Summary of content

This course aims to equip students with knowledge and skills required for successful development of digital twins. Digital twin (DT) models are continuously evolving digital profiles comprising of the historical data superimposed on the current behaviour of the physical asset. The digital twin is based on cumulative, real world measurements across a wide range of operational parameters. Whilst in theory the DT concept is sound and it can indeed enable real value, in practice, deploying such models at full scale is still very challenging, for several reasons:

- Realization of digital profiles is heavily dependent on existence of large amounts of historical data.
- Access and processing of raw data relevant for asset's performance is not always possible.
- Unclear requirements on how domain knowledge should be channelled in the DT design process.
- Data visualization and human factors are topics that present significant challenges while superimposing the real time operational and digital twin values in the most visually impactful way.

This module will allow students to become familiar with the components of a DT framework - data, models, algorithms. As part of the module, students will investigate requirements for DT, domain knowledge, data ontologies, modelling dimensions, architectures, and ML/AI development platforms for instantiation of DT. They will also be exposed to COTS software packages capable of developing and validating a design methodology for DTs targeting assessment of components' health and overall asset/process performance.

The module will be tackling the design of DT for Assets, as well as DTs for manufacturing processes.

Module structure

Assessment methods