Precise grasp is a vital application in robot manipulation while real-time slip detection can help the control system to monitor grasp force and compensate for position deviation. However, commonly-used robots hardly have slip detection modules, and the lack of real-time slip information obstructs the robot from making timely force or position adjustments. This paper proposes a real-time slip detection method using 2D images and surface height maps provided by a tactile sensor. Marker-centered sub-height maps, which indicate the local tangential and normal deformation of the sensor surface, are compared based on fast normalized cross correlation (NCC). The object status is determined from the integration of all the local slips. The experiment validates the proposed method with the average classification accuracy of 88.89% for 21 daily objects in 90 grasp trials, including translational slip, rotational slip, and stable trials. Frame detection accuracy of 94.56% for slip frames and 98.81% for stable frames are achieved. The real-time capability of the algorithm is also tested. The proposed method demonstrates the feasibility of detecting slip using the dual-modal outputs from single tactile sensor which can be installed on robot grippers. Moreover, real-time classification results can be used to monitor force or position regulation strategies and assist precise manipulation, including assembly work and deformable objects operation.
Tactile-Based Slip Detection Towards Robot Grasping
Lect.Notes Computer
International Conference on Intelligent Robotics and Applications ; 2023 ; Hangzhou, China July 05, 2023 - July 07, 2023
2023-10-16
15 pages
Article/Chapter (Book)
Electronic Resource
English
Slip detection , Tactile sensor , Robot grasping Computer Science , Artificial Intelligence , Software Engineering/Programming and Operating Systems , Computer Applications , User Interfaces and Human Computer Interaction , Computer Communication Networks , Special Purpose and Application-Based Systems
Tactile-Based Slip Detection Towards Robot Grasping
TIBKAT | 2023
|Tactile Feedback for Multifingered Dynamic Grasping
British Library Online Contents | 1997
|Fully Tactile Dexterous Hand Grasping Strategy Combining Visual and Tactile Senses
Springer Verlag | 2023
|Stable grasping under pose uncertainty using tactile feedback
British Library Online Contents | 2014
|