King's College London

Dr Shan Luo from the Department of Engineering has been awarded an Engineering and Physical Sciences Research Council (EPSRC) Open Fellowship to revolutionise how tactile robots - with a human-like sense of touch - are used in manufacturing, retail and healthcare.

Working with partners such as Unilever, Ocado, Shadow Robot and MathWorks, Dr Luo will develop a series of simulations that predict how a robot will interact with its environment in real-time. By providing feedback to automate the refinement of the robot’s sensing, structure and configuration at the same time, the programme will also teach it to perform its job better.

Over the next decade, robotics have been predicted to boost the productivity of the UK economy by 22%. Tactile robots are slated to be a major driver of this in areas such as manufacturing and healthcare, taking on tasks such as packing delicate objects and enhanced physical rehabilitation.

However, despite the recent rollout of new touch-based sensing technology such as GelTip, significant technical barriers remain to unlocking the full potential of tactile robotics. Often there is a gap between accuracy from testing in the lab and practical application, making real-world deployment challenging.

Building on his previous experience building simulation models for tactile sensors and algorithms to interpret touch data, Shan will use his five-year EPSRC Open Fellowship to find solutions to these challenges. His fellowship, entitled TacDiff, aims to develop a new paradigm of data-driven designs for tactile robots, through differentiable simulations that provide much more granular feedback to assist in training.

Through this project, Dr Luo will develop robots that can learn from interactions with their environment, predicting their next movement, assessing outcomes, and continuously improving their performance. This approach will pave the way for a new generation of robots that are more dexterous, adaptive, and capable of operating safely in real-world settings.

Dr Luo explains, “Robots with a sense of touch could be revolutionary for the way we think about automated tasks. With a soft and dexterous touch, robots could help pack delicate objects with speed, assist in manufacturing, or even support remote surgical operations, but current approaches are limited. The components for a robot are built in isolation and training a robot in real-world settings are prohibitively expensive, leading to an accuracy gap between the lab and the real world.

“Through TacDiff, I aim to create a new paradigm of data-driven tactile robot design, where structure, sensing and control evolve together. This will lay the foundation for the next generation of robotics.”

The fellowship will also facilitate the creation of TaRoSim, an educational toolbox designed to make advanced tactile robotics accessible to students in high school and higher education. By offering interactive resources and hands-on learning experiences, TaRoSim hopes to inspire future innovators and build a skilled workforce ready to drive future advancements in robotics.