Since 2010, I've been part of an organization called the Smart Manufacturing Leadership Coalition. SMLC is working to define an overarching operating system for manufacturing. Why do we need to do that? Can't we just use what we've got? It isn't that simple. Among the biggest issues manufacturing companies say they are facing are integration of data across their supply chains and workforce development to replace retiring workers.
Data integration is extremely hard to do. For example, one of the attendees at the October SMLC meeting noted that she needs to get particle size distribution and elemental concentration and moisture for her input feedstock, and what she gets is a paper (or electronic copy of a paper) certification. It isn't the same thing, so the company has to have wide acceptance brackets for its feedstocks, which leads directly to non-optimal performance. It's sometimes possible to build brittle application program interfaces (APIs), but as another participant said, "They last as long as the people who wrote them are employed there." It's critical to improve this kind of data transfer because, if it isn't competitive to make the products here in the United States, they will be made elsewhere.
Workforce development is another sore point. It's clear that there's a workforce shortage, especially in the technically literate job area. If we want to keep manufacturing here, we need to have people who want those jobs and who qualify to work in them.
Large companies can afford workforce development and training. They can support integrated supply chains (to the extent their suppliers will let them). But small- and medium-sized companies just can't afford to implement the systems and support structures necessary to use smart manufacturing technologies.
And these technologies already, for the most part, exist. Mark-up languages, data conduits like OPC and fieldbus, databases, low-cost sensors, standards such as ISA88, 95, 99 and 06, and IEC standards such as 61499 all exist. They are all tools, but not every plant that could use them does.
Can you integrate all your systems? Sure. Just buy everything from a single vendor. But what if you don't want to do that? Worse yet, what if you have to start telling your suppliers that they also have to use the same vendor you do. What do you think their answer will be? Likely it will be, "Not only no, but absolutely not!"
So what to do? Think of how a computer operating system works. One of the reasons that modern enterprises work is the predominance of Windows (and to a lesser extent Linux and Apple) as an operating system. Wouldn't it be easier to interconnect supply chains in an open manner using best-of-breed products if there were an overarching operating system for manufacturing? Of course, it would. But how do you get that system?
SMLC is not looking to replace current control systems or even current ERP systems. Rather, the system to be developed is going to be designed to make interconnection and interoperation easier, or even possible.
Even though SMLC is really still in its infancy, some really interesting ideas have already come forward. What would happen if you used, instead of cycle time or response rate, the concept of workflow? Suppose you could spread the cost of this workflow-based programming across a number of companies by using Software as a Service (SaaS) and cloud computing? Now small companies could afford to use the tools, too. We'll be talking about this revolutionary concept more in the future.
SMLC will help define smart manufacturing for the future. If you're interested in the future of manufacturing, come join in.