Should I Automate My Manufacturing Process? If So, How Much Should I Automate It?
, by Jim Ryan, 6 min reading time
Tags
Okay, longish post, for those of you trying to decide whether to automate your manufacturing process and, if the answer is yes, how much to automate it. Lots of links to smart people's thoughts on this below. They shed light on it.
Step back for a moment. If in 2010 you thought that the trough of disillusionment on the Gartner Hype Cycle for autonomous cars would be very deep and would arrive by 2020 or so, you were right. You may have had special insight into how hard the problem was. Or you may simply have been someone who was a bit more jaded about "shiny" new technology and less susceptible to their charms than those who were very excited about an immanent arrival of UGVs. (Perhaps you still listened to vinyl because it sounded better, and so forth.) Whatever the reason, if that was you in 2010, then you simply weren't hypnotized by the idea of cars without either steering wheels or driver's seats.
How do you tell whether you are susceptible to being hypnotized by a "shiny" technological solution? Here are some warning signs:
- You take great delight in the cutting-edge nature of the solution
- You find yourself calling critics of the solution "naysayers," "backwards thinkers," or "stuck in the past."
- You assume that since the solution is high tech, it will offer workarounds for any pain points entailed by adopting it instead of previous technology. These are the pain points noted by the aforementioned critics. You wave them off without really taking them seriously. Why? Because shiny!
- You fear being looked at as old fashioned or irrationally slow to adopt good new technology.
- In general, love high tech and can't wait for a Star Trek-type utopia in which there are unmanned air-and-ground taxis that show up wherever and whenever you want, dark factories, delicious meals that come out of a wall for free whenever you want, and interstellar warp drive for inexpensive vacation trips.
Here's the point. The fact that a solution is amazingly high tech or clever, cutting edge, or even "bleeding edge" does not make it the right fit for a certain application. Yet, it is still amazing, which can cause it to be selected. I have seen this sort of mistake cost lots of money and time.
Let's take a hackneyed example: wifi vs ethernet cables. Remember ethernet cables? Perhaps your grandparents used them. There are still some true believers. They still have advantages in some applications. They may offer more security, reliability, and speed. They are not as subject to man-in-the-middle attacks as wifi is.
So, what about factory automation? Obviously there are many cases being made today for increasing automation, industrial revolution 4.0, IoT (internet of things), smart factories, and even dark factories. And there are plenty of articles about the trends driving the move to automation. There is no question that there are good reasons to automate.
Let's take a look at some of the arguments that are strongly in support of greater automation. Kip Hanson has an article at SME that, after beginning with a refreshingly blunt paragraph on the horrible labor shortage the U.S. is facing, states, "To say that success in manufacturing is now an automate-or-die proposition might be hyperbolic—but only slightly." After discussing (again with refreshing bluntness) the expensive but possible high-ROI path toward greater automation, Hanson says that "shops that master automation open the door to the grand prize of manufacturing: lights-out operation, which promises to double or even triple output without additional machine tools, floor space, or the skilled employees that no one can find anyway." Perhaps, but the key word is "promise". He implies, following Courtney Ortner of Absolute Machine Tools, Inc., that "everyone" who "takes the leap of faith" "is a success story" and makes no mention of the possibility that full automation may not be the optimal solution for every floor. Everyone? Really? And managerial decisions shouldn't amount to leaps of faith.
Another example of the hard push for automation is the verbiage in this article, also at SME, by Hooi Tan: "It's Time to Join the Fourth Industrial Revolution" and failure to do so now will "threaten to leave manufacturers in the dark ages" and, quoting Francisco Betti of WEF, "the future belongs to those companies willing...." and "lighthouses are illuminating the future of manufacturing...." While Tan's article is helpful, none of this rhetoric is an argument that full automation is optimal or prudent for any factory.
Few people want to lag "behind" technologically. There is psychological pressure to "keep pace" with others in degrees of automation. Here's a video on why the UK is "behind" on robotics and automation adoption. The video states that this is a "problem" but gives no argument why this is a problem. It says that the UK is a world leader in manufacturing but is 24th in the world for automation adoption (measured in terms of automation density per 10,000 factory workers). So, why shouldn't we believe, instead, that the UK is in the sweet spot, automation density-wise, whereas other countries have adopted automation too aggressively? The video doesn't offer a reason.
Perhaps a more judicious and careful approach to the question of whether and how much to automate might be Michael Bell's Welding Automation ROI and Part Quality, which points out that "It’s important to weigh each consideration against the benefits of an automated cell." The considerations he discusses - infrastructure, workforce, and technology - affect the ROI. Other examples of cautious approaches to the question of whether and how much to automate are this one by Steven Calhoun, this one by Liz Salter, this one at Marlin Steel, and even the discussion of semiautomatic welding (a category which includes cobotic welding and other modalities) here by Austin Weber, who points out that "fully automated welding is often not ideal for every application." All of these approaches to the question of whether and how much to automate suggest that you weigh a variety of considerations germane to your particular circumstances. They do not make rhetorical or emotional appeals. That's prudence.
Finally, we should, as always, keep the Gartner Hype Cycle in mind, for instance here, where Alex Peron surmises that for industrial IoT (internet of things) we are at the peak of inflated expectations on the cycle. (The article is undated.) There is always a trough of disillusionment in the cycle. It may be deep and it may be shallow, but it comes. Factory automation is no exception.
What about weld cells? Would anyone ever consider a wifi replacement for the ethernet cable? After all, wifi is amazing and it isn't vulnerable to welding slag. No, wifi is vulnerable to electromagnetic interference while ethernet cables are not. But what about the weld slag? You can protect your ethernet cables from weld slag! Simply cover them with Weld Dynamix's Activflex Cable Protection - in seconds and without disconnecting your cables at either end! Weld Dynamix weld cell products are ingenious technology. Are they "shiny"? I don't know, but they are definitely the right solution to the endless grind of replacing sensors and cables in your weld cells.