DreamFLEX: Learning Fault-Aware Quadrupedal Locomotion Controller for Anomaly Situation in Rough Terrains

2025-02-09

AI Summary

To address motion failure in quadrupedal robots operating on unstructured terrain—caused by hardware faults such as motor overheating and joint locking—this paper proposes a real-time, perception-aware, and adaptive robust gait control method. The core contribution is the first end-to-end neural network that jointly performs fault estimation and gait modulation: it explicitly models fault type and severity, and integrates deep reinforcement learning with neural state estimation to predict online fault vectors and remap control policies using joint torque and position observations alongside fault priors. Evaluated in simulation and on the A1 physical robot, the method significantly improves traversal success rates under faults, achieving an average 42% enhancement in stability over baseline controllers. It enables sustained locomotion across challenging terrains—including gravel, sloped surfaces, and uneven ground—demonstrating strong resilience to hardware degradation.

Technology Category

• Intelligent Robots: State Estimation
• Natural Language Processing: Safety and Robustness
• Planning, Routing, and Scheduling: Replanning and Plan Repair

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
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Recent advances in quadrupedal robots have demonstrated impressive agility and the ability to traverse diverse terrains. However, hardware issues, such as motor overheating or joint locking, may occur during long-distance walking or traversing through rough terrains leading to locomotion failures. Although several studies have proposed fault-tolerant control methods for quadrupedal robots, there are still challenges in traversing unstructured terrains. In this paper, we propose DreamFLEX, a robust fault-tolerant locomotion controller that enables a quadrupedal robot to traverse complex environments even under joint failure conditions. DreamFLEX integrates an explicit failure estimation and modulation network that jointly estimates the robot's joint fault vector and utilizes this information to adapt the locomotion pattern to faulty conditions in real-time, enabling quadrupedal robots to maintain stability and performance in rough terrains. Experimental results demonstrate that DreamFLEX outperforms existing methods in both simulation and real-world scenarios, effectively managing hardware failures while maintaining robust locomotion performance.

Problem

• Develop fault-tolerant controller for quadrupedal robots
• Handle joint failures in rough terrains
• Maintain stability under hardware anomalies

Innovation

• Fault-tolerant locomotion controller
• Real-time failure estimation
• Adapts to joint failures