In manufacturing facilities and process plants, maintenance errors account for a huge proportion of downtime causes.

In Part 1, we covered ways to generally reduce maintenance errors. In this Part 2, we will cover ways to reduce the specific maintenance errors that occur in your facility.

The single biggest challenge in identifying maintenance errors is getting people to acknowledge them and report them. Management must clearly communicate the expectation is that humans will make mistakes. There may be a little shame in making one, but there is far more shame in hiding it so it can bite again. Thus people won't be punished for the occasional mistake, but they will be punished for a single instance of covering one up.

A caveat applies here. There is a difference between making a mistake and behaving stupidly. It's not just a "mistake" if you put a DMM in current mode and connect it between phase and ground on a 480V feeder. It is a mistake if you normally use a 400:5 current clamp and don’t notice that your helper brought a 200:5 clamp, and your readings are consequently off by half.

Those who ignore history are doomed to repeat it. Those who hide mistakes are doomed to repeat them. Thus, mistakes that are acknowledged offer opportunities for preventing recurrence.

In the case of the current clamp, what happens if you report your mistake and it is then on the list of items to discuss at the weekly shop meeting? Craig suggests extra training for this and also for the torque screwdriver mix-ups. Bill suggests color coding. Mike suggests a bar code that the CMMS can read and cross-check against the asset number. These are all good suggestions. But getting this kind of problem-solving requires that people report their mistakes.

Many people misunderstand a culture of excellence to mean you have a place where perfection is expected. That is an unrealistic, dangerous, and self-defeating expectation. What you want is a place where mistakes are expected. A genuine culture of excellence is based in no small part in learning from those mistakes.

If people are afraid to even admit their mistakes, much less offer them up for a group learning experience (or system change), they will tend to hide their mistakes. And repeat them. To avoid this, you need to eliminate the fear part of the equation. If someone reaches out for help with “I think I messed up” or “I want to tell you about a mistake I made”, how should a supervisor or senior crew member react?

Realizing that we all make mistakes, what is the proper response? Does it include berating the individual or explaining how much that mistake cost the company?

You must remember the mistake was not an intentional act, nor was it an act of stupidity. With that framing, you can look at system causes that might have led to that mistake. You listen carefully and respectfully, perhaps asking for more detail. This helps the person reporting his mistake to feel safe and respected. And it helps both of you, or perhaps the crew or whole shop, solve for contributing factors.

You can also take a pre-emptive approach. For example, you quiz individuals on how they would do (or did) a particular maintenance task. And then you watch them do it. You ask Larry, who has been lubricating motors for the past six months, to demonstrate how to lubricate a motor.

You make these notes:

1. He did not identify the grease already used in that motor.
2. He did not remove the drain plug.
3. He did not wipe the zirc clean.

Any of these of these mistakes, on its own, can destroy a motor. The fact that Larry makes all three, even in a dry run, also reveals the solution. Larry was not trained in how to lubricate a motor. Now you have a good idea of how to eliminate three maintenance errors that are causing downtime.

Another option is to take the data driven approach. For this, you will need a maintenance coroner. This person will need to look at a CMMS report that identifies failure modes by type and shows the quantity of each. This person needs to spot patterns in the data.

Let’s say “motor bearing failure” shows up as the top failure mode. This mode should be fairly rare, so this rate is abnormal. What causes motor bearings to fail at an abnormal rate? Bonding errors, for sure. But if you check the bonding and find no deficiency, that leaves these possibilities:

• Alignment problems.
• Soft foot, weak base, and similar.
• Load shock.
• Improper lubrication, which is a maintenance error.

If your downtime mortician finds the first three are insignificant, you don’t need to ask Larry for that demo. You already know lubrication is being done incorrectly. This particular problem has multiple solutions, including color coding of lubricant containers, signage on equipment, determination of the correct grease to use in the first place, and training in lubrication practices.