ACORN: Adaptive Contrastive Optimization for Safe and Robust Fine-Grained Robotic Manipulation

2025-05-10

AI Summary

Conventional embodied intelligence research overemphasizes task success rate and cumulative reward, neglecting critical robustness and safety requirements for real-world deployment. Distribution shifts and unforeseen disturbances in realistic environments frequently induce catastrophic failures in fine manipulation tasks. Method: We propose the first safety-oriented, four-dimensional resilience evaluation framework tailored for manipulation tasks, quantifying policy safety under environmental perturbations. Additionally, we design a plug-and-play adaptive contrastive optimization framework integrating expert trajectory alignment, structured Gaussian noise injection, and dual-perturbation high-diversity negative sampling—enhancing robustness without compromising task performance. Contribution/Results: Experiments across multiple manipulation benchmarks demonstrate up to a 23% improvement in safety metrics, significantly mitigating failure induced by distributional shift while preserving efficacy.

Technology Category

• Intelligent Robots: Manipulation
• Natural Language Processing: Safety and Robustness
• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness

Application Category

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

Abstract

Embodied AI research has traditionally emphasized performance metrics such as success rate and cumulative reward, overlooking critical robustness and safety considerations that emerge during real-world deployment. In actual environments, agents continuously encounter unpredicted situations and distribution shifts, causing seemingly reliable policies to experience catastrophic failures, particularly in manipulation tasks. To address this gap, we introduce four novel safety-centric metrics that quantify an agent's resilience to environmental perturbations. Building on these metrics, we present Adaptive Contrastive Optimization for Robust Manipulation (ACORN), a plug-and-play algorithm that enhances policy robustness without sacrificing performance. ACORN leverages contrastive learning to simultaneously align trajectories with expert demonstrations while diverging from potentially unsafe behaviors. Our approach efficiently generates informative negative samples through structured Gaussian noise injection, employing a double perturbation technique that maintains sample diversity while minimizing computational overhead. Comprehensive experiments across diverse manipulation environments validate ACORN's effectiveness, yielding improvements of up to 23% in safety metrics under disturbance compared to baseline methods. These findings underscore ACORN's significant potential for enabling reliable deployment of embodied agents in safety-critical real-world applications.

Problem

• Addresses lack of robustness and safety in robotic manipulation tasks
• Introduces metrics to quantify resilience to environmental perturbations
• Proposes algorithm to enhance policy robustness without performance loss

Innovation

• Adaptive contrastive learning for robust manipulation
• Gaussian noise injection for diverse negative samples
• Double perturbation technique to minimize overhead