A Multiple Working Mode Approach to Hammering With A Modular Reconfigurable Robot

Abstract

Robot joints could be damaged by impulsive forces due to impact when the robot performs hammering operations. This thesis presents the development of a strategy that allows a modular reconfigurable robot to safely perform nail hammering. To do this, a multiple working mode approach is applied to switch the selected joint(s) to passive mode with friction compensation to allow free rotation during impact. Analytic impulse models are used to predict joint impulses which can be computed offline or online and serve as criteria for mode switching. Joint impulses are constrained in case the hammer collides with such an object of infinite effective mass. Advantages of the proposed approach may include savings on space, weight, costs, and complexity for a limited range of nail/board environments. An experimental study validates analytic models of hammering and effectiveness of multiple working mode approach.