Digital Twins Are on the Rise

As our faithful readers might know, we are a data & AI consulting firm specialized in two industries: logistics and construction. While they may seem different on the surface, they actually have a lot in common (we'll keep this for a different blog post). A technology that holds much promise for both will be the topic of today: "digital twins."

What is a digital twin?

One could refer to a digital twin as a high fidelity digital simulation or replica of an offline artefact or process.

Common traits of digital twins are:

1. They constitute a virtual model of a real thing
2. They simulate both the physical state and behavior of the "thing"
3. They are unique
4. They are connected to the physical thing

The evolution of digital twins

Digital twins might sound new - yet they are older than what most of us would think. In fact, in the aerospace industry, they go back decades. Why? The reason is relatively simple: they allow for simulations and testing on high value assets found in inaccessible operating conditions.

Imagine the following: you want to test the effect of different types of "bird strike" on a Rolls Royce Airbus A380 (that's an airplane for those who don't know) engine.

Solution 1: you fire dead birds into a running engine somewhere in a wind tunnel and verify what happens. Cost of the experiment (in case the engine is totally ruined): around $ 30 million.

Solution 2: you simulate a bird strike on a digital twin of the engine. Cost of the experiment: < $ 1 million.

Admittedly, an aircraft engine is an extreme case and constitutes a "physical" thing. You could also apply the digital twin concept to a process. Example: you could map how a sales process looks like in a company (from customer first contact to closing the deal and aftercare). The advantage is that - not in the least because of more powerful AI algorithms - you could simulate a multiplicity of scenarios. In logistics, a digital twin could offer a (near) real-time overview of the entire supply chain. One could see which goods are stored in which warehouses in which exact quantities. Next, operators would be able to track these goods across the different transportation lines (including their costs). As such, 3PL (third-party logistics) companies would be able to see (near real-time) how their margins evolve across the chain (predicting loss of margin when getting stuck in traffic or delivering the wrong cargo).

Underlying technologies

"Digital twins" do not exist in a vacuum, their success goes hand in hand with a number of "emerging" technologies such as:

• IoT (internet of things)
• Cloud computing
• API (and open standards)
• AI (artificial intelligence)
• AR/VR/XR (augmented, virtual and extended reality)

We expect much from digital twin technology in the industries we serve. Some of the bottlenecks such as data quality, interoperability and cyber security will require a decent data strategy, data infrastructure and know-how to limit the risks. We're happy to learn about providers out there. In turn, we at Nemeon can provide advice as to the "data side" of digital twin projects.