Devices and systems for adaptable robots

From soft robotics to nuclear decommissioning, our research develops flexible, intelligent machines inspired by nature and built for real-world challenges.

About the team

In the Devices and Systems for Adaptable Robots team, we develop adaptable robotic systems that operate safely and efficiently in uncertain environments.

Our research spans soft and compliant devices, bioinspired mechanisms and intelligent mechanical transmissions. By combining structural flexibility, embedded sensing and nature-inspired design, we create robots capable of robust interaction and autonomous function in dynamic and extreme conditions – from collaborative tasks to nuclear decommissioning.

Research topics

Explore all our interaction robotics research:

Compliant and soft robotic devices

Rigid robotic systems face inherent limitations when interacting with uncertain environments, where adaptability and safety are critical. Compliance in robotic structures enables passive adaptability, mitigating impact forces and enhancing interaction control. Compliant and soft robotic devices exploit material deformation and mechanical intelligence to achieve robust functionality in dynamic settings.

Our research focuses on the design, modelling, and control of compliant and soft robotic systems that leverage structural flexibility to improve interaction safety, energy efficiency and environmental adaptability.

We investigate mechanical architectures, embedded sensing strategies and novel actuation paradigms to enhance force distribution, shape adaptability and functional versatility. By integrating bioinspired principles and advanced manufacturing techniques, we develop robotic solutions that dynamically adjust to varying constraints, optimising performance across diverse application domains.

Nature-inspired systems and robots

Biological systems exhibit remarkable adaptability, efficiency and resilience, often leveraging compliance to navigate complex environments and dynamic interactions. These natural strategies provide a foundation for engineering robotic systems that transcend traditional design limitations.

At CROSS-Tech, we systematically analyse biological principles to extract functional mechanisms that enhance robotic performance. By integrating bioinspired compliance, distributed actuation and adaptive morphology, we develop mechatronic solutions that address key challenges in robotics, such as energy-efficient locomotion, robust manipulation and autonomous environmental interaction. Our research spans biomechanics-driven modelling, material-driven design, and control strategies, translating nature’s solutions into high-performance robotic systems with enhanced adaptability and robustness.

Mechanical transmission and tribology for adaptive systems

Robotic systems operating in uncertain or extreme environments require transmission mechanisms that balance precision, adaptability and resilience. Traditional rigid transmissions struggle with dynamic interactions, while sensor-based control strategies are often impractical in harsh conditions, such as high-radiation or debris-laden environments.

At CROSS-Tech, we develop mechanically intelligent transmission architectures that embed compliance at the mechanical level to enable safe and efficient force modulation. Our research focuses on Continuously Variable Transmissions (CVT), Variable Impedance Actuators (VIA) and tribological optimisation to enhance power transfer efficiency, mechanical robustness and adaptive interaction capabilities. These technologies enable robotic manipulators to adjust stiffness on demand, operate with minimal reliance on external sensing and withstand challenging operational constraints.

In collaboration with the University of Tokyo, we are applying these principles to the design of a dedicated manipulator for nuclear decommissioning tasks. This system leverages compliant mechanical transmissions to ensure safe and controlled interaction with its environment, contributing to decommissioning efforts at sites such as Sellafield (UK) and Fukushima Daiichi (Japan). By embedding adaptability at the transmission level, we enable robotic solutions capable of autonomous operation in extreme and unstructured environments.