Mega Trends - Digital Twins?

Image: Australian Mega Fauna

Recently I was asked about what I thought about the impact of ICT Mega Trends on technology/business strategy.  The 1st question is what are "Mega" trends? My guess is that they are the new buzz words coined in the last 6 months which will be replace equally as quickly?  Just like Mega Fauna?!

The latest Garner mega trends is here.  Most of these are obvious but a couple caught my attention: Digital twins and Meshes.

Trend No. 5: Digital Twin

Within three to five years, billions of things will be represented by digital twins, a dynamic software model of a physical thing or system. Using physics data on how the components of a thing operate and respond to the environment as well as data provided by sensors in the physical world, a digital twin can be used to analyze and simulate real world conditions, responds to changes, improve operations and add value. Digital twins function as proxies for the combination of skilled individuals (e.g., technicians) and traditional monitoring devices and controls (e.g., pressure gauges). Their proliferation will require a cultural change, as those who understand the maintenance of real-world things collaborate with data scientists and IT professionals.  Digital twins of physical assets combined with digital representations of facilities and environments as well as people, businesses and processes will enable an increasingly detailed digital representation of the real world for simulation, analysis and control.

So this isn't a term I'd heard before but I think it's just talking about modeling and simulation from sensor data. There are a few complexities here however. For example, one model doesn't fit all sizes. In practice multiple models are useful for reasoning and predicting different metrics/properties of systems. One is too complex. How will multiple models be managed? So more likely digital twins.





See https://www.youtube.com/watch?v=v9KXXz9B-K8 

The other problem is that you will need to be able to start reasoning about the whole system, including the digital twins. This is sort of like the observer/observed paradox (in science and social sciences). i.e. observing something changes the behaviour of the thing being observed. But in the case of the a system of observer/observed components this will result in the behaviour of the system being changed continuously.  Whoops.

What's a real example of  a digital twin? Here's one for wind turbines.

My suspicion that one impact of this is that the distinction between the observed and observer systems will vanish. In practice many "edge" devices (observed) will contain the code for observation (at least for semi-autonomous operation), maybe synced with the "centralised" or in practiced distributed observer systems.  So self-observation, perhaps slaved some of the time to other systems (thinks Minds, AI-slaved systems, and Avatars/Drones from the Culture novels).

Over the last 10 years I've designed and built a "digital twin" system for monitoring, modelling and reasoning/predicting the behaviour/changes for enterprise software systems. This consumes APM data, builds multiple models (depending on the need) and enable simulation and other solutions for future predictions. The problem is getting the level of abstraction right so you can simplify the observed system without just replicating it in it's entirety (which is one approach but impossible in practice and hard to do in real-time). And any abstraction results in simplification and loss of accuracy.  In practice you also need to be able to do this in real-time, continuously and incrementally faster than the requirement to cope with change and make prediction and react and fix problems. Basically a cloud problem at this point (you need more resources for the observer system than the observed system). But what if the observed system is also just the cloud. Uh oh. Infinite recursion............  Oh at that point the Singularity probably occurs (or even sublimation).

I.e. to iterate is human, to recurse. divine.




And and given my software performance modelling example it's obvious that you will end up with a hierarchy of observed - observer relationships. I.e. each observer can be be observed and have it's own observer(s). So it won't just be "physical" systems that are being observed, but software systems (and hence recursion).

Mesh systems another time!