SurGE: Surrogate Gradient-guided Evolution for Co-design of Legged Robots with Parallel Elasticity

This research presents SurGE, a framework that computes surrogate gradients of the design objective through a differentiable pipeline consisting of a kinodynamic single-rigid-body (Kino-SRB) model and a design-aware control policy. SurGE injects these gradients into CMA-ES via mean shift with cosine-annealed step decay, achieving improved performance in both simulation and hardware experiments on a hopping robot with unidirectional parallel spring.