Tensegrity mechanisms for robotic manipulation

Manipulation is at the heart of many industrial processes, and is historically one of the first applications of robotics. A robotic manipulator consists of a base linked to an end effector by a motorized kinematic chain. Industrial robots are traditionally made up of one or more kinematic chains (in series or parallel architecture) linking the base to the end effector via solid, rigid bodies articulated together by mechanical links. Given the high dynamics and power required of these industrial robots, human and machine workspaces are generally kept separate for obvious safety reasons. These industrial robots are therefore ill-suited to interaction robotics applications where man and robot work side-by-side, or even share the manipulation task in the case of co-manipulation. In this case, it is necessary to consider alternative robotic architectures in order to design robots that are both lightweight and robust, enabling them to perform a manipulation task efficiently while ensuring safe interaction with the human worker.

The use of tensegrity mechanisms in this context is the subject of this article. These mechanisms are derived from tensegrity [C 2 471] , a class of structures composed of bars held in compression by a network of tensioned cables. Tensegrity mechanisms are obtained by actuating one or more of the structure's elements. These mechanisms can withstand heavy loads while remaining light, compliant and capable of adapting to their environment. Their application is therefore of great interest in the field of interaction robotics. This article provides specific details of their use in this context, as well as the technological and fundamental challenges associated with their use.

The first section introduces the basic principles of tensegrity mechanisms and their consequences on their behavior and modeling, then carried out in a static case. The second section focuses on modeling methods and simulation tools specific to these systems. In the third section, we present the analysis tools used to evaluate their performance. The fourth section highlights the design issues involved in exploiting tensegrity mechanisms in the context of robotic manipulation. The fifth section introduces the control strategies available to control not only the mechanism's pose, but also its intrinsic stiffness. Applications of tensegrity mechanisms in robotic manipulation are presented in the sixth and final section.