Generalized recognition of single-ended contact formations

Contact formations have proven useful for programming robots by demonstration for operations involving contact. These techniques require real time recognition of contact formations. Single ended contact formation (SECF) classifiers using only forces/torques measured at the wrist of the robot have been shown to be quite effective for this purpose. To function properly, however, previous SECF classifiers have required a sizable training set and a constant pose between the force/torque sensor and the manipulated object. Thus, if an object is regrasped and the pose changes, one expects to have to repeat the creation of the training set. We discuss the impact sensor-object pose changes have on two successful classifiers. Experimental data shows that they perform poorly when sensor-object pose changes. We derive, experimentally verify and discuss a method to regain the performance of both classifiers while minimizing the retraining necessary.