Code-as-Monitor: Constraint-aware Visual Programming for Reactive and Proactive Robotic Failure Detection

2024-12-05

arXiv.org

AI Summary

To address open-set fault detection in closed-loop robotic systems, this work formulates both reactive anomaly identification and proactive fault prevention as a spatiotemporal constraint satisfaction problem. We introduce the novel “code-as-monitor” paradigm, wherein vision-language models (VLMs) generate executable monitoring code; further, we propose constraint geometric primitives to enable constraint-aware visual programming and real-time execution. Our method integrates VLMs, declarative constraint modeling, geometric abstraction, and a lightweight visual monitoring framework. Evaluated across three simulated environments and a physical robot platform, it achieves a 28.7% improvement in fault detection success rate and reduces average inference latency by 31.8% under strong disturbances. The approach significantly enhances the robustness of long-horizon closed-loop control grounded in open-loop policies.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Language and Vision
• Natural Language Processing: Safety and Robustness

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Automatic detection and prevention of open-set failures are crucial in closed-loop robotic systems. Recent studies often struggle to simultaneously identify unexpected failures reactively after they occur and prevent foreseeable ones proactively. To this end, we propose Code-as-Monitor (CaM), a novel paradigm leveraging the vision-language model (VLM) for both open-set reactive and proactive failure detection. The core of our method is to formulate both tasks as a unified set of spatio-temporal constraint satisfaction problems and use VLM-generated code to evaluate them for real-time monitoring. To enhance the accuracy and efficiency of monitoring, we further introduce constraint elements that abstract constraint-related entities or their parts into compact geometric elements. This approach offers greater generality, simplifies tracking, and facilitates constraint-aware visual programming by leveraging these elements as visual prompts. Experiments show that CaM achieves a 28.7% higher success rate and reduces execution time by 31.8% under severe disturbances compared to baselines across three simulators and a real-world setting. Moreover, CaM can be integrated with open-loop control policies to form closed-loop systems, enabling long-horizon tasks in cluttered scenes with dynamic environments.

Problem

• Detect and prevent robotic failures reactively and proactively
• Unify spatio-temporal constraint satisfaction for real-time monitoring
• Enhance monitoring accuracy with constraint-aware visual programming

Innovation

• VLM-generated code for real-time monitoring
• Constraint elements for accuracy and efficiency
• Closed-loop systems for dynamic environments