Using Machine Learning to Build Digital Twins of Urban Traffic Scenarios – TAKEN

A ‘digital twin’ is a digital model of a real physical artefact or environment that mirrors the state of its real-world counterpart (ideally) in real time by integrating data available via the ‘Internet of Things’ (IoT). A digital twin of a roadway would represent the physical road and its current traffic flow and provide a basis for implementing traffic management applications or assessing the impact of changes to the infrastructure on traffic flow. The goal of this project is to develop a framework to allow digital twin of urban road networks (e.g., the roads around Trinity) to be developed.

The focus of the work will be on using machine-learning and/or other sensor fusion techniques to reconstruct accurate traffic flows representing the current traffic on the road from available sensor data, which may be delayed, incomplete and/or inaccurate. The project will build on prior projects concerning digital twins of motorway scenarios.