Abstract
Industrial collaborative robots have safety features that, if tuned correctly, make possible their physical interaction with humans. However, in addition to mechanical safety, reducing any kind of psychological discomfort or stress that the robot may induce is also crucial for Human-Robot-Interaction (HRI). On this regard, the term psychological safety has been coined to identify the absence of such unpleasant effects during the HRI experience. Several approaches propose to ensure safety by planning a motion subject to safety constraints. Among these, minimum jerk trajectories have drawn the interest of the robotics community for their advantages from the mechanical point of view and their similarity to human motions. In this work, we revise and comment the most common approaches that have been applied to compute this kind of trajectories through a common framework and discuss them with respect to the authors’ recently introduced method based on the variational formalism.