SlipSense: Multimodal Sensing for Online Slip Detection in Legged Robots
Iris Szu-Yao Liu, Chien Chern Cheah, Meng Yee (Michael) Chuah
AI summary
Problem
Current slip detection methods rely on insensitive kinematic estimates or bulky, fragile force sensors that fail under dynamic locomotion, leaving legged robots vulnerable to early-stage slips on variable terrains.
Approach
The system fuses a lightweight foot-mounted pressure and inertial sensor with an LSTM force inference model and a self-supervised anomaly detector to identify early slip events in real time during dynamic quadruped gait.
Key results
- Detection of early-stage slips at 24.1 ± 6.4 mm displacement with 85.9% accuracy
- 3.3-fold finer resolution and 24% higher accuracy than kinematic baselines
- Lightweight 65g multimodal footpad sensor enabling real-time 3D force inference
- Self-supervised anomaly detection model trained exclusively on stable contact data
Why it matters
This framework provides a practical, high-sensitivity slip detection foundation for legged robots, enabling future force-aware controllers to enhance stability and safety in dynamic, unstructured environments.
Abstract
Legged robots rely on accurate ground interaction awareness to traverse variable terrains, such as slippery sur- faces. Existing slip detection methods often rely on kinematics and proprioception, which lack the sensitivity to detect early- stage slips that occur prior to catastrophic instability. Thus, this paper presents SlipSense, a novel framework for online force-based slip detection using a custom lightweight sensorized foot for quadrupeds to detect slip. The framework integrates a multimodal sensor design with a LSTM-based model to infer ground reaction forces and detect slip-indicative anomalies during locomotion. The proposed framework is deployed on a Unitree Go1 quadruped to demonstrate blind online slip detection over a slippery terrain. Our method detects early- stage slips down to an average displacement of 24.1 ± 6.4mm with an overall accuracy of 85.9%. This represents a 3.3- fold finer detection resolution and a 24% relative accuracy improvement over a standard kinematic baseline that uses foot velocity inferred through state estimation. The work in this paper serves as a foundation for force-aware gait adaptation in legged robotic locomotion, allowing future controllers to estimate terrain friction and adjust constraints, thus improving the overall stability of the system.