Background:
Human-Robot Co-working tasks are a very promising direction in Robotics research with a clear industrial relevance. The robot acts as a helper to a human worker and is in physical interaction with the human to achieve a task. Up till now co-working tasks have been addressed mainly using a fixed base or ground mobile manipulators. However, Human-Aerial Robot co-working came into prominence recently with a project like the EU H2020 Aerial-CORE project, in which the University of Twente is the leader of the Aerial Co-Working Work Package.
The envisioned application is for workstations located at elevated places, for example where workers are operating on power transmission or communications towers. In these applications, it becomes necessary to endow these systems with the ability to interact with humans in a safe way. A safe robot should be able to move around humans without colliding with them. However, humans are dynamic and it is very difficult to guarantee collision avoidance. In the case, that a collision is inevitable, a ”safer” robot is one that minimizes as much as possible the exchange of generalized forces during an impact and thus maximizes the safety of the human.
Assignment:
Problem statement:
Considering a redundant mechanical system composed of a cart carrying a retractable horizontal pole. The cart is significantly heavier than the pole. The system is redundant with respect to the positioning of the tip of the pole. Both the Cart and the pole are controlled via linear actuators. The actuators are implementing a Proportional Damping Controller (PD) with respect to a predefined set-point. The system in question is placed in a scenario where an accidental collision happens with an object obstructing its way. An exchange of forces happens between the Robot and the Obstacle.
Possible research questions
• What is an appropriate metric to assess the safety of the impacted obstacle
• How do the Internal PD gains affect the value of the safety metric
• Based on the relation between the gains and the safety metric, design PD gains which are optimal from a safety point of view.