Imperial College London, Soma Design Group KTH

Soma Skins will be interactive soft materials, augmented with actuation and sensing, that can be quickly and flexibly attached to both human and robot bodies to experiment with new interaction possibilities and capture data. Our starting point is our partner KTH’s Soma Bits, a toolkit of soft materials that are placed on the body and triggered to inflate, vibrate, and heat up during soma design workshops. Initial work will extend these with pressure, touch, and temperature sensors to capture data about bodily interactions, and collaborate with partner Ultraleap to add ultrasonic mid-air haptics as a further modality. Subsequent work will collaborate with the Additive Manufacturing group at Nottingham to explore how the 3D printing of sensors and pneumatic muscles into soft materials, building on their work for prosthetic hands and Prof Nanayakkara’s soft robotics team at Imperial to explore how techniques for creating membranes with variable stiffness and control.

The Soma Skins team is exploring how AI can recognise interactions defined from the user’s perspective rather than treating individual senor readings separately.  Multiple sensing modalities — including air pressure, force-sensitive resistors and thermal sensors — are collected simultaneously from SomaSkin prototypes, while interactions such as squeeze, tap and stroke are performed. Somabotics Research Fellow, Kieran Woodward’s AI tools are then used to learn the multimodal patterns associated with these interactions. Initial tests are promising, and we are moving towards more systematic user data collection.

Work is also exploring how conversational AI and LLMs could become part of the SomaSkin design process. Rather than programming interactions node by node, users could describe an intended bodily interaction through natural language, combine this with bodily demonstration and experience provisional behaviours through SomaSkin, then iteratively refine the interaction with the AI. The aim is to make AI a conversational and embodied partner in authoring sensor–actuator behaviours.

A further direction is to apply SomaSkin more broadly as an embodied interface between people, musical AI and robotic systems. This could include augmented musical instruments and performances, as well as physical human–robot interaction where sensing and actuation distributed across human and robot bodies enable richer forms of physical communication.

Kia Hooks soma design grooup at KTH 

Additive Manufacturing at Nottingham 

Ultraleap

Prof Nanayakkara’s group at Imperial 

Interns:

Danruo Zhang – Danruo has a background in AI and industrial R&D, with experience in federated learning, explainable AI, SNNs and applied AI. Her MSc project at Nottingham Trent University explored explainable federated learning for Alzheimer’s prediction.

Xiang Li – Xiang has a background in architecture, urban design and planning, with an MSc in Architecture and Urban Design from the University of Edinburgh and professional experience in landscape, architectural and urban design.

Tian Liu – Tian is studying Business Management at the University of Southampton and has experience in video production, digital content and marketing, including previous work with NetEase Media.

Xiaoyan Wang – Xiaoyan’s background is in embedded systems and digital fabrication, contributing technical expertise in hardware development and fabrication.

Outputs & Highlights

Decoupled, wearable soft robotic rehabilitation device for the upper limb. Greig J, McInnes M, Chadwick ED & Giannaccini M.E. Wearable Technologies , Volume 6 , 2025 , e38

AI and Intelligent Synthetic Skin: Advancing Beyond Human Skin for Sensory Human-Robot Interaction (HRI), Zhou F, Chamberlain A & Benford S. Third UK AI Conference, 23 June 2025, London, UK

SomaSkin has supported embodied prototyping around AI LENS, extending its exploration of real-time interaction with generative AI beyond conventional screen-based interfaces. AI LENS uses live generative transformation of video and allows people to engage with AI through movement, framing and physical interaction with the system.

SomaSkin has supported the prototyping and exploration of soft robotic interactions within The Garden, an interactive installation for children developed with Makers of Imaginary Worlds. The project investigates how soft, responsive objects can support multisensory interaction through touch, pressure, movement and sound, with particular attention to very young children and their interactions with adults.

In work connected to Horizon PhD candidate Angie Higgins, SomaSkin has been used in co-design workshops with older adults living with chronic pain to explore how soft robotic interactions might support pain management. The toolkit enabled participants’ bodily experiences — such as comfort, discomfort, rhythm, intensity and location — to be translated into adjustable and repeatable physical interactions, allowing their lived bodily knowledge to directly shape the design process. Angela Higgins’ wider PhD research similarly focuses on using embodied participatory design with older adults to imagine robotic technologies for health and wellbeing.

SomaSkin has been explored as an interface for augmented musical interaction, connecting bodily sensing and physical or haptic feedback with AI-supported musical performance. This aligns with the wider Traditionable Machines research into how musical AI can become embodied and integrated into instruments, musical practice and live performance.