Decentralized PLM?

Decentralized PLM?

PDM (or PLM) was born as idea of centralization of data and processes under single roof (actually database).  Back 30-40 years the idea of central data storage combined with data and file control was very much appealing to large organization. Enterprise PDM and later PLM systems were born. Major PLM vendors today are an outcome of this history. The world seems to be a good place for these large PLM vendors, but something is telling me that crisis in PLM industry is inevitable. You can ask me why, because in my view, this is one of the most important questions in PLM industry today.

Here is the thing… Major PLM vendors developed walled garden of systems. Each one is beautiful and has everything you might need. But to get the value of this system (often called PLM platform) you need to commit to it in full. Nothing wrong- companies can still rely on a single vendor. Other enterprise systems – ERP, CRMs can give you lot of examples where it is actually works. The latest trend of Innovation platform development is new packaging to the idea of siloed PLM systems.

But the main challenge I can see for a future of walled garden of PLM platforms is related to the future of distributed work in manufacturing. Single company is not capable to manufacture products. Think about large automotive OEM. Such manufacturing company is responsible for approximately 20-25% of all work to build an automobile. Rest is build for suppliers and their suppliers. And this is where things are getting really interesting. Try to map an advanced supply chain network onto walled garden of PLM platforms and if you’re CIO or responsible for engineering and manufacturing IT you can easy lose your sleep. Because the reality is a distributed network on connected companies working together. This is not good for PLM walled garden. Outside of large OEMs, the situation is even more complex. The jungle of relationships, contractors, suppliers is making manufacturing life distributed and complex. Single centralized system cannot provide support for manufacturing companies. Things get stuck and each time it happens, companies are moving to emails and spreadsheets as a lifesaver option.

Earlier today, my attention was caught by Chris Dixon article – Why decentralization matters. Internet is a greatest example of distributed system. In his blog, Chris is suggesting some interesting observations and comparison about internet evolution between first era of distribution, centralizing of GAFA and future steps. Here is my favorite passage:

The internet is the ultimate software-based network, consisting of a relatively simple core layer connecting billions of fully programmable computers at the edge. Software is simply the encoding of human thought, and as such has an almost unbounded design space. Computers connected to the internet are, by and large, free to run whatever software their owners choose. Whatever can be dreamt up, with the right set of incentives, can quickly propagate across the internet. Internet architecture is where technical creativity and incentive design intersect.

Decentralized networks can win the third era of the internet for the same reason they won the first era: by winning the hearts and minds of entrepreneurs and developers.

An illustrative analogy is the rivalry in the 2000s between Wikipedia and its centralized competitors like Encarta. If you compared the two products in the early 2000s, Encarta was a far better product, with better topic coverage and higher accuracy. But Wikipedia improved at a much faster rate, because it had an active community of volunteer contributors who were attracted to its decentralized, community-governed ethos. By 2005, Wikipedia was the most popular reference site on the internet. Encarta was shut down in 2009.

The lesson is that when you compare centralized and decentralized systems you need to consider them dynamically, as processes, instead of statically, as rigid products. Centralized systems often start out fully baked, but only get better at the rate at which employees at the sponsoring company improve them. Decentralized systems start out half-baked but, under the right conditions, grow exponentially as they attract new contributors.

What does it mean for future of PLM developers? Existing PLM platforms are very much static. They are developer with a very slow speed. There are many reasons for them – complexity, resources constraint, but most importantly the ability of manufacturing companies to consume these systems. After investment in a specific version of PLM software, large manufacturing OEM is thinking how not to get involved into future development at least 5 years. But then the question of solving business problems comes. And large PLM vendors can barely provide solution to solve these problems.

Opposite to this static state, decentralized distributed PLM system are always live. Like Facebook – they have no choice, but to improve the service, connect more systems and build intelligence based on communication with many users. Now, apply it to manufacturing companies and typical examples of distribution in manufacturing. Decentralized systems are the only way to help manufacturing companies to communication and coordinate their activities.

What is my conclusion? Decentralization is the next step in maturity of PLM systems. Moving from large central system to distributed decentralization can provide a unique approach on connecting data and orchestration processes between OEMs, contractors and suppliers. This is a new world of manufacturing as I can see it. Just my thoughts…

Best, Oleg

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.



Share This Post

  • Pingback: GHZ Partners | IT Services, Software()

  • David

    The trick will be in identifying what processes get decentralised and what stay centralised. IMHO the creation of a coordination 3D model (talking big things here like an oilrig) probably is going to work better in the short term (as long as there is no universal XMLish format for 3D data) having several locations and partners creating a single coordinated clash free model will be far easier on one system, perhaps the data will be decentralised but it will be created in one format. MCAD data etc. may come from separate CAD systems which effectively feed into the master model. Data creation and document creation and management on the other hand can be highly decentralised as the lowest common denominator (pdf or XML metadata) is more easy to handle in a consistent way. Managing the standards and configuration that allow mapping and transformation of data into what looks like a central data store to the user (single source of truth) probably also needs to be centralised and so on.

  • beyondplm

    David, You raised an interesting question – can centralized and decentralized systems coexist together? I have a proposal. Data is linked (and centralized), but process is decentralized. However, the important element in my view is how to “link” real data and stitch it into a “centralized linked space”?