Variable Admittance for pHRI: From Intuitive Unilateral Interaction to Optimal Bilateral Force Amplification

In this paper, a novel control architecture for physical human-robot interaction with a multi-degree-of-freedom (multi-dof) robotic manipulator is presented. An admittance regulation is used as a single control structure to effectively manage unilateral and bilateral interaction modes. This control approach is known to lead to intuitive manipulation for the operator during unilateral interactions. However, the effectiveness of this type of control for bilateral amplification on a rigid environment is often questioned. Here, variable admittance parameters are used in order to accommodate and optimize the system response to any potential dynamics. Three interdependent varying control laws are therefore presented, namely, a common unilateral variable admittance control, a bilateral gain scheduling control, and a continuous transition control. A 7-dof Kuka LWR 4 is used to conduct the experimentation and demonstrate the effectiveness of the control algorithms. A video showing different pHRI tasks with the novel control architecture is also provided.