Slip Detection

Slip detection research focuses on enabling robots to reliably grasp and manipulate objects by identifying the onset of slippage. Current efforts utilize diverse sensor modalities, including tactile sensors (e.g., GelSight, PapillArray, Uskin) and vision, often integrated with machine learning models like deep autoencoders, support vector machines, and temporal convolutional networks to process sensor data and classify slip events. Successful slip detection is crucial for improving the robustness and safety of robotic manipulation in various applications, from industrial automation to planetary exploration, by allowing for real-time adjustments to grasp forces and control strategies. This research also contributes to a deeper understanding of human tactile perception and its role in dexterous manipulation.