Earlier this week, I shared my thoughts about connected PLM and how the organization of a semantic data layer can provide a foundation to build a new approach in data aggregation and communication. For the last two decades, semantic technology was used for a number of projects, and data management initiatives helped to create a data model to unify product data and to facilitate data integration in different industries. In product lifecycle management, semantic technologies were mostly used internally by software vendors and large customers that found the technology of organizing data models useful and capable to help access data. Although such database technology didn’t provide a universal way to model data, I’ve heard about multiple projects in the PLM industry looking at how to leverage semantic technology to model complex data sets, business models, and the way to break and replace existing monolithic PLM systems.
Check my earlier article OpenBOM Semantic Data Modeling Layer for Connected PLM to learn more. If you’re using semantic technological stack in your current projects, I’d love to talk to you. I’d like to thank all my readers that commented on the article online and offline. I appreciate your support and sharing your insight. My favorite comment and the challenge, which was greatly articulated by Matthias Ahrens, PLM Solution Architect at HELLA.
…the challenge for the enterprise (single company or network of collaborative companies) will be… to create an open physical product ontology model that can be across multiple systems. …” If such a common denominator is determined by the involved parties, it could act like a “translating” dictionary resp. common “language” to help to interpret the originals within the different source systems. If such a semantic layer is created in an open standard like RDF / OWL, it will prevent a vendor lock-in for this knowledge. This leads to the capability to move the knowledge easily from one framework to another one, if necessary. A few hypers callers provide RDF / OWL compatible triple stores. As consequence, even the path towards a cloud solution for covering such knowledge keeps left open.
These and other comments made me think about possible options of semantic layer development and how it can be supported by PLM vendors, PLM IT companies, and IT departments of large enterprise industrial companies. Here are some thoughts about what can be done.
What is the semantic data layer and why should we care
The semantic data layer represents the set of schemas and ontologies (RDF and OWL) that can describe the data entities in a form that can be consumed by different systems more effectively than existing data standards. If you never heard about it check it out here – Semantic Web W3C. One of the main reasons is that RDF and OWL has a mechanism that allows recombining the data from multiple data sources as well as to expandable to customize it with additional semantic data elements. Such a model is much better than relational databases with fixed data schemas capable to model an entire product lifecycle including engineering processes, supply chain networks, and multiple enterprise systems. Industrial companies can see significant improvements and create an enterprise transformation process by using semantic data layers into their PLM systems because they don’t have to spend as much time integrating all existing enterprise systems and IT systems.
Open Source PLM – History and Future
Open source was around for more than two decades, yet PLM had very little exposure to open source. Check some of my earlier articles about Open Source and PLM. Recently I shared my thoughts about what will be the next open-source model for PLM. While the jury is still out if open source PLM initiatives can generate enough interest from both PLM vendors and industrial companies, the trend towards the development of open source projects outside of PLM space is strong. The assumption that PLM space is very small has some legs when you think about vendors, but if you think about industrial companies, consulting, and services, the picture can be different. The early belief was that vendors will be triggering open-source initiatives and creating a foundation for open source PLM. However, it didn’t happen. The great attempt was done by Aras Corp and it included a lot of great initiatives. But Aras code was not open and the application was only available for free download and use. The next opportunity round can be triggered by large IT providers and manufacturing companies spending time and effort building their PLM solutions.
How To Create A Common Ontology?
Standards are a great way to unify PLM systems and to provide efficient management of product data during the development process and connected PLM transformation. However, standards are hard and expensive and require a great deal of agreement between multiple industries, vendors, and companies. So, while the rationale to create a common standard is here, from the standpoint of PLM technology, it is not very much achievable and not widely adopted. The value of data standards is the biggest discussion and if the same approach can be used to create a “common ontology”, it can be a failure. An example of some kind of successful ontology building is the schema.org project,t but it was supported by major search vendors. The same cannot be said about PLM vendors. Therefore, an opportunity is to create a bottom-up approach and to make industrial companies and IT invest in this development as a tool that can help them to support manufacturing transformations and information systems. Will it work? The time will show, but it requires some support from software vendors.
How to use Open Source Advantage?
Open source was a powerful initiative and supported the building of multiple huge software products and technologies. How to use it to build common semantic ontologies and support complex business processes for large industrial companies? I can see 2 independent processes that together can contribute a significant value creation and business model to make modern technologies such as RDF / OWL and modern SaaS PLM succeed.
By defining ontologies and semantic models as open-source, manufacturing companies will create a bottom-up process of data organizations. Industrial companies have huge incentives to make it work to enable semantic layers to unify complex data. The initiative should be coming from large IT and global manufacturers. They have all interests to organize an ontologies for the product model, change management, and different aspects of lifecycle management.
Open-source software components
This one can be tricky because it is related to product lifecycle management vendors. Existing PLM systems are closed products. No source code is available. There is an opportunity for vendors to open source for some of their products. Modern PLM business models are moving fast towards subscriptions and therefore provide a ready foundation for open source business.
What is my conclusion?
A business model is one of the hardest things to change both for the industry and in each vendor independently. Semantic technology can become a trigger for a new open-source initiative. A new reality in PLM business with a massive switch of industrial companies and IT providers to software development can reverse the track of “closed PLM solutions”. New PLM architectures can enable better modularity and create an opportunity to switch some of the components and building blocks compatible with the new semantic layer to open source. What process can be a trigger and what vendor would like to take a “first mover” advantage in the future of open PLM systems? This is a great question with no answers yet. Just my thoughts…
Disclaimer: I’m co-founder and CEO of OpenBOM developing a digital network-based platform that manages product data and connects manufacturers, construction companies, and their supply chain networks. My opinion can be unintentionally biased.