Exploiting mechanofluidic instabilities to enable soft autonomous machines

During the past decade, using soft materials to build robots has gained significant traction in the scientific domain. In contrast to more traditional ‘rigid’ robots operating in settings where precision and speed are essential, soft robots target applications where human interaction, unstructured environments, and robust behaviour are key. Despite these exciting developments, most of the current electronic control, intelligence, and power systems of soft robots are too bulky for embedded use and therefore limit their applicability.



To imprint soft robotics with bio-inspired forms of autonomy, we are in the process of developing smart fluidic circuits that harness nonlinear mechanical behavior to replace electronic control, that directly interact and respond to their environment. As an illustration, one of our recent breakthroughs involves an elastic valve with a slit that demonstrates hysteretic behavior. We have demonstrated the ability to utilize this valve to convert a continuous flow into a pulsatile flow, a crucial requirement for enabling programmable locomotion and activating components like a soft robotic heart that we are developing. Similarly, alternative fluidic circuits feature nonlinear components such as kinking tubes, which exhibit instabilities to enhance actuation frequencies and movement speed. Importantly, these soft components change their behavior considerably in response to interactions with their environment, resulting in useful self-sensing behavior, that can be utilized for dynamic synchronization and autonomy.