Emerging Technologies

Digital Twins for Industrial and Infrastructure Applications

A digital twin is often sold as a three dimensional model of an asset, which is a visualisation; a twin is a model that is connected to the real thing and used to answer a question the real thing cannot be asked safely. This course builds toward that. Delegates work through defining the question first, since the required fidelity follows from it and nothing else. Model types are covered: geometric, physics based, data driven and hybrid, with what each can and cannot answer. Data connection follows, including sensing, sampling rate and the latency the use case tolerates. Calibrating a model against real behaviour and keeping it calibrated as the asset ages are addressed, since drift is the usual reason twins are abandoned. Simulation and what if analysis are covered. Integration with maintenance and operations systems follows. The course closes on cost, on ownership and on judging a vendor claim.