As oft intimated on this blog, product data in all its forms is located and sourced in different places in manufacturing companies. Accessing product data is, moreover, confounded by how different systems deploy different methodologies (or languages) for how they keep track of it. For example, an item known in engineering as a “Part” may be called a “Component” in production. Or a AutoCAD drawing data may have data derived from Solidworks assemblies. What I’m trying to say here is that the mixture of different vocabularies or vernacular used in a typical manufacturing company makes it difficult for different systems to have a macro understanding of how product data located in different places is connected together. This is something that we at Inforbix understand very well. Let me explain.
I had a chance to watch Tim Berners-Lee talk at the gov2.0 expo in Washington, DC last year. He introduced the notion of Linked Data. What he said resonated well with our own ideas at Inforbix. Here’s a video clip of that same talk:
Tim Berners-Lee broadly defined linked data as cherry-picking common attributes or languages to identify connections or relationships between disparate sources of data. His central argument is that linking data together ought to be as simple as declaring the equivalence between common properties. He illustrated this with a bag of American potato chips. He noted that in the UK, the equivalent would be labeled simply as a bag of Crisps. However, the main ingredient, potatoes, is common to both countries. That’s a link. The nutritional information on the package is described in a language understood by the USA but not by it’s equivalent in the UK. However, both have identical caloric measures that are in common. That’s a link. By creating a few such links (e.g. another common link is USP codes), identifying the bag of chips (or crisps) becomes easy because linked data doesn’t require that everyone agree on all the terms that define it. Certain things, such as a number on the bag, can be ignored if other more relevant and universal links are established. Tim’s thesis is that data need not be structured top-down to be well organized and accessed. Rather, he posits that a system that links data together tremendously simplifies how data is organized (or not) and how it’s accessed.
So, this is my conclusion: linked data is the future of how product data is organized. Today, data resides in silos in most manufacturing companies and is organized using top-down systems that don’t effectively link all the product data in a company together making it difficult to access. At Inforbix, we are thinking how to harness the power of linking data together. Join us in helping make product data accessible by everyone that needs it. Sign up for our Beta program by following this link. We’ll supply all the chips (or crisps) you can eat.
Best, Oleg
Pingback: se5 window cleaners