Belichaamde intelligentie voor zachte robots door middel van mechanische instabiliteiten

Advances in robotics have historically followed progress in digital computation, leading to increasingly powerful centralized control architectures. While effective, this approach contrasts with biological systems, which achieve adaptability and efficiency through decentralized control and tight coupling between morphology, materials, and environmental interaction - an approach known as embodied intelligence. Two fundamental architectural differences underlie this contrast: rigid robots rely on discrete, single-function components, whereas biological systems exhibit continuous morphologies with multifunctional elements. Soft robots naturally align with biological architectures through their compliance, functional integration, and sensitivity to mechanical interactions. Their ability to undergo large elastic deformations further enables the deliberate exploitation of mechanical instabilities as computational and control resources. This work investigates how such instabilities can be harnessed to shift control functionality from centralized software into the physical body. Three soft robotic systems are presented. First, a bio-inspired actuation strategy enables rapid, high-energy motion in inflatable actuators by storing elastic energy and releasing it through pressure-triggered mechanical detachment, experimentally demonstrated via repeated jumping. Second, a hybrid electro-pneumatic central pattern generator exploits the mechanical bistability of a truncated conical membrane and a soft strain sensor to generate self-sustained, tunable oscillations with modular control. Third, the Pneumatic Multilevel Schmitt Trigger introduces multilevel pneumatic logic, enabling compact memory, demultiplexing, and oscillatory circuits with increased per-component functional density. Together, these contributions demonstrate how mechanical instabilities can enhance embodied intelligence, enabling robust, decentralized, and efficient soft robotic systems.