Co-optimized volumetric muscle design for large dynamic deformations in versatile biohybrid robots

Biohybrid robots use engineered living muscle tissue as actuators, offering self-organization, adaptability, and self-healing. Current designs focus on improving muscle physiology but overlook the interplay between muscle and scaffold in the robot's function. We present a computational and experimental pipeline to co-optimize the form and function of centimeter-scale bioactuators. Using a soft-body simulation framework, we modeled unified muscle-scaffold systems and applied an evolutionary algorithm and targeted parameter sweeps to maximize range of motion. The resulting bioactuators, fabricated by integrating skeletal muscle tissue with microgrooved hydrogel scaffolds, feature a continuous interface enabling efficient force transmission and large, rapid deformations. They achieved up to an eleven-fold increase in range of motion over previous designs of similar muscle volume, powering robots capable of jumping, swimming, walking, and gripping, and scalable into multi-unit systems. By unifying modeling, optimization, and fabrication, we demonstrate how such a pipeline accelerates the development of high-performance biohybrid robots.