Towards Autonomous Robotic Assembly: Using Combined Visual and Tactile Sensing for Adaptive Task Execution

Abstract Robotic assembly tasks are typically implemented in static settings which parts kept at fixed locations by making use of part holders. Very few works deal with the problem moving industrial applications. However, having autonomous robots that able to execute dynamic environments could lead more flexible facilities reduced implementation efforts for individual products. In this paper, we present a general approach towards robotic combines visual and intrinsic tactile sensing continuously track within single Bayesian framework. Based on this, it is possible implement object-centric skills guided estimated poses parts, including cases where occlusions block vision system. particular, investigate application peg-in-hole assembly. A tilt-and-align strategy using Cartesian impedance controller, combined an adaptive path executor. Experimental results multiple combinations provided analyzed detail.