Researchers have proposed a new approach called AdaControl for designing robot morphology and control systems. They found that traditional methods of evaluating fitness can lead to biased selection towards robots with high morphological intelligence, but low true potential. This bias restricts the exploration of design space and compromises diversity. The team developed AdaControl, which monitors this bias and allocates sufficient control learning for unbiased evaluation. Experiments on simulated soft robots showed that AdaControl outperforms traditional methods in discovering diverse high-performing designs while reducing computation by up to 80%. --- Why it matters: This matters because it provides a new understanding of the relationship between robot morphology and control systems, which can lead to more efficient design processes and improved performance. The proposed approach can be applied to various robotic systems, enabling researchers and engineers to better optimize their designs and explore new possibilities. Source: https://arxiv.org/abs/2608.23100