Professor Ashiq Anjum and Professor Lu Liu
A digital twin is a digital representation of a physical asset that spans its lifecycle, is updated from real-time data, and uses modelling, machine learning and reasoning to help decision-making. Currently it takes months and years to operationalize digital twins because the process is manual and time consuming. There is no standard theoretical model of digital twins that can be considered as a reference point for developments, optimization and deployments. Without a sound theoretical model of a digital twin, this is a tedious task to map potential new data sources on it or propose new adaptations once the digital twin has been put into practice. This project therefore aims to investigate a system model of digital twins that can learn from engineering, infrastructure as well as operational data and iteratively evolve over time as new data and insights become available.
The resulting model will learn, optimize and adapt to evolving states, representing a desirable or expected functionality of a physical asset, enabling users to generate potential new scenarios as well as data that will be a realistic emulation of the real world experimental environment.
The proposed model of the digital twin will be adapted and deployed in edge devices for performance aware model execution as well as for latency aware immersive analytics. The model will be deployed within edge devices such as digital assistants and handheld devices so that they can acquire, process and interpret data locally.