Shaping the Evolutionary Dynamics of Robot Morphology via Adaptive Control Learning
2026-08-24 • Robotics
RoboticsArtificial Intelligence
AI summaryⓘ
The authors studied how a robot’s body shape (morphology) and control system (brain) evolve together over generations. They found that good body designs help the robot learn faster (morphological intelligence), but there is also a hidden potential for performance that is often underestimated. By improving how learning is evaluated, their method called AdaControl avoids biases that favor only fast learners and helps find better, more diverse robot designs more efficiently. This challenges a previous idea about how body and brain influence each other during evolution. Their approach speeds up search for good designs while maintaining quality.
robot co-designbi-level optimizationmorphological intelligencecontrol learningmorphological evolutionfitness evaluationgenetic algorithmBaldwin effectvoxel-based soft robotsoptimization efficiency
Authors
Junru Song, Yang Yang, Yaqing Xu, Ying Wen, Wei Peng, Guozhen Li, Wei'en Zhou, Wen Yao
Abstract
Robot co-design via bi-level optimization couples within-lifetime controller learning for fitness evaluation with cross-generational morphological evolution. Prior work has established that well-adapted morphology facilitates faster control learning, a property termed morphological intelligence. Yet how control learning reciprocally shapes morphological evolution remains unexplored. This paper examines both directions for a holistic account of brain-body interplay. We first show that morphological contributions to control learning decouple into two orthogonal dimensions. We formalize the convergence speed as morphological intelligence and identify the performance ceiling as a complementary quantity termed true potential. A concise functional relation is then established to jointly characterize both quantities from individual learning curves, which, when aggregated at the population level, capture evolutionary profiles. Through extensive experiments on simulated voxel-based soft robots, we reveal that premature fitness evaluation systematically underestimates true potential and biases selection towards fast learners. This restricts design space exploration, compromising both optimization efficiency and morphological diversity. Notably, the widely recognized morphological Baldwin effect emerges as an artifact of this bias rather than a general evolutionary tendency. We therefore propose AdaControl, which monitors disproportionate selection for morphological intelligence during evolution and allocates minimally sufficient control learning for unbiased fitness evaluation. With AdaControl, a simple genetic algorithm rivals state-of-the-art generative-model-based co-design methods in discovering diverse high-performing designs while cutting computation by up to 80% versus exhaustive control.