By integrating a small amount of robot-collected recovery data, the method enables on-demand activation of recovery policies through a gating mechanism that dynamically predicts when to trigger corrective intent and facilitates policy transfer. Evaluated on real-world tasks, EgoRecovery significantly outperforms baselines—including purely robot-learned recovery, direct human-robot co-training, and intent transfer—achieving substantially higher recovery success rates.

Abstract

Robust embodied robots should be able to recover from failures and retry tasks in order to operate reliably in unstructured and noisy real-world environments. Achieving this capability requires training policies on data that captures recovery behaviors. However, collecting such data through robot teleoperation is difficult to scale, as it is time-consuming to induce diverse failure states, perform corrective actions, and reset the environment. This challenge is further exacerbated by the high diversity of failure modes, which demands substantially more recovery data than success demonstrations. In this work, we show that egocentric human data capturing failure recovery processes provides a scalable alternative. By efficiently arranging task-level failure configurations and recording short recovery segments, human operators can generate more than 10x as much valid recovery data per hour compared to robot teleoperation under our protocol. To address the embodiment gap between human and robot, we propose EgoRecovery, a co-training framework for learning recovery behavior, where human recovery demonstrations are aligned to a compact corrective-intent space shared with robot data, which captures the timing and magnitude of correction. Only a small number of robot recovery demonstrations are required to connect this intent to executable robot actions. At deployment, a learned recovery gate predicts when correction is needed from robot observations and activates the corrective intent only in recovery states. Experiments on real-world recovery tasks show that EgoRecovery improves success from failure starts over robot-only recovery, direct co-training with human recovery data, and direct intent-transfer baselines.

Problem

• failure recovery
• embodied AI
• human demonstration
• recovery data
• real-world robotics

Innovation

• EgoRecovery
• failure recovery
• human demonstration
• corrective intent
• embodied AI