You will often hear the terms digital twin or virtual twin spoken and it’s certainly a subject we hear a lot more about at XD Innovation. We believe this is a technology that will become more prevalent over time and more mainstream. However……..what is it?
Digital twin and virtual twin are in fact two different methods. Although incredibly easy to confuse I will try and break these down to a digestible interpretation. You will often hear them used universally and sometimes contextually incorrect but there is a difference between the two which I’ll do my best to outline below.
Digital twin, in the eyes of Gartner who are the go to for all software technology research and our good friends at Dassault Systèmes. A digital twin is a digital representation of a physical object, a process, a person or an abstraction which has a process that can be twinned. It can be a living breathing virtual model of the real. Fed by data you can play out scenarios with the digital twin before you apply them in real life to avoid issues which may have been unforeseen. This can go a long way in speeding up time to market and reduce costly re-engineering or shutdowns. The key thing to remember here is that the digital twin is always a digital representation of a real-world object or process.
Now it gets a bit trickier, virtual twins are easier understood if based on a hypothesised scenario. As outlined by Dassault Systèmes, they go beyond the capability of a digital twin. Rather than working with a physical object, they can start to formulate in the concept stage of a product or process. As the project moves from concept to design, to engineering and so on. The virtual twin will evolve in line with these milestones. It will give a representation of your vision and once equipped with enough data. It can be a simulation of your intentions. Like with a digital twin, you can expose this simulation to scenarios like weather, or a failure and the outcomes of this can feed the design process and help steer engineering. It gets to a cycle like phase where the virtual twin evolves with all new data added. Once you move to a physical product or process you are now into the digital twin world as described above. But with the advantage of all the data and knowledge captured and incorporated on the journey to this stage. A virtual twin aligns with the lifecycle of the product and in the connected world we now live in. It can be a cradle to grave process.
The largest deployment of virtual twin technology I am aware of is with Team Tempest. The next ambitious 6th gen fighter jet collaboration between multiple geo’s and engineering companies. A more relative example would be with Claas Tractors. They have a virtual twin methodology in place at their production lines in Le Mans. This has been heavily utilised for the engineering and manufacturing of the Axion 900 Terra Trac. No mean feat! The video link in the document gives some insight into the value this offers their operations.
The value these methods bring to business is the chance to test scenarios before putting them into practice. This can save time, save material waste and contributes to a more sustainable process. Healthcare, automotive, aerospace and manufacturing have started to embrace these technologies at scale and no doubt they will become more accessible as time goes by.
So why am I banging on about it to you? Despite the big rollouts given in this blog. It works for any size of business, wherever you are disrupting and innovating it has its place to bring visons to fruition with less headaches. I also believe that as we learn to leverage Space, the data captured will play its part in developing products and processes and you will see these inputs in future digital and virtual twin models. Be it agricultural data or feedback from Space manufacturing.
My last comment on this topic is like all data heavy practices, it’s only as good as what you put in. Data integrity is critical at any phase but paramount in this application. The GIGO (garbage in garbage out) acronym has been around for years and is more prevalent now than ever. You need to know what you want from the twin, ensure you feed it with good data and don’t use data unnecessarily. Just because you capture it does not mean you have to use it.