Is there a value in PLM data standards?

Is there a value in PLM data standards?

Standards are important. They play an important role in our life. There are a different type of standards. There are government, industry, city and professional standards. There are standards for environmental impact (filtering of exhaust fumes), food (labeling), transportation (rail track), information (ISBN), accounting standards (GAAP). There are many professional standards outlining processes, practices, and vocabulary. An example of a professional standard can be a set of rules that technicians need to be certified to perform a specific task. But another example of the standard can be a vocabulary used to describe a product, its lifecycle stages, and other parameters.

I’m getting a spike of questions and opinions in recent discussions on the topic of standards in PLM. In case you missed some of my recent articles, here are some links and comments.

Do we need a standard like PLCS to build a digital thread?

The discussion about what is special in PLCS that will enable digital thread creation.

Initially Eurostep targeted extended enterprise collaboration with ShareAspace/PLCS. It is an obvious case where a neutral, standards based hub would play a role. All participants could then keep their internal processes and tools and ShareAspace/PLCS is rich enough to allow for the sharing of data from Systems Engineering, PDM, PLM, ERP, MRO etc. Good for collaboration to break out of silos and keep the digital thread also in collaboration.

A comment above is from Hakan Karden, CEO or Eurostep. His main point – use standards to build a hub. Otherwise, you will be stuck in silos.

Hakan published another article – The need for standards in Smart Manufacturing Systems. Here is an interesting passage:

Production lifecycle data management standards define general models of data integration, sharing, exchange, and hand-over for lifecycle support of production facilities. A selected set of important production lifecycle data management standards are shown in Table 9. A study of information modeling [53] found that ISO 10303 AP 239 (PLCS) has the most potential to model a manufacturing system for its lifecycle.

As it sounds from this passage, the key role of standards is to exchange and share data. I also captured the following picture from the NIST report. It shows the positioning of PLCS, PLM XML and LOTAR supporting the lifecycle.

In a very unexpected way, the conversation in another article – 3 Reasons For Not Growing Existing PDM Into The Full PLM system found its way to standards again. As part of the question of how to separate a monolithic PLM system into enterprise services, the comments were made about the usefulness (or not) of standards. Here are a few passages:

Horia Bradau: True, standards would be necessary for breaking the monolith into services. There are attempts for data and process standardization (OAGiS, ISA95, STEP etc) but they are usually limited in scope, or niche specific, or competing, or too complex, or obsolete, or…

Jos Voskuil: if we move to more granular Enterprise Services, we also need to align more on standards to avoid massive point-to-point solutions

Oleg Shilovitsky: what’s wrong with existing web standards to integrate enterprise services?
Jos Voskuil: It is about data standards – semantics to reduce complexity

This is not the first discussion about standards, but this is my PLM data standards epiphany moment. It all comes down to a set of attributes, vocabulary, and formats of data representation. Ultimately, if I have two systems, I need to match the way the same data is stored in both systems otherwise, data sharing and exchange will be a data management nightmare. And each time I’d like to add or change data representation in one of the systems, I will have to ensure that this step is coordinated across multiple systems.

Here is a conflict…

As an industrial company, product development activity is different and hardly can be presented in a standard way. Therefore, companies are demanding PLM flexibility as an ultimate requirement of every PLM project success these days. Even more, companies are demanding flexibility and configuration (no customization) to be supported by PLM software. Therefore, we can see flexible PLM platforms are on the rise. Unless… some a critical business process is dependent on a standard (eg. a supplier cannot get an order from OEM without supporting a specific standard).

As a software vendor, the last thing needed is to coordinate data management development with other vendors. It will make system dependencies nightmare and will bring a total collapse preventing systems from future development, extension and expanding. Unless… there is a requirement from a customer to support a standard, otherwise the software won’t be purchased.

The middle ground is to define a set of data standards for data exchange and sharing. As much as the idea makes sense, it is expensive, complex, competitive, specific for some companies and implementations and makes no sense for others. In other words, without funding, it won’t work.

Is there a way to find a solution to make progress or we are in a stalemate position? As much as I like standards and believe in the value of standard development, in reality it looks more like a stalemate. There is a potential for industry data standards driven by large OEMs and their supply networks. At the same time, the growing demand for connectivity in outsourcing and contract manufacturing, a growing number of online services and complexity makes these standard activities way too complex. Meantime, future development on cloud/SaaS solutions and web services can simplify the integration, but at the same time to create a new wave of integration semantic challenges.

What is my conclusion? There is a demand for connectivity, integration and data sharing. From early design and requirements to capturing information about customer activities via IoT and connected devices, there are too many questions is how to pass data back and forth and also to make sense of the data. There is a need to rethink standards and make it part of the process that can create value. If the industry will be able to do so, there is a chance that this value creation process will support data standards. Without such value creations, all PLM data standards will be dead on arrival. Just my thoughts…

Best, Oleg

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

Share This Post

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

  • I dealt with this issue, in a small way, on a previous project…

    This was for a Detroit-based tooling supplier to the US 3 Auto makers. Each of the automakers required CAD and meta-data to be uploaded into the OEM’s independent “PLM” systems. Within the supplier’s data model we were maintaining separate data structures for each OEM, but realized that in reality each OEM required nearly identical data. At one point we brought the OEM’s and the supplier into the Siemens offices (I was with SPLM at the time) and brought it up.

    I had a slide in which I showed ice cubes in a glass of water representing the OEM’s and suppliers. The ice cubes are distinct, but they each impact the others. A colleague mentioned that “OEM’s compete on what they sell, but cooperate on what they buy”. In fact, this was already happening in the area of tooling part design, using something called NAAMS parts. The OEM’s and suppliers (working through USCAR) had already come to the conclusion that common-izing on tooling part design was in everyone’s best interest; there are now thousands of common NAMMS part designs; the tooling designer pulls the part from a NAAMS catalog, and local Tier-2 suppliers were comfortable in their manufacture. The point in the meeting was to also common-ize on meta-data.

    My point here is that the solution did not lie in a more complicated data mechanism; it was found through interpersonal negotiation amongst OEM’s and suppliers, and emphasizing common value. This difficult data problem had an easy human solution.

    —————————–
    (My pet peeve…)
    IMO – “PLM” is misused here; the product lifecycle encompasses four phases, Create, Build, Service, and Dispose, and there is no one database (or vendor) who encompasses the entire scope. The self-called PLM databases are limited to Create; MES is the domain of Manufacture, and MRO the domain of Service. (Disposal databases haven’t really appeared as yet.)

    There are 1-to-many relationships between each phase, and to the point of Oleg’s article some sort of commonality will be required to implement Digital Twins and Threads. (The physical instances of aircraft and homes last for many decades. How will we manage their digital instances?)

    But crypto-currencies have the same problem; I’m not the first to see a role for blockchain.

    Pat

  • beyondplm

    Patrick,

    thanks for sharing the story! It is straight to the point to demonstrate that standards can only be made for usefulness and needs. In your story OEMs were interested to have it done. So, as soon as people problem was solved, the agreement about data is much easier today.

    You raised an interesting point about 4 stages. There are few vendors today that are pushing digital thread stories and how to connect everyone with a single database. It sounds like a nice story, but reminded me “one version of truth” story when PLM just started 20 years ago.

    Oleg