We are all familiar with having to enter a password or e-mail twice, to ensure we have done it correctly.  A custom book binder has taken this technique a step further.  In their application, it was critical that the author’s name be entered correctly (go figure), and sometimes operators would jump the gun and enter what they assumed was the correct spelling of the name.  For instance, Mortenson vs. Mortensen.

They used a method to force the entry operator to really look at the author’s name, on the order form.  The name would be entered normally, then the computer asked that they do it again backwards.

Name:          Samuel Hoskins

Verification:  leumas sniksoh

If the verification didn’t match, an error was displayed.  This was further mistake-proofed by programming the software to always capitalize the first letter of each name.

This is a nice simple error proof method.

That’s the headline, all over the news today. According to ABC news:

“What should have been a blessed time for actor Dennis Quaid and his wife, Kimberly Buffington, turned into a time of anguish and anxiety, after their newborn twins nearly died from an accidental overdose of a blood-thinning drug.

“Zoe Grace and Thomas Boone received a massive overdose of the blood-thinning drug Heparin” used to keep IV catheters from clotting” some time after their Nov. 12 birth at Cedars-Sinai Medical Center in Los Angeles. The incident was first reported by celebrity Web site TMZ.

aerobics

“While not mentioning the Quaids specifically by name, the hospital released a statement that confirmed that three of its patients had received 1,000 times the prescribed Heparin. Instead of 10 units per millimeter, the patients received 10,000 units.

“This was a preventable error, involving a failure to follow our standard policies and procedures,” the hospital said. “Although it appears at this point that there was no harm to any patient, we take this situation very seriously.” “

As I read the various stories about the accident, I was struck by statements like “This was a preventable error, involving a failure to follow our standard policies and procedures…”. The problem with that is, people tend to rely on policies and procedure rather than on real devices.

Yes, if “someone” was being more careful the accident would not have happened. However, if you were to Mistake-Proof the procedure, the hospital would have something in place that would prevent the problem from happening, even if “someone” forgot. I see others in the healthcare community are thinking the same way as I am. From one of the news sites’ public comments section :

“This has happened repeatedly WITH THIS DRUG. Couldn’t the pharmaceutical companies stop packaging the “heparin flush” in vials that are VIRTUALLY IDENTICAL to the higher, therapeutic dosages of heparin? Although, nurses are responsible for checking labels, it shouldn’t take a rocket scientist to see that this is an accident waiting to happen over and over again.”

The person making that comment is on the right track, it would be very interesting to do a Mistake-Proofing event for this issue. I am certain the hospital could come up with a reliable M-P device in a matter of hours, and I think the commentator has a good suggestion by looking at the packaging to make sure that similar medicines look different.

Mistake-Proofing strongly applies to healthcare, just as it does to manufacturing. People make the same errors; they forget, they misidentify, they misinterpret, etc. It is our job, as successful Mistake-Proofers to continue to find real physical devices to keep the root cause error from happening.

When you really think about it, Mistake-Proofing devices are all around us. They are so common that we don’t think of them as M-P devices. Remember, a Mistake-Proofing device is put into place so that it is impossible for the error to occur. Eliminate the error and you eliminate the resulting defect.

Look at all the M-P devices on a modern automobile! I was looking over my car and tried to find the most obvious ones and here is my certainly incomplete list. Please e-mail a comment and let me know if you think of more.

Convenience:

• • Bell sounds if you leave the key in the ignition and you open the door. This tries to keep you from locking the keys inside!
  • A sunroof usually has two sliding components; the glass and the headliner modesty panel. If they were independent of each other it would be possible to close the modesty panel and forget to close the sliding glass panel. This is solved with a simple sliding stop which prevents the modesty panel from closing if the glass is open.

Safety - there are tons of M-P devices in a car!

• • Foot must be on the brake in order to shift from “park” to “drive”.
  • White back-up lights illuminate when car is put into reverse. Sure, back-up lights have been used for years and years, but they really are a M-P device!
  • Cruise control disengages if you step on the brake (or step on the clutch).
  • Anti-skid brakes. No matter how hard the driver tries to slam on the brakes, the brakes will not lock up and will remain at their most effective braking.
  • Seatbelt bell; reminds you to click-it.
  • Child-proof door locks for the back seat.
  • Electric window locks so the kids can’t play with them, either.
  • Airbags are a Mistake-Proofing device.

More:

• • The gas cap may have a tether so you can’t lose it.
  • The gas filler also has a restrictor that prevents you from trying to fill it with diesel fuel.
  • The gas gap also has a torque limiting ratchet, which prevents you from tightening it to the point where you can’t unscrew it.
  • Some new cars have an interlock so you can’t open the trunk if the trunk isn’t in the “park” position.

Mistake-Proofing devices really are all around us and they were created to prevent a bad situation. This can be easily done in the workplace, but first we need to make up our mind to make it happen. After that, it’s all down hill.

A manufacturing company had received a couple of customer complaints regarding a product. This particular product used a colored plastic connector to designate certain characteristics about the item and how it should be used by the end user. The color gives a visual clue to the customer for product identification.

The plastic connectors were made by a molding vendor and shipped to ABC Company in totes. The operator would dump the parts into a vibratory feed bowl and they would be assembled into the product on an automated machine. On very rare occasions, a wrong color part made it into the product and resulted in a customer complaint. The company goal is zero defects.

A Mistake-Proofing event was held. On examination of the problem, we found the wrong color part could be introduced almost anywhere in the value stream. The part could have been introduced at the vendor or it might have been left over in one of the totes. A material handler may have been involved or perhaps the products assembly machine wasn’t cleaned out during changeovers.

Since there were so many opportunities where the mix-up could occur, we really couldn’t pinpoint a root cause. And you know what? The true root cause of this defect really didn’t matter! We just needed to make sure it didn’t happen again.

We needed to focus on developing a Mistake-Proofing device that would not allow the problem to occur, no matter what the root cause.

We were able to locate some color sensors from Turck. These teachable sensors look for a certain color as the parts pass by. Also, the machine’s PLC was programmed to modify the job set-up screens. Now during the changeover sequence, the operator tells the automated assembly machine which color part to use; yellow, red, blue, etc. In operation, the machine looks for red, red, red, red. If a green block shows up, the machine stops and alerts the machine operator to remove the wrong color connector.

This way, they will never again have the wrong color part used in the assembly. In Mistake-Proofing, it can often be more productive to spend time on the preventative action rather than on root cause analysis.

A lot has been written about The Visual Workplace. To me, visuals are the informational components that supports KISS (Keep It Simple S****d). Meaning, use the visuals so someone can tell at a glance if the proper condition is being satisfied.

One of the simplest visuals is a properly marked pressure gauge. Suppose you have a critical condition that is required for the process and suppose you have to maintain a certain pressure range. At the worst end of the spectrum, we might expect an operator to monitor the gauge and if the gauge shows a problem, to take action.

So, we are asking a person to check this gauge every once in a while to see if everything is okay. This sounds like we are setting the operator up for failure. What if the pressure changes? How is our operator expected to notice it? Look at it every five minutes? Ten minutes? Every half hour?

And how about when she looks at it? What should she look for? 60 psi? 65 psi? And what if the gauge is hard to read?

In this situation we can pretty much be assured that the gauge will go unobserved and even if she were to look at it, some interpretation would be probably be needed. How can we make this situation better?

First, I would modify the gauge. Establish what the operating ranges are and then mark the gauge with bright colors. This can easily be done with colored tape on the gauge, or if you want to get fancy, many gauge manufacturers cans customize the gauge for you.

 

To go one better, you could install a pressure switch in the line to be monitored. Keep in mind we are KISS and KIC (Keeping It Cheap). The switch could be hooked up to lights so the operator could tell at a glance if the satisfactory condition was met.

The next level of Mistake Proofing could be to connect the pressure switch in a way that an incorrect pressure condition would cause the machine to stop. This “benign failure would prevent any bad parts from being produced.

Let me show you another example of an effective visual. A polyethylene extrusion and printing company had a very costly defect. Multicolored printing was being applied onto plastic film on a large high-speed high-volume machine that is the size of a boxcar. After each color is applied, the film runs through an oven. Another color is applied, back through the oven, and so on. The heat is provided by steam fed from boilers in another building.

In this case, the boiler went down and resulted in a loss of heat in the printing ovens. The operator didn’t notice the loss of heat until a substantial quantity of product was run. That entire product batch had to be scrapped.

There was a temperature gauge on the printing machine, but there wasn’t anything in the system that obligated the operator to check it. So naturally, when the steam supply problem occurred it went unnoticed.

A Mistake-Proofing event was held and the solution was elegantly simple. The plant engineering department installed a pressure switch in the steam line AT THE PRINTING MACHINE and connected it to a large yellow rotating light. Previous to this event, a warning light had been located outside the steam plant with the hope that a maintenance mechanic would notice it.

Now if pressure is lost, the printing operator is alerted well before the temperature drops in the ovens. This gives him enough time to shut the machine down and prevent making all that waste.

Try and make your visuals so they can convey abnormal conditions, at a glance.

A persistent defect in assembly areas is that of missing components. Certain parts may be assembled and missing one or more components. Sometimes this is caught in a downstream process or inspection station, but sometimes that part can make it all the way to the customer!

First, a walk out to the shop floor to observe the process is in order. Take a close look at the workstation and ask yourself if we are setting the operator up for success or for failure?

One of the greatest weapons against missing components is simply using One-Piece-Flow. OPF means the operator picks up an assembly and performs all of the operations at the work station, from start to finish. Making piles of assemblies drastically promotes the ability to miss a part.

Let’s say our operator is assembling a widget and performs three simple operations before passing the part to the next work station…

Todd likes to take 10 parts at a time and set them on the work station. Todd picks up the first assembly and performs operation A on each part. Then he does operation B ten times, then C, etc. When Todd is finished, he picks to the pile of 10 parts and pushes them over to the next work station… Todd thinks this the best way and he thinks it is faster.

But there is trouble in River City. Let’s say that Todd picked up his usual 10 units and completed operation A on all the pieces. Then he starts into operation B but only completes nine pieces when the supervisor comes by and they chat about Todd’s upcoming vacation to Las Vegas… They talk about five minutes, then the supervisor leaves.

Todd picks up the whole pile of 10 parts and performs operation C on all of them. He doesn’t notice the single unit missing operation B and passes the 10 units downstream. A defect is born.

Let’s look at the same workstation with Doris as the operator. Doris received training in OPF and understands how powerful it is.

When Doris makes a part, she picks up one assembly and performs operations A, B, and C before she sets the unit back down. Then she picks up the second assembly and repeats the process.

When distractions, such as a wandering supervisor, come around Doris’ workstation she has a much better change of not skipping an operation because she has only one part at the work station! Doris has much better control what is happening. She also has less WIP, and if there is a problem found with some of her equipment, the chances are she will catch it sooner than if she was working with big batches of WIP.

One-Piece Flow is a very powerful tool and its principles should be applied all over the shop floor.

The traditional approaches to handling human error seem to be based on “We need to try harder”.

The first step in successful Mistake-Proofing is to acknowledge that everyone makes mistakes. Once you get over that hurdle, we simply need to ask ourselves how to keep our system functioning even though an error may occur.

We can prioritize how well we accomplish this. We can say that a successful Mistake-Proofing device should either:

1. Prevent an error from occurring in the first place.
2. If we can’t prevent the error, then let’s detect the error.
3. If we can’t detect the error, let’s at least detect the resultant defect and stop the process and correct the situation.