Revolutionizing Soft Robotics: A New Control System for Safer Interactions
Imagine a soft robotic arm gracefully bending around a delicate bunch of grapes, adjusting its grip in real-time as it lifts the fruit. Unlike rigid robots, which often avoid contact with the environment and humans for safety, this arm senses subtle forces, stretching and flexing like a human hand. Every motion is calculated to avoid excessive force while achieving tasks efficiently. This innovation comes from the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and the Laboratory for Information and Decisions Systems (LIDS), aiming to create robots that can safely interact with humans and delicate objects.
Soft robots, with their deformable bodies, hold the promise of a future where machines seamlessly move alongside people, assist in caregiving, or handle fragile items in industrial settings. However, their flexibility makes them challenging to control. Small bends or twists can produce unpredictable forces, increasing the risk of damage or injury. This challenge has spurred the need for safe control strategies for soft robots.
The research team, led by Assistant Professor Gioele Zardini, aims to adapt safe control and formal methods from rigid robots to soft robotics. They focus on modeling complex behaviors and embracing contact rather than avoiding it, enabling higher-performance designs without compromising safety or embodied intelligence. This approach is shared by recent parallel work from other groups.
The team developed a new framework that combines nonlinear control theory, advanced physical modeling, and real-time optimization to achieve "contact-aware safety." At the core of this approach are high-order control barrier functions (HOCBFs) and high-order control Lyapunov functions (HOCLFs). HOCBFs define safe operating boundaries, ensuring the robot doesn't exert unsafe forces. HOCLFs guide the robot efficiently toward its task objectives, balancing safety with performance.
The lead author, Kiwan Wong, explains that the approach involves complex derivations of soft robot dynamics, contact models, and control constraints. However, specifying control objectives and safety barriers is straightforward for practitioners, resulting in tangible outcomes. The robot moves smoothly, reacts to contact, and never causes unsafe situations.
Maximilian Stölzle, a co-lead author, highlights the gap in "cognitive" intelligence in soft robots compared to rigid serial-link manipulators. This work helps bridge that gap by adapting proven algorithms to soft robots and tailoring them for safe contact and soft-continuum dynamics.
The team tested the system in various experiments, demonstrating its adaptability and safety. The arm maintained precise force while pressing against a compliant surface and adjusted its grip to avoid slippage when tracing curved objects. In another demonstration, the robot manipulated fragile items alongside a human operator, reacting in real-time to unexpected nudges or shifts.
Looking ahead, the team plans to extend their methods to three-dimensional soft robots and explore integration with learning-based strategies. By combining contact-aware safety with adaptive learning, soft robots could handle even more complex, unpredictable environments.
The research team's work, supported by various scholarships and programs, was published in the Institute of Electrical and Electronics Engineers' Robotics and Automation Letters, marking a significant step towards safer and more versatile soft robots.
