by Dragan Savic on Linked InIn 2018 it was #AI on the top of the Gartner hype curve, but #digitaltwins (together with #deeplearning) are taking ...

by Dragan Savic on Linked InIn 2018 it was #AI on the top of the Gartner hype curve, but #digitaltwins (together with #deeplearning) are taking ...by Dragan Savic on Linked In

In 2018 it was #AI on the top of the Gartner hype curve, but #digitaltwins (together with #deeplearning) are taking that position now in #watermanagement. We are normally slow.

Quite often definitions I have come across for digital twins in the water sectors are basically describing a model. For years in water management, we had real-time models, also called online, state estimation and data assimilation models, which all take advantage of the available real-time data to update model predictions. What is one thing that makes digital twins different from any of the above?

I can see only one and that is two-way communication between the physical (i.e., water systems prototype) and digital (i.e., model) twin (as opposed to one-way where information from the prototype is used by the model). While that makes a lot of sense in manufacturing, in water systems it is only useful if we have remote control (automation) capabilities where pumps, valves and other elements of a water/wastewater system can be controlled remotely.

Is that it?

Any other?

For those in the know, can you please share your thoughts. I think that we as a community need to come up with a basic idea about the concept. This is my call for action to make sure that we develop a common understanding.

SOURCE: https://www.linkedin.com/in/dragansavic/