Traditionally, design of robots and their actuation systems relied heavily in the intuition of its designers. Intuitive heuristics, such as reachable workspace and maximum continuous torque in worst-case configurations, have been important design tools for a long time. However, this rarely leads to an optimal robot design in terms of capabilities or (energy) efficiency. These problems become even more significant as robots are moving out of structured lab/industrial environments into the real world, and are expected to perform varying tasks, often on battery power. Therefore, there is a clear need to design robots in a more holistic manner: The best robot for the task.
Simultaneously, the need for more dynamic robots has recently promoted the development of actuators using high torque electrical motors combined with lower gear ratios - quasi-direct-drive or proprioceptive actuation. The result has been actuators with reduced reflected inertia, higher speeds and efficiency, and generally reduced output impedance with better interaction capabilities.
In this assignment, the student will focus on numerical optimisation methods of design parameters of both a robotic system (e.g., link lengths of an arm) and its actuation (motors, gearing). The goal is to maximise performance and efficiency of the system, in a single task and over a range of potential tasks (e.g. pick and place of various objects between varying positions). Different optimisation methods may be employed e.g. based on computed torque or full co-design with trajectory optimisation. A simulation study will be performed to compare different approaches and performance metrics.
This assignment is a collaboration with the engineering team of maxon Benelux and the Robotic Drive Systems team of maxon Switzerland. Maxon will support the research by providing insights into development tools and methods, provide data of drive system components and support during commissioning and operation of prototypes. A proof-ofconcept prototype is within the scope of the project: The student will be able to design a prototype system with drive components and electronics from Maxon, and structural design/manufacture in reinforced polymers and/or metal manufactured at RaM.
The best robot for the job: Optimising robot and actuation design through co-design
Finished: 2024-08-26
MSc assignment