Digital Twins for Human Assistive Robot Teams

Previous research has identified difficulties in testing assistive robotics technologies with patients in realistic contexts, as well as a lack of relevant safety standards and regulations for assistive robots for use in health and social carei, particularly those being designed to support vulnerable end-users. Assistive robots for people with long-term conditions or aging-related impairments need to adapt as conditions and impairments change.

To ensure adoption of these technologies, care professionals and regulatory bodies need assurance of system safety, in particular, trust in the autonomous system that learnt behaviours will remain clinically efficacious. This is hard to prove and realise, as testing would need to happen with a diverse range of vulnerable people, over an extended duration, and in different use-contexts.


Additional challenges arise due to technology appropriation changing the original ecosystem it was designed for. A new assistive robot embodiment and the subsequent new approaches to assistance, are likely to alter existing care pathways.

The question then is “how may we foresee and plan for impact that will be multi-faceted and complex and offer assurance that an assistive robot will continue to function as per best practice approaches in care”.

Developing appropriate and relevant approaches for simulating operational use, building on existing digital twin frameworksii, particularly where there is a close proximate physical interaction between the robot and user operating as a team, requires an agile platform that can lead to deeper and more nuanced understanding of inter-related interaction aspects.

Project Aim:

The aim of this project is to investigate to what extent using a digital twin paradigm can be used to simulate, verify, and validate adaptation of assistive robots to support users with complex and changing physical needs.

Building on research from existing TAS Resilience node, the previous research of the team and state of the art in digital twins and bio-mechanical simulation and VR software, this project proposes to develop a digital twin of at least one existing physically assistive robot linked to a co-evolved user model and use this as a demonstrator to test the extent to which this provides valid information to understand how the system should and does adapt to users whose physical condition is changing over time.

Project Team

University of Nottingham: Dominic Price, Praminda Caleb-Solly, Donal McNally

University of York: Xinwei Fang, Sinem Getir Yaman, Radu Calinescu

Heriot-Watt University: Mauro Dragone