Visuo-Tactile Sensors for Robotic Grippers Design and Challenges: A Review

Abstract

Visuo-tactile sensors have become an important sensing technology for robotic grippers because they transform physical contact into high-resolution visual data that can be processed using computer vision and learning-based methods. This review examines visuo-tactile sensors for robotic grippers from the perspectives of sensing principles, integration methods, applications, and deployment challenges. First, the review relates the standard marker-based and intensity-based sensing taxonomy to a four-stage pipeline: mechanical transduction, optical encoding, image acquisition, and tactile inference. This structure explains how contact deformation is converted into tactile images and then into estimates of geometry, force, slip, texture, and contact state. Second, the review categorizes integration methods according to how visuo-tactile sensors are incorporated into grippers, including external modules, fingertip-replacement sensors, integrated tactile finger-pad designs, and distributed or large-area tactile surfaces. The comparison highlights how each integration strategy affects compactness, contact coverage, calibration, durability, and suitability for robotic manipulation. The review further discusses major applications in grasp stability, slip detection, force regulation, object recognition, in-hand manipulation, and contact-rich control. Finally, key deployment challenges are identified, including miniaturization, optical packaging, optical-noise robustness, sensor durability, thermal drift, calibration transfer, real-time inference, embedded computation, distributed synchronization, dataset limitations, and standardized benchmarking. Overall, visuo-tactile sensing offers a powerful route toward more dexterous, adaptive, and contact-aware robotic grippers, but broader deployment depends on balancing sensing resolution, mechanical integration, material robustness, calibration repeatability, electrical architecture, and computational efficiency.