Stochastic Locomotion Emerging From Body–Environment Interactions in a Mexican Jumping Bean–Inspired Robot

Biological adaptive locomotion has long inspired robotic solutions for navigating unstructured environments. However, animal‐inspired systems—and, more recently, plant‐inspired ones—are typically considered separately. This study merges strategies from both kingdoms through a design inspired by the Mexican jumping bean. Specifically, the moth larva Cydia saltitans , which hatches inside the seed of Sebastiania pavoniana , generates temperature‐induced multimodal locomotion enabled by the seed's peculiar curved–flat geometry. Translating this principle into robotics, Bean‐Bot is a frugal, 1.3 cm thermo‐responsive jumping module enclosed within a bean‐shaped shell. Upon temperature increase, it achieves terrain‐sensitive locomotion—including jumping (up to 30 cm), flipping, rolling, sliding, climbing, and self‐reorientation—with 81.5% of jump events transitioning into one or more additional locomotion modes after landing, capabilities absent in conventional jumping robots. Unlike control‐centric or tethered multimodal systems, the electronics‐free, untethered Bean‐Bot achieves multimodality through a dual‐kingdom‐inspired embodied design that couples material properties, morphology, and the environment. Its bean‐inspired geometry passively modulates jump trajectories and enables terrain‐dependent mode switching. This bioinspired strategy demonstrates an aspect of physical intelligence by achieving multimodal behavior through the tight coupling of temperature‐responsive materials, shell morphology, and environmental interactions, with minimal computation. Through temperature variations, Bean‐Bots move stochastically and may offer new strategies for navigating unstructured environments.