Advanced AI-Based Slip Detection on an ESP32 for Bionic Grippers

Slip detection is a critical aspect of robotic manipulation, enhancing the stability and dexterity of robotic grasping systems. Conventional slip detection approaches often rely on computationally intensive hardware, limiting their applicability in lightweight and mobile robotic platforms. In this work, we present an edge-computing solution for slip detection using an ESP32 microcontroller, enabling on-device inference without external dependencies.The developed system is implemented on a Fin Ray-inspired soft robotic gripper, mounted on a UR3 robotic arm. Experimental validation using diverse objects with varying material properties demonstrates that our ESP32-based solution achieves reliable slip detection in real-time, with the responsiveness required for dynamic robotic grasping applications. By leveraging lightweight AI models, our proposed method provides an efficient, cost-effective and scalable solution for slip detection in mobile and stand-alone robotic systems.