Collaborative robotics and healthcare have progressed in the last decades, being now in a wide topic with multiple branches, such as surgery, laboratory robots, exoskeletons, prosthesis, therapy and much more. In the case of therapy, collaborative robots are already in use for rehabilitation and available for different application types, e.g. for improving the rehabilitation in lower or upper limbs.

In this project the focus relies on the workspace restrictions on rehabilitation exercises, in order to prevent the patient from doing exercises trajectories incorrectly and potentially deviating the manipulator from its goal position towards an non-allowed space. Overall, the workspace restrictions have been done using an approach based on virtual constraints and in case of not letting the patient deviate the robot, it is based on the cooncept of Artificial Potential Field.

From a practical point of view, the project was developed in order to improve the robotic platform called ROBERT, a collaborative assistant used by therapists, mostly on the recovery from stroke patients.

ROBERT is a platform with a Kuka LBR Med on the top, which has an end-effector that can be attached to the patient's leg with a tool. Afterwards, the robot helps the patient to complete physical routines.

The main goal of this project is to provide a proof of concept under a new perspective on path planning for manipulators in clinical applications.