The future of Digital Twins is a diverse one — both in meaning and in set-up. Ignoring this reality will hamper the success of Digital Twins.

Here, the focus is on set-up and design considerations for successful implementations of Digital Twins within a disperse service and data landscape (thus excluding here the IoT ingest and data model part). An initial technology mapping is presented.

Digital Twin

What is regarded as a Digital Twin depends on its goal. One example is a Digital Twin of real-life assets, like a car fleet. Based on the status of the fleet decisions will be made in real time about routing, drivers’ brakes, etc. These systems will rely also on external data like traffic information, and environmental conditions like the weather. As decision will progressively be based on advancing algorithms, the patterns of historical occurrences will become important and will include gradually more diverse and rich data. The digital version becomes gradually rich enough to building up understanding of the interactions of the important actors. With that understanding, simulations can be built, and automated decisions can be made with a selection procedure on the results from simulation runs. In other words, one is building a digital truth as well as parallel future possibilities around the assets it concerns.

With another goal in mind, another type of Digital Twin will be built. If, for example, one seeks mitigating actions within a changing climate, the data, the algorithms, the simulations, the regarded time intervals differ greatly. Resulting in decisions or recommendations that differ among Digital Twins — even if they concern partly the same objects. Decisions from Digital Twins depend on their KPIs (key performing indicators) fitting their goal and the approaches. Digital twins will help to optimize operations, but will also help to address societal issues, although underlying social, economic, ethical, responsible, environmental considerations should be made in parallel among a broader audience as this impacts the KPIs a Digital Twin will be designed for.

#digital-twin #cloud #geospatial #big-data #architecture #data-science