Colby - great article.  I tend to think of a DT in layers - GIS, potentially CAD, SCADA, IOT data other than via SCADA, hydraulic model and asset data, including physical locations in 3 dimensions, linked to histories, maintenance data etc.  (You could post other layers too - topology, meters and consumption, weather....).  Once you have that lot integrated then as you say - simulation, operations and system of record are possible from one place, and applying AI in many cases is an incremental add-on, using the data that has already been assembled. There is also an organizational/political dimension to this as well - each of those layers is owned by a different functional silo.  They will each probably have to be convinced that they can contribute their data to the digital twin without losing control of it - and that if they do, the whole will be greater than the sum of the parts, so that they get back more than they contribute.

Published by

Colby - great article.  I tend to think of a DT in layers - GIS, potentially CAD, SCADA, IOT data other than via SCADA, hydraulic model and asset data, including physical locations in 3 dimensions, linked to histories, maintenance data etc.  (You could post other layers too - topology, meters and consumption, weather....).  Once you have that lot integrated then as you say - simulation, operations and system of record are possible from one place, and applying AI in many cases is an incremental add-on, using the data that has already been assembled.

There is also an organizational/political dimension to this as well - each of those layers is owned by a different functional silo.  They will each probably have to be convinced that they can contribute their data to the digital twin without losing control of it - and that if they do, the whole will be greater than the sum of the parts, so that they get back more than they contribute.