Title: Towards Adaptive Robots: Embedding Physical Intelligence through Geometry and Smart Materials
Abstract: Robots operating in complex and changing environments must adapt not only their behaviour but also their physical form, mechanical properties and modes of movement. This raises fundamental questions: How can intelligence be embedded directly into a robot’s body to enable adaptive and reconfigurable movement? How can a robot’s physical form contribute actively to the sensing–decision–actuation cycle?
This talk explores these questions through two complementary approaches to physically intelligent robot design. The first draws on geometric principles derived from the ancient arts of origami and kirigami. Our work demonstrates how folding, cutting and structural constraints can encode movement and functionality directly into robotic mechanisms. Examples include a kirigami-inspired metamorphic mechanism with programmable motion modes and soft–rigid hybrid robots incorporating kirigami-inspired skeletons. These systems demonstrate how carefully designed geometry can generate deployable, compliant and reconfigurable movements while reducing mechanical and control complexity. The second approach investigates how smart materials can enable actuation, sensing and adaptable mechanical properties to be combined within the robot body. In particular, new designs of self-sensing variable-stiffness artificial muscles combine actuation, stiffness modulation and sensing, significantly simplifying robot construction. I will present our electrically driven artificial muscle and its application in a wearable hand tremor-suppression glove, demonstrating how integrated sensing and actuation can support responsive physical interaction and active movement regulation.
Together, these studies demonstrate how the co-design of geometry, structures, smart materials and control can enable the robot body to contribute actively to the sensing–decision–actuation cycle. This approach offers a pathway beyond fixed mechanical architectures towards robots that can adapt their form, movement and mechanical properties in response to changing tasks, users and environments.
–
Dr Ketao Zhang is a Reader in Robotics at Queen Mary University of London, where he directs the Robotic Systems Research Group within the Centre for Advanced Robotics (ARQ). He is also the Industrial Engagement Lead for the Centre for Intelligent Transport.
His research focuses on embodied and intelligent robotics, integrating robot body design, sensing, control and learning to enable robust interaction with complex environments. His work spans reconfigurable mechanisms, aerial robotics, and soft robotics, underpinned by expertise in robot kinematics, dynamics and the co-design of robotic bodies and control.
Dr Zhang has published more than 100 articles in leading journals and conferences, including Nature and IEEE and ASME Transactions. He led the development of an aerial additive manufacturing approach using teams of autonomous drones, opening new possibilities for building and repairing structures in remote, high or difficult-to-access locations. This research was published in Nature and featured on its cover in September 2022. He received the Institution of Civil Engineers’ Howard Medal for his research on drone-based additive manufacturing.
Dr Zhang is an Associate Editor of IEEE Robotics and Automation Letters, Mechanism and Machine Theory, and the ASME Journal of Mechanical Design.