Some Techniques for continuous improvement in maintenance failures
By Joel Levitt
Continuous improvement of failure is a five step process. The five steps are the same for continuous improvement to an accounting, a purchasing or a housekeeping department problem. Find mistakes, breakdowns in communications, items that slipped through the cracks and apply these steps. The goal of all continuous improvement is either reduction to the inputs of:
1. Labor (operator, mechanic, clerk, staff and contract)
2. Management effort (reduce headaches, non-standard conditions requiring management inputs)
3. Maintenance parts, materials, supplies
4. Raw materials
1. Improve reliability (uptime) and repeatability of process
2. Improve output
3. Improve safety for the employees, the public and the environment
The five steps for continuous improvement of failures:
Commitment: Organizations that are committed to continuous improvement commit the time to do the deep analysis necessary. They allow and encourage maintenance workers participate on problem solving teams that involve other parts of the organization including production, engineering, quality, cost accounting and marketing. Commitment also is a long term choice. In good times and bad the maintenance department must always be looking at improvements.
Measurement: A necessary prerequisite to a continuous improvement attitude is establishing ways of measurement of the maintenance department. The process of setting up measures is called benchmarking.
Review the production and then the maintenance incident history (an incident could be a breakdown, a series of breakdowns, PM's for a machine, a series of minor adjustments, or other maintenance activity). Review from six different points of view:
1. Economic analysis
What is the cost of the incidents, the downtime, per year? What is the return on investment of a projected improvement? How much should we spend to fix this?
2. Maintenance analysis
How disruptive is this breakdown. Does this breakdown cause mechanical or electrical problems elsewhere? What is the honest opinion of the maintenance `old timer' experts? Is the root cause a faulty or inadequate maintenance procedure? Is the root cause related to inadequate training?
3. Statistical analysis
How often does the incident occur? Is there a pattern or trend? What is the mean- time-between-failures (called MTBF). What is the mean-time-to-repair (MTTR)? Can the MTBF be managed by replacing the component more frequently then the failure frequency?
4. Engineering analysis
What was the mode of the failure? Is a structural analysis of the broken parts indicated? What happened just before the failure? Why did the breakdown take place? Was there a failure of the PM system? Do the PM task lists look at this failure mode? Are we looking at the root cause?
5. Operations analysis
Does this event impact operations? Does the event force the failure of other parts of the process? Can we bypass the problem with a back-up or standby unit? Is there a scrap or start-up exposure? Is this indicative of a failure of the operations system?
6. Marketing/business analysis
What is the impact of this failure on the customer (internal/ external)? Can we afford to have this type of event happen? Can this event impact quality? Does this impact morale? Is there an impact outside our sphere of influence (environmental, competition)? How high should the priority be to deal with this problem?
No matter how knowledgeable individual professionals are, they cannot know all of the ramifications of a maintenance event. If continuous improvement is seriously pursued multi-departmental teams will be necessary on an ad hoc basis to attack problems.
Action: After an investigation then institute the improvement in one or more of 4 areas.
1. Modify maintenance or PM procedure
Add tasks to catch the particular failure mode earlier on the critical wear curve. Increase the technology of the tasks. Increase frequency of the tasks. If economic and business analysis shows that we are spending too much on PM in relationship to other costs do the opposite of the above (reduce frequency, depth, etc.). Investigate improved lubes if this is a wear problem.
2. Modify machine (maintenance improvement)
Improve the machine so that it doesn't break or need adjustment. Improve the tooling. Make it easier to do the maintenance tasks. Automate some of the maintenance tasks. Remove the source of the problem (redirect the dirt so it doesn't fall on the cylinder). Add automated lubrication systems. Add instrumentation to machine.
3. Modify part
Make the part easier to produce. Improve the tooling. Change the shape, material, finish. Reduce the number of steps. Reduce (or increase) part's the specifications. Simplify the steps.
4. Modify production process
Improve the whole process. Improve incoming materials. Make the transfers more bullet proof. Look for improved technology. Look for a whole new and more reliable process. Give headache processes to vendors who are expert in that process.
Continuous improvement is the only antidote to the constant pressure of competition. It is essential that management realize that continuous improvement in the maintenance department is everyone's business and can only be achieved with everyone's input.
The goal of all continuous improvement is either reduction to the inputs of: Labor, management effort, maintenance parts, materials, supplies, raw materials, energy, capital and overhead, or improve reliability and repeatability of process, improve output and improve safety for the employees, the public and the environment. Continuous improvement of failures is a multiple step process. These steps are: Commitment, Measurement, Investigation, and Action. Organization should be committed towards continuous improvement and devote time, and encourage workers. The company should establish ways to measure the maintenance department. The process of setting up measures is called benchmarking. Review the production and then the maintenance incident history. After investigation, take action. Improve the tooling. Make it easier to do the maintenance tasks. Automate some of the maintenance tasks.