For almost 40 years we’ve been using process models, ASMs, containing HI, human intelligence (as opposed to AI). And for all those years modellers have been extracting data from lab logs and SCADA databases, transferring them into process models, setting up simulations, scenarios, calibrations, writing reports; all manually. And model developers were making models more and more complex, adding phosphorus, sulfur, greenhouse gases, sidestream and granular processes, energy and carbon footprint modules and many other features.

This is a lot of manual work which is good for the consultants’

timesheets and consequently plants became more sophisticated and efficient. Then came the Digital Twin (DT) idea. Let’s automate all this work.

In 1992, without ever thinking “DT”, I already set up a system for San Jose, California that autocalibrated and was used for tuning controllers to minimize effluent NO2. Sensor technology was not the same as today… In the talk, I will highlight basic principles of Digital Twins and explain a few live examples that are operating around the world.