9.1Select and Implement Controls

No problem-solving event should ever be considered complete until some kind of safeguard is put in place to minimize the probability of a recurrence of the problem. Wouldn’t it be a shame after all of the work a team did to determine the root cause of a problem and then implemented a solution that some kind of control of the factor that caused the problem wasn’t implemented, and the problem came back into the process. It is important to remember that the purpose of a control is to give an advanced warning that something may be wrong before it is actually wrong. For example, diabetics routinely monitor their blood sugar levels to make sure that the level of their blood sugar stays within an acceptable range. They do this to avoid damaging their vital organs.

By far and away the best type of control is mistake-proofing or what the Japanese call a poka yoke. Mistake-proofing involves the installation of a sensing or control device within the process that will prevent errors from being made. The advantage of these devices is that defects are prevented, rather than detected. A poka yoke is based upon the premise that 100 percent inspection can be successfully achieved if done by a nonhuman device, rather than by a human who has been proven to not catch all errors. Fail-safe devices can be used effectively to prevent such things as the incorrect positioning of parts or the omission of parts. Unfortunately, sometimes due to the nature of the measurement or because it is cost prohibitive, not all processes are capable of utilizing a fail-safe device. But there are other alternatives.

If it is not possible to implement a fail-safe device, then a manual measurement and a control chart is the second choice. Because the control chart is based upon the normal statistical variability of the process, data is collected and control limits are calculated for the process parameter or product characteristic. Control chart theory states that as long as the measured or calculated data point falls within these limits and the process has been deemed capable, then we are reasonably sure that defects will be caught. The only problem with these type controls is that they are based upon samples, rather than 100 percent inspection.

A control chart is simply a run chart with control limit lines equally spaced from the process average. There are many different varieties of control charts, but in my opinion, the most effective is the x-bar and R chart. This chart simultaneously monitors the location of the process parameter, or product characteristic being measured, relative to the historical average of the process and the amount of variation present. If the calculated average and range remain within predetermined limits, then the process is said to be “in control” and is permitted to continue running. If the process goes outside these control limits, then the process is said to be “out of control” and is stopped while corrective action is taken.

There are many excellent reference books on control charts that summarize other rules that dictate whether a process is out of control, so I suggest you do a Google search to learn more about them.

Figure 9.1 is an example of a control chart that one of my teams developed to control product coming off of a shear machine. The product being controlled with this control chart was a long rectangular piece of metal that not only had to be a certain width but also had to be parallel along the entire length of it. The x-bar (i.e., average) portion of the chart was used to control the width of the rectangle, and the R (i.e., range) portion was used to control its side-to-side parallelism. This chart proved to be a very effective method of control for a problem that had been “fixed” many times in the past. As you can see, the procedure for collecting and recording data is located in the box at the far right side of the chart. Directly beneath the procedure are “rules of action” that define out-of-control conditions and actions the operator should take if or when the process actually goes out of control.

Figure 9.1Control chart of ring shear machine.

The third, and least attractive, alternative is an audit of the process. Audits are intended to be a review of a procedure or process or system to determine if the person performing the work is doing so according to instructions. Audits can work well, as long as the subject of the audit is well defined and the auditor knows exactly what to look for. The problem is that audits are somewhat subjective and have a tendency to evolve over time.

Our split inner liner team members were forced to choose this alternative. They knew now that excessive application of cement was the cause of split inner liners and updated the work method to include photos of a “good” tank. They also added written instructions on how best to apply the cement. The control that the team selected was to, on a weekly basis, monitor the work methods of the cement operators for compliance to the new method. Process audits can be effective if done correctly, but any time there is the option of using either a fail-safe device or a control chart, we should always choose one of these as our control tool.

The key to effective audits is a complete definition of what the auditor is looking for as the process is reviewed. My recommendation here is that the team develop a checklist of items that are important. Since the team knew how important the correct application of cement was to the elimination of inner liner splits, its checklist focus was on this.

The checklist contained a section for application method that attempted to quantify the number of times the operator’s mop was dipped into the solvent per tank side; the number and length of strokes of the mop on each side of the tank; the number of times the operator applied cement to the same location; the number of passes over the joint area; the pressure exerted on the mops by the operator; and the direction of application of the cement for each layer. The team believed that if the cement operators performed the work method to a prescribed work method that the incidence of split liners would be significantly reduced or eliminated.

9.2Document Your Success

One final and important activity to complete is to document your success. Documenting how your team successfully solved the problem does several things for the organization. First, it reinforces that the use of a structured and systematic approach to problems really does work. Second, your report is there for future problem-solving teams to follow as a roadmap. Third, if the problem you have solved should ever return (God forbid!), then your team’s report is available as a ready reference.

The actual report does not have to be lengthy. In fact, each of the first sixteen steps could serve as the format for the report. I always recommend to companies that they should start a file of “Best Practices,” which is simply a collection of solved problems, and your report belongs in this file for future generations of problem solvers.

9.3Celebrate Your Success

The culmination of a successful problem-solving event should always be recognized and celebrated. After all, the team was able to come together as a group, follow a systematic and structured process, and end a problem that was impacting the business. I always end a successful event with doughnuts or some other form of recognition for a job well done, and you should too.