RELIABILITY 101

F undamental reliability begins to take shape when a work community shifts its attitude of maintenance from "fix it when it breaks" toward something that is an ongoing and important function. It starts with two simple questions. In your work area:

1. Who is responsible for reliability?

2. Who is responsible for production output?

The answer should be everyone . This book has advocated the use of cross-functional teams and the development of active learning and multiskills for everyone. To sustain high OEE numbers, employees in the different manufacturing functions need to have basic understanding of concerns and aspects of the other functions before making decisions on priorities about their own functions.

7.1 Fundamental Reliability

Fundamental reliability is most powerful if plant managers, production and operation managers, maintenance managers and production supervisors champion it. This leadership team for the area must take stewardship of the overall system. It must balance production and production capability over the long term. They should build on the idea that "If you are not promoting reliability, then you are promoting failures."

Sharing a set of basic concepts and definitions will provide a platform for communication so that everyone can contribute to reliability, availability, and maintainability (RAM).

Because everyone is responsible for RAM, the expectations of that responsibility should be understood. These are:

1. Eliminate failures.

2. When failure occurs, reduce the impact or consequence of the failure.

3. When failure occurs, use both short-term and long-term considerations to optimize repair, and restore the systems.

The first of these is most important. If you could eliminate failures, then the other two would not be necessary. Optimizing repair for both the short term and the long term means that the fastest solution is not necessarily the best solution. Take the appropriate time to collect data, confirm root cause, and allow good workmanship so that the event is correctly understood. Prevent failures from reoccurring. All too often, the fastest fix requires a second time, or more, to correct the problem. A fast fix doesn’t promote learning from mistakes.

Two of the most important steps in promoting RAM are collecting and analyzing data. These steps provide good information for problem solving and setting strategy. The data does not consist only of measurements and numbers, but also includes samples, diagrams, drawings, pictures, observations, procedures, and practices. Just as modern forensic study reconstructs a crime scene, many root cause sources of equipment failures can be uncovered by scientifically studying the failed part, then reconstructing the failure circumstances. This process can be the most exciting aspect of reliability: carefully determining the root cause of a chronic equipment failure, creatively injecting a simple solution, and monitoring results that show significant improvement on area throughput. Nearly everyone enjoys playing detective when they read mysteries. The same holds true for equipment reliability. Everyone should participate in solving the mystery. All factors should be suspected before eliminating them from being the root source problem. To set the investigative stage, let’s set some common guidelines that we can use to identify these root source problems.

Begin by obtaining OEE data and categorizing the losses and summarizing event details, as outlined in chapter 2. Rank the loss events using the value fulcrum suggested in chapter 5. As appropriate, factor in actual business parameters (e.g., known costs) for various loss events. Use cross-functional teams to work on the top two or three items. Do not allow discretionary resources of people, time, or money to be directed at anything other than these top items. This focus will provide the greatest impact and the fastest rate of improvement for your area. Everyone wins when area effectiveness jumps significantly.

When resources can’t be used directly on cross-functional OEE teams, focus them on developing best practice methods and procedures. This step could start with clarifying procedures and set points, collecting and categorizing all downtime minutes, and reviewing maintenance strategies. It would also include applying simple Predictive Maintenance techniques and statistical process control (SPC), then using these tools for proactive, conditioned-based maintenance (CBM).

Assume that equipment reliability of a certain subassembly is designated one of the top three items. Everyone in the community should be focused on solving this item. The necessary resources should be extended with priority; other discretionary activities should be subordinated to this investigation. Following the Theory of Constraints (TOC) steps modeled for shutdowns in Chapter 6, we have identified an important parameter for our attention. We will exploit the study and root cause analysis of this limiter. Meanwhile, we will subordinate other activities from distracting our resources. Once the root cause is determined, we will elevate the changes to eliminate the problem, using designed experiments and monitoring the results to prove that our analysis and action have addressed the problem. With this proof in hand, we will go back and tackle the next most important OEE limiter.

One of the most successful work centers where I worked used this exact approach. Over a three-year period, the work center showed dramatic improvements on an established system, having 70 percent throughput improvement with 12 percent fewer employees.

Once a specific subassembly is identified as a top priority, all pertinent data about downtimes, frequencies, and event details is required to begin root cause analysis. The work community's ability to collect and maintain a good database on all downtimes is very important. Good practices can prevent countless hours being spent guessing and verifying data after the fact. Good practices can reduce the risk of bad data being used which leads to incorrect causes. A good database will help you focus on the right information and allow your analysis to zoom in on the root cause.

These are among the reasons why a good database is valuable to your company. Everyone must understand this aspect, accept the responsibility, and maintain the discipline to report and record all the necessary data for each interruption. As the cost of personal computers spiral down and the need rises for vast amounts of information to be collected, organized, Pareto charted, analyzed, graphed, reported, and shared, every company should strongly consider using a computerized maintenance management system (CMMS). Good databases sorted quickly via computer systems are very useful for reliability studies.

With data collection in mind, let’s review some key reliability definitions.