Construction and registration of 3D humans and environments into digital twins for safe human-robot collaboration

This project aims to develop novel and adaptable digital twin systems for safety critical human-robot collaboration (HRC) systems in manufacturing and healthcare, etc. The project differentiates itself from existing digital twin systems in a way that accurate three-dimensional (3D) human and environment models will be built through deep learning and Simultaneous Localisation And Mapping (SLAM) techniques. Such 3D models will be next registered into digital twins as references to capture and refine diverse human shapes, subtle behaviours, and dynamic environmental objects. One can image that by connecting and synchronising such a digital twin with its counterpart real HRC system, risks caused by dynamic objects including humans approaching proximity to robots will be monitored and predicted, to prevent injuries and other incidents.

This project is a great opportunity to build your skillset and knowledge in developing and applying cutting edge techniques to improve human wellbeing and promote sustainability. You will work with a supervision team of domain expertise from robotics, computer vision, deep learning, to generic AI, to successfully deliver the project. You will gain full support from the Department of Computing and Mathematics and Centre for Advanced Computational Science (CfACS) to facilitate and disseminate your research.

Aims and objectives

This project aims to develop novel and adaptable digital systems for safety critical human-robot collaboration (HRC) systems in manufacturing and healthcare, etc. Such digital twin systems will help to monitor and predict risks such as potential collisions of robots with surrounding dynamic objects including humans, to prevent injuries and reduce financial costs.

Find out more.