Metal fabrication and the Internet of things

First, a disclaimer: If decades from now people find this blog in the deep recesses of some digital archive, they’ll probably laugh at just how wrong I am. If they aren’t laughing, then we’ll be living in a very different world.

Futurists have opined away about what they call the “Internet of things,” the fact that every device potentially could be connected. We won’t have just smart phones, smart tablets, smart computers (or whatever we call them years from now), and smart cars. Nearly everything we own will be smart. They’ll be able to diagnose themselves, order parts for themselves, or notify the owner that they’re on their last leg and it’s time to buy a replacement.

In a recently posted video, economic theorist and author Jeremy Rifkin grouped the Internet of things into three categories: energy, communication, and logistics. Even now, billions of sensors are on machines, roads, appliances, warehouse replenishment points, and everywhere on the factory floor, front and back office, and in retail stores.

“We have sensors all across the system feeding big data back,” he said. “There are 13 billion sensors now, and IBM says by 2020 we’ll have 30 billion sensors connecting every ‘thing’ with every being. And by 2030, the most recent forecast, we’ll have 100 trillion sensors, connecting all of us in one vast, lateral neural network made up of three operating engines: a communication Internet converging with an energy and logistics Internet.”

Rifkin goes on to say that the physical world of manufactured goods will follow what has happened with the information world and distribution of ideas. Today the Internet has allowed anyone to create content, often at near zero marginal cost. Got a camera and a laptop? You can post a blog or a video. He said this eventually may migrate to the physical world where, thanks to 3-D printing, every consumer will become what he calls a “prosumer,” able to both produce and consume, all at near zero marginal cost.

I don’t agree with his argument entirely, because it assumes technology will make it easy for anyone to engineer a good product. Sure, quality and safety standards conceivably could be built into the creation platform, be it some future version of CAD or anything else. But quality and safety are just baselines. The Internet’s just a platform, and it can’t by itself make everyone good creators of content. I suppose the same would apply if every one of us were given the power to manufacture things in the home. So I don’t buy the idea of a third industrial revolution where everyone can make things and ideas flow freely across a so-called “big data” platform.

But I do buy the idea that we may be at the dawn of big change, and it relates to those 100 trillion sensors and the supply chain. If you compare value-added time to nonvalue-added time in any value stream, you’ll probably find that the value-added time—the time a part spends being laser cut, bent, welded, painted, and assembled—is very small, in the low single-percentage range. Spread that out to the entire supply chain, from mining the ore to distributing the steel and all the way to Walmart, and you discover an immense amount of wasted time. And much of it—the warehousing, the bulk orders, the inventory buildup, and so on— is there because we can't precisely predict what or when we'll buy.

What if those trillions of sensors gathered so much data that we could predict economic behavior? If we have predictable demand, the entire supply chain could produce at the pace of customer demand, no faster and no slower. Humans are difficult creatures to predict, but if that prediction were just a little better, it may lead to some dramatic changes. If an entire supply chain eliminate a good portion of that nonvalue-added time, then costs would plummet; probably not to the “zero marginal cost” that Rifkin talks about, but closer to it.

Imagine a future in which you have trillions of sensors able to predict customer demand throughout the supply chain, monitor machine conditions to prevent unplanned downtime; and a future with machine tool technology and manufacturing methodologies allowing shops to change over between jobs within seconds (some of this technology is already here), all synced with customer demands. In short, imagine a future in which the majority of activities in the supply chain add value.

When you look at examples from lean manufacturing, quick-response manufacturing, the theory of constraints, or any other improvement methodology, you see how long jobs spend between processes just waiting to be cut, machined, bent, shipped, and sold. If the Internet of things can eliminate a significant portion of that waste, we may be in store for a very different future.

Or not, and if that’s the case, at least I’ve given future readers a good laugh.