A data-driven computational framework to provide deformable solids with the sense of touch

A data-driven, probabilistic, computational framework will be developed that provides a deformable solid with the sense of 'touch', so that it can detect the shape and mechanical behaviour of its environment. The framework will rely on three modules: a mechanical model to simulate the contact between the deformable solid and its environment, a machine learning module to rapidly emulate the mechanical simulations, and a probabilistic framework to identify the shape and mechanical behaviour of the solid’s environment.