Learning-Based Fault Detection for Legged Robots in Remote Dynamic Environments

Abriana Stewart-Height; Seema Jahagirdar; Nikolai Matni

cs.ROApr 3, 2026

Learning-Based Fault Detection for Legged Robots in Remote Dynamic Environments

Authors: Abriana Stewart-Height, Seema Jahagirdar, Nikolai Matni

Organizations: Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge MA 02139, USA · Department of Electrical and Systems Engineering, University of Pennsylvania, Philadelphia, PA, 19104, USA

Abstract

Operations in hazardous environments put humans, animals, and machines at high risk for physically damaging consequences. In contrast to humans and animals, quadruped robots cannot naturally identify and adjust their locomotion to a severely debilitated limb. The ability to detect limb damage and adjust movement to a new physical morphology is the difference between survival and death for humans and animals. The same can be said for quadruped robots autonomously carrying out remote assignments in dynamic, complex settings. This work presents the development and implementation of an off-line learning-based method to detect single limb faults from proprioceptive sensor data in a quadrupedal robot. The aim of the fault detection technique is to provide the correct output for the controller to select the appropriate tripedal gait to use given the robot's current physical morphology.

Figures & tables

Fig. 1 : Flowchart of Proposed Fault Recovery Framework

Condition	Number of Points	Size	Average Time per Trial
None	70	36 MB	$\sim$ 90s
LF	30	8 MB	$\sim$ 60s
RF	15	8 MB	$\sim$ 120s
LB	24	29 MB	$\sim$ 90s
RB	31	32 MB	$\sim$ 90s

TABLE I : Dataset Breakdown

Fig. 2 : Actual input values (blue) Vs. Reconstructed input values (red) for the intact quadrupedal robot pronking

Fig. 3 : Distribution of reconstruction errors for the intact quadrupedal robot when executing fore-aft pronking.

Explore similar work

Learning Fault-Tolerant Locomotion with Adaptive Gait Timing
Aug 7, 2026Giovanbattista Gravina, Luca Rossini, Carlo Rizzardo +2Legged Robotics Robot Locomotion
Humanoids and Human-Centered Mechatronics Research Line, Italian Institute of Technology, Genoa 16163, Italy
Mixture-of-Experts RL for Fault-Tolerant Legged Locomotion
Jun 24, 2026Giulio Turrisi, Ozan Pali, Luca Oneto +1Legged Robotics Mixture of Experts
Dynamic Legged Systems Laboratory, Istituto Italiano di Tecnologia (IIT), Genova, Italy. · Dipartimento di Informatica, Bioingegneria, Robotica e Ingegneria dei Sistemi (DIBRIS), Universita di Genova, Genova, Italy
SlipSense: Multimodal Sensing for Online Slip Detection in Legged Robots
Jun 23, 2026Iris Szu-Yao Liu, Chien Chern Cheah, Meng Yee Michael ChuahTactile Sensing Robotic Tactile Sensing
School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore · Institute for Infocomm Research, Agency for Science Technology and Research, Singapore

Aug 7, 2026J/K move · Enter open · S save

Learning Fault-Tolerant Locomotion with Adaptive Gait Timing

Giovanbattista Gravina, Luca Rossini, Carlo Rizzardo, Arturo Laurenzi+1

Humanoids and Human-Centered Mechatronics Research Line, Italian Institute of Technology, Genoa 16163, Italy

Legged Robotics Robot Locomotion RL for Legged Locomotion Quadruped Locomotion

Open paper AskPDF arXiv

Hardware failures require legged robots to rapidly reorganize coordination and gait timing to maintain stability and mobility. This is particularly challenging for larger quadrupeds, where increased mass and tighter actuation limits reduce the feasibility of aggressive, high-frequency compensation strategies often observed on smaller platforms. In this work, we propose a deep reinforcement learning approach for fault-tolerant locomotion under actuator power loss. The method employs an asymmetric actor-critic architecture in which the critic has access to privileged information during training, while the actor learns to reconstruct a corresponding latent representation from proprioceptive observations. We introduce a latent-alignment loss that encourages consistency between actor and critic representations. Additionally, we augment the action space with a learnable gait frequency parameter, enabling adaptive gait timing in response to terrain variations and actuator degradation without predefined faulty-leg strategies. The approach is validated in high-fidelity simulation on uneven terrain and real-world experiments on flat ground using a 68 kg quadruped robot.

Abstract

Explore similar work

Learning Fault-Tolerant Locomotion with Adaptive Gait Timing

Mixture-of-Experts RL for Fault-Tolerant Legged Locomotion

SlipSense: Multimodal Sensing for Online Slip Detection in Legged Robots