Data-Driven Modelling of Deformable Materials for Robotic Manipulation

这个项目是关于开发和实施数据驱动的方法，用于机器人对变形材料的操纵。这将需要使机器人能够使用多模式反馈(例如，触觉感知、扭矩和视觉反馈)逐步学习物理属性，并在积极探索期间进行因果推理，以确定材料的结构。该项目的抱负之一是开发最终将对生物和医学产生影响的技术。

This project is about developing and implementing data-driven methods for robotic manipulation of deformable materials. This will entail enabling robots to learn physical properties incrementally, using multimodal feedback (e.g. tactile sensing, torque & visual feedback), and to reason causally during active exploration, in order to figure out the structure of a material. One of this project's aspirations is to develop technology that will ultimately be impactful in biology and medicine.