Last post, I talked about the lean concept of poka-yoke or “mistake proofing.” The idea of poka-yoke is that even the best worker can make costly or dangerous mistakes when their attention slips even just for a moment. This is certainly true in the world of childproofing, where any parent will tell you that toddlers can get into amazing dangers during the split second your back is turned.
That is why you need to have passive backups, to keep mistakes from happening when your attention is distracted for a second. A good example of this kind of backup is the cabinet lock, which keeps cabinet doors shut when the prying fingers of a toddler go to work on them when you aren’t looking. In theory, a cabinet lock should make it impossible for a toddler to open a cabinet they shouldn’t, whether you are aware of them doing it or not.
I say “in theory” because one of our cabinet locks, a very popular model (in fact, the only model available at any of our local stores), has a crucial flaw that makes its protection much less reliable than it should be. Here’s a picture of the offending lock:
It looks pretty good, right? You loop the U-shaped piece around your cabinet door handles, pass them through the ratcheting crosspiece, and push them together until the doors are secured. The tricky button-pressing motion required to open the lock keeps toddlers out.
Or at least, that’s how it’s supposed to work. In practice, there are four different ways the two pieces can come together, depending on their orientation. Unfortunately, only one of the four orientations will actually engage the ratchet and secure the lock. The other three orientations will not engage the ratchet and the lock can simply be pulled apart without resistance by the busy hands of a toddler.
Even worse, the three incorrect and unsafe orientations look exactly as secure as the correct orientation! You can’t tell just by looking whether you got the one right orientation or one of the three wrong orientations. Now, it is true that you can tell whether you got the right orientation by listening for the ratchet and testing the device carefully after you’ve engaged it. In other words, it’s just a training and experience issue, right?
Wrong. You can’t train mistakes out of existence.
My wife and I are both well aware of the issues, but when you have a hot pot on the stove and a timer beeping and toddlers pulling at your pant legs, it’s easy to forget to listen for a ratchet or give the lock device one more pull to doublecheck it. Even the best-trained, most-motivated workers can make mistakes when they are tired or distracted, even just for a moment. And a moment can be far too long when safety is at stake.
What this situation needs is not just training but effective poka-yoke. It needs to be physically impossible to put the lock pieces together the wrong way. To give credit where it is due, the manufacturers of the cabinet lock in question made an attempt at poka-yoke: The U-shaped piece has a small nub on one side, and the crosspiece has a small notch on one side. If the nub goes into the notch, the pieces are aligned correctly. Problem solved, or so it seems.
Unfortunately, the design of the lock made the holes in the crosspiece too big, and the diameter of the U-shaped piece too small, for the nub to be an effective poka-yoke device. With just a slight effort, the crosspiece can be pushed onto the U-shaped piece whether the nub is correctly oriented with the notch or not. (We know this because we’ve unwittingly done it ourselves many many times.) Which leads to my main point:
Ineffective poka-yoke is worse than no poka-yoke at all.
It should be obvious why: An ineffective safety measure gives you a false sense of security, and that may be even more dangerous than having no safety measures at all. A driver with no seatbelt will drive more carefully than a driver with a seatbelt that looks fine but will actually do nothing at all to protect him in an accident. Overconfidence can be dangerous.
So what can be done? With no other cabinet locks available, we are stuck with this one for now. So we made do. If you look at the photo again, you’ll see some red lines. We drew these on the lock with a marker to make it obvious to us when the device is in a correct orientation and when it is not. These visual reminders help, and qualify as a very weak poka-yoke aid, but a physical change in the device itself to make it impossible to engage it incorrectly would be much better.
As it is, we are still relying on our attention to always notice when the red dots are misaligned, and that means we are vulnerable. Hopefully a better solution will present itself.