ECoDe: A Sample-Efficient Method for Co-Design of Robotic Agents
This addresses the challenge of data-intensive reinforcement learning in robotic co-design, offering a more efficient method for researchers and engineers, though it appears incremental as it builds on existing multi-fidelity and warm-starting techniques.
The paper tackles the sample inefficiency in co-designing robotic agents by optimizing both controller and physical design simultaneously, proposing a multi-fidelity exploration strategy that uses a universal policy learner to warm-start controller learning, resulting in superior performance across various agent design problems with observed design simplifications and non-intuitive alterations.
Co-designing autonomous robotic agents involves simultaneously optimizing the controller and physical design of the agent. Its inherent bi-level optimization formulation necessitates an outer loop design optimization driven by an inner loop control optimization. This can be challenging when the design space is large and each design evaluation involves a data-intensive reinforcement learning process for control optimization. To improve the sample efficiency of co-design, we propose a multi-fidelity-based exploration strategy in which we tie the controllers learned across the design spaces through a universal policy learner for warm-starting subsequent controller learning problems. Experiments performed on a wide range of agent design problems demonstrate the superiority of our method compared to baselines. Additionally, analysis of the optimized designs shows interesting design alterations, including design simplifications and non-intuitive alterations.