Model-based is a topic in the air of many discussions about future digital transformation in engineering and manufacturing. I honestly admitted my confusion about “model-based” in my article earlier this year – Model-based confusion in 3D CAD and PLM. Jos Voskuil – one of the best advocates of systematic model-based approach and PLM, put together few articles that I think worth reading. The last one is Model-based: Digital Twin – the PLM side (you can find links to other article inside).
This article caught my special attention because in my view, it touches one of the most fundamental aspect of PLM development – future data management architectures. Digital Twin is an interesting topic. It forces companies to think about how to connect dots and data silos, which by itself is a big challenge. But digital twin brings a good portion of business value and profit into data management and IT related discussions about PLM.
According to Jos’ the future is federated. The following picture presents kind of “federated picture” of potential future digital enterprise foundations. The most funny part in this picture in my view that it has absolutely nothing to do with the notion of federation. It doesn’t even present any connections between systems. What we can see is competing platform domains – ERP, CRM, PLM, MES etc. Each of these platforms will be happy to become a center of enterprise digital universe. Since the picture is created by PLM consultant, I’m not surprised to see CIMdata’s Product Innovation Platform in the middle.
Here is passage from Jos’ article:
For an R&D department the information from an individual Digital Twin might be only relevant if the Physical Twin is complex to repair and downtime for each individual too high. Imagine a jet engine, a turbine in a power plant or similar. Here a Digital Twin will allow service and R&D to prepare maintenance and simulate and optimize the actions for the physical world before.
The second part where R&D will be interested in, is in the behavior of similar products/systems in the field combined with their environmental conditions. In this way, R&D can discover improvement points for the whole range and give incremental innovation. The challenge for this R&D organization is to find a logical placeholder in their PLM environment to collect commonalities related to the individual modules or components. This is not an ERP or MES domain.
The question about logical placeholder in PLM for IoT data is conflicting with the idea of federated future. Because, if future is federated, systems will be interconnected and obtaining information from other systems. There is no point to search for a logical placeholder. So, we are going to see future fight for data ownership is coming soon.
The discussion about logical placeholders made me to think again about federated data architectures. Federation was around for the last 20-25 years. I can see 2 big approaches:
1- Federate architecture
2- Federated databases
The difference might be not big from the beginning, but long term impact is huge, in my view.
Federated approach is focusing on coordinated information share and exchange which is organized by describing of common models (concepts and behaviors). Information can be transferred using various mechanisms, but components (or systems) are independent. This is what we can see in most of PLM/ERP/MES environments these days. It demands few important architecture elements such as federated identity, SSO and few others.
Federated databases is a different approach, which creates meta data management system with federated model and query plan. By using this type of database, you can actually run queries in multiple databases based on some query plan. A federated database, or virtual database, is a composite of all constituent databases in a federated database system. There is no actual data integration in the constituent disparate databases as a result of data federation In such way federated databases are focusing on federated query plan and ability to prevent multiple siloed databases to be physically and logically integrated as a single piece.
What is my conclusion? As much as companies are focusing on digital twin discussions, I found most of them are missing the point of data architecture and data integration. No surprise… Each vendor believes that the data will be captured and located in the their own database – PLM, MES, ERP, CRM. But the reality is different in my view. The data is most probably will create a new information slio and will be demanding integration with other data silos and systems. So, when you start you digital twin projects, you need to allocate time to spend on data architecture to support digital twin models. All vendors will be coming with systems to support digital twin activity and IoT data storage. We have a new kid in the game – IoT data. The fight for IoT data is in the future. Just my thoughts…
Want to learn more about PLM? Check out my new PLM Book website.
Disclaimer: I’m co-founder and CEO of OpenBOM developing cloud based bill of materials and inventory management tool for manufacturing companies, hardware startups and supply chain. My opinion can be unintentionally biased.