Tag Archives: custom

From Automation to Fabrication?

Here’s my simplified idea of what factories do: they make a whole lot of copies of one thing really cheap.

The “really cheap” part only comes with scale. Factories don’t make “a few” anything. They’re dependent on a mass market economy. Things need to be cheap enough for the mass market to buy them, but they also need to change constantly. As humans, we have an appetite for novelty. As the speed of innovation increases, factories spend more time retooling as a result.

The result is more demand for flexibility in automation. Just look at the rise in Flexible Automation and, more recently, Robotic Machining.

Where does this trend point? We’ve already seen low cost small scale fabrication machines popping up, like MakerBot and CandyFab. These are specialized versions of 3D Printers. Digital sculptors can design their sculpture in software, press print, and voila, the machine prints a copy of their object.

Now imagine a machine that’s part 3D Printer, 6-axis robot, laser cutter/etcher, and circuit board fabricator all-in-one. Imagine our little machine has conveyors feeding it stock parts from a warehouse out back, just waiting for someone to download a new design.

That kind of “fabrication” machine would be a designer’s dream. In fact, I don’t think our current pool of designers could keep up with demand. Everyone could take part in design and expression.

I don’t see any reason why this fictional machine is actually beyond our technological capabilities. It’s certainly very expensive to develop (I’m going to take a stab and say it’s roughly a complex as building an auto plant), but once you’ve made one, we have the capability to make many more.

For more ideas, take a look at what MIT’s Fab Lab is working on.

Off-the-Shelf or Custom Automation?

If you’re like me, you’re a fan of customizing:

…and certainly in the automation industry you see a lot of custom control solutions. In fact there’s always been this long-running debate over the value of custom solutions vs. the value of off-the-shelf “black box” products.

I’ve noticed this rule: the closer you get to the production line, the more custom things you’ll see. Just look at the two ends of this extreme: production lines are almost always run by PLCs with custom logic written specifically for that one line, but the accounting system is almost always an off-the shelf product.

There’s a good reason for this. Accounting methodologies are supposed to be standardized across all companies. Businesses don’t claim that their value proposition is their unique accounting system (unless you’re talking about Enron, I suppose). Automation, however, is frequently part of your business process, and business processes are the fundamental business proposition of a company. Fedex should definitely have a custom logistical system, Amazon needs to have custom order fulfillment, and Google actually manufactures their own servers. These systems are part of their core business strengths.

So when should a company be buying off-the-shelf automation solutions? I say it’s any time that “good enough” is all you need. You have to sit down and decide how you’re going to differentiate yourself from the competition in the mind of your customers, and then you have to focus as much energy as possible on achieving that differentiation. Everything else needs to be “good enough”. Everything else is a cost centre.

If you follow that through logically, it means you should also seek to “commoditize” everything in the “everything else” category. That bears repeating: if it’s not a core differentiator for your company, you will benefit if it becomes a commodity. That means if you have any intellectual property sitting there in a non-critical asset, you should look for ways to disseminate that to the greater community. This is particularly important if it helps the industry catch up to a leading competitor.

There are lots of market differentiators that can depend on your automation: price, distribution, and quality all come to mind. On the other hand there are other market differentiators that don’t really depend on your automation, like customer service or user-friendly product designs. Ask yourself what category your company fits in, and then you’ll know whether custom automation makes sense for you.