Lean Maintenance is a key ingredient to a successful Lean Manufacturing environment.Lean Manufacturing seeks to cut out three types of waste:
The latter concept is "mura" or "unevenness". If an employee's work day is divided between being idle for five minutes and working frantically for ten, then her workload should be smoothed out so she can work steadily all day.
In a Lean Manufacturing factory:
Therefore unscheduled downtime is a serious problem in a Lean Manufacturing facility. Unscheduled repairs interrupt the flow of work-in-process, thereby adversely affecting delivery time. There might not be a place for work-in-process to be stored even if the preceding operations can continue. The costs of unscheduled downtime are enormous in such a factory.
Lean Maintenance supports a Lean Manufacturing approach by preventing unscheduled downtime. The method is to reduce or eliminate the factors that cause machine breakdowns.
The four most common root causes for machines to break down are: operator error; programming error (for computer-controlled tools); inadequate maintenance; and environmental causes.
Operator errors can be as simple as improperly placing material or using too much force on a tool.
Numeric-control or computer-guided machines may have program bugs. A simple example would be trying to drill too quickly for the type of material.
Regular maintenance operations might be neglected, especially in a busy factory. An operator could forget to add lubricants. The operator might not be trained to perform all the necessary steps.
Many environmental conditions can take their toll on equipment: heat, humidity, vibration, or airborne chemicals.
Of course, some of these conditions may combine. An operator might neglect to clean away dust, and unusually humid weather could change that dust into a sticky clog over a cooling fan.
Lean Manufacturing uses customer orders to pull production from raw materials. It is most successful when the company has a fairly reliable mix of orders coming in at a steady pace.
It is, indeed, possible to set a maintenance schedule so that the schedule pulls the maintenance activity. However, by itself, a schedule cannot prevent unexpected breakdowns.
Such breakdowns are equivalent to casting the manufacturing schedule aside because the most important customer has placed a rush order that pre-empts all others.
On the production side, a new customer order may trigger purchasing new raw material. This lead time should be built into the customer's delivery schedule. However, a machine breakdown that requires an expensive, out-of-stock replacement part immediately harms the schedule. It is uneconomical to stock every possible replacement part to avoid this situation.
The performance parameters of the machines are more readily available than the likelihood of breakdowns for those machines. This type of analysis requires research into the maintenance and repair histories. See the article about Preventive Maintenance] for more background.
To conclude: Lean Maintenance differs from Lean Manufacturing because the maintenance must be planned, where the manufacturing can be "pulled" by customer orders:
There is one similarity: both require careful planning based on the actual capabilities of the machines, whether to produce or to require preventive maintenance.
Recall that the primary concept for Lean Manufacturing is to reduce several types of waste. For Lean Maintenance, the equivalent "waste" concepts to be avoided are:
Many of the gurus advocating Lean Maintenance point to blindly scheduled maintenance as a waste, especially if it involves replacing expensive components or incurring high labour costs or lengthy downtime.
Note the word "blindly", however. Their advice is to measure, say, the wear of a part or the viscosity of the lubricant. This is "condition-based maintenance", which defers maintenance until there truly is a need. Two requirements for condition-based maintenance are: measure the critical indicators that will change before the machine fails; and measure with enough lead time to schedule the maintenance and procure any required parts.
Refer to the repair history as a guide for how long a machine actually did run before a breakdown. This can provide a rough plan for scheduling maintenance. Also use that repair history as a guide to what should be tested in order to implement condition-based maintenance.
To borrow a phrase from automotive maintenance, "While you've got the hood open, you should also…". Scheduling several maintenance tasks for the same time can be much more efficient than arranging for repeated shutdowns.
Good planning also ensures that the right parts and tools are available, along with people who have the appropriate skills.
Of course, a "Lean" approach would also examine how efficiently the actual work is performed. Are there extra trips back to a storeroom for supplies? Would it help to invest in a maintenance trolley? Would better tools help the employees do a better job? What about training session from the manufacturer?
Clearly the "frantic hurry" phase of maintenance comes during an unscheduled breakdown. In this situation, the repair person is most likely to make mistakes due to time pressures. This has the least likelihood of having all the necessary tools and parts.
Many maintenance managers aim to keep their maintenance crew scheduled to about an 80% load. This leaves some time for emergency repairs but the crew still is busy throughout the shift.
Another example of sporadic maintenance is a maintenance blitz. Beyond the obvious problems of overloading the maintenance team's schedule, this blitz also means that skills may have become rusty in the intervening time.
You should pause here and take the Lean Maintenance / TPM Assessment quiz.
The following precepts help guide the planning for Lean Maintenance:
Each machine, or at least each model of machine, has its own maintenance needs. Do not assume that each needs the same schedule.
Each operator should perform regular inspections and maintenance on their machines. These people should know the machines well, and can certainly take part in minor maintenance. Invest in training them for these tasks. Add instruction sheets and find other ways to make it simple.
One example of making maintenance easy is an automobile's dipstick: bright yellow handle like the other two things the driver is encouraged to check; only one place to put it in; has markings for "too low", "in range" and "overfull".
Post a schedule for the maintenance activities, with boxes for the employee to initial and date, beside the instruction sheet.
Train the operators how to write work orders or how to report trouble verbally. This helps the repair people, and their schedulers, know what is wrong or where to start looking.
If less than 10% of maintenance activity is "repairs", you might ease back on the preventive measures. If it is more than 20%, put more work into preventing the worst breakdown and repair situations. If your maintenance people can get everything done in the regular day shift, you're doing well.
Without a computerized maintenance management system (CMMS), how do you log, track, and report on patterns of maintenance and repair activity? What do you use to determine the right frequency for preventive maintenance? How do you determine how many spare parts you need to keep a machine going without making emergency orders to your supplier?
In the best maintenance environments, a significant portion of the work schedule involves training and mentoring the operators to take on regular maintenance work.
None of this comes together without management support and commitment.
Lean Maintenance is a key ingredient to a successful Lean Manufacturing environment.
Like Lean Manufacturing, Lean Maintenance seeks to avoid waste:
Wasted "activity" includes having expensive replacement parts in inventory longer than was necessary; ideally these should be ordered "just in time" for scheduled maintenance.
Inefficient activities may refer to delays while performing maintenance, such as chasing around for tools or replacement parts. Another type of inefficiency comes with poor scheduling: an example is replacing two parts on two separate occasions, when both tasks could have been done during the same outage.
Ideally the maintenance team, like the production operators, should work at a steady pace that is productive but not rushed. A repair crew is more likely to make mistakes when the pressure is on to fix a machine and get the factory back into production.
Generally consultants advise integrating Lean Maintenance with the production disciplines, such as Six Sigma or 5S, already implemented in your factory.
However, this article provides advice on what any Lean Maintenance program needs and should accomplish. It is possible, in fact, to implement Lean Maintenance without formally adopting any of the other disciplines. Some industries, such as mining and pulp-and-paper, have serious maintenance requirements and may be actively seeking ways to improve. In many other industries, maintenance is not often management's first area of concern when looking to improve productivity and profitability. Even so, inefficient maintenance is expensive and unscheduled breakdowns are even more costly: it may be very worthwhile to investigate what Lean Maintenance can do for your factory.
Every Lean Maintenance Program needs:
Here are more details on these requirements:
This may be the largest hurdle in planning a Lean Maintenance program. To address the needs of the factory, you need an inventory of each machine. For each, you need to know:
All this information includes both labour and parts requirements.
The ultimate goal for Lean Maintenance is to minimize the cost of performing preventive maintenance while also minimizing the risk of an unscheduled breakdown. The cost of maintaining one machine is usually much less than the cost of its breakdown, because that might shut down operations altogether.
The usual goals for scheduling the maintenance crew are:
The goals for containing costs in spare-parts inventory are:
Critical replacement parts are those that are necessary for a machine to operate, and have a long lead time for purchasing. These are too expensive to keep in stock, but highly valuable in that they are needed to keep the machine running.
On the other hand, the Lean Manufacturing approach is to procure raw materials "just in time" for production. The same principle should apply to replacement parts: purchase them "just in time" for the next maintenance activity, rather than just after the last one.
This drives the need for accurate historical information on the intervals between failures of a critical replacement part.
Make a separate plan for maintaining each model of machine. You may need to engage experts to analyze the stresses and causes of each machine's past failures, and the values of the components to test for "condition-based" maintenance planning.
Part of the process is to meticulously log the maintenance and repair history.
Applying root-cause analysis to every failure (or every need for maintenance under condition-based maintenance) should lead to important clues for extending the useful life of the machine. As an example: let's say that lubricating oil begins to break down, triggering an oil change as maintenance. Why is the oil breaking down? Let's say it is dust getting past a filter. The long-term solution is either to replace the filter more frequently or to find another way to prevent dust from getting there (by adding a fan to blow the particulates in a different direction or by adding a filter to trap dust where material is being cut).
As you acquire data, you need to evaluate the effectiveness of the Lean Maintenance program, by comparing various metrics between the "before" and "after" time periods:
Of course, you will add the cost of planning and implementing Lean Maintenance to the "cost" ledger.
Most factories see a significant reduction in overall cost, because unscheduled repairs are extremely expensive due to lost productivity.
Ongoing evaluation is important to the success of a Lean Maintenance program. Besides giving the program a cost/benefit analysis, ongoing evaluations provide the guidance for further improvements.
The following are useful metrics for any maintenance program, but especially useful for Lean Maintenance:
Let's examine why these numbers are important, and their targets or trends.
The goal is to increase the ratio of scheduled versus unscheduled work orders, with an achievable target of 90%. Whether you make your schedule based on past repair history or, preferably, using condition-based testing, scheduling has several benefits over repairing after breakdowns:
The target is to schedule about 90% of the tasks. Many maintenance gurus estimate that even fewer breakdowns indicates that too much time and effort are being spent on maintenance.
The benefits here are the same as for the number of work orders.
The goal here is to schedule the maintenance crew to about 80% of their capacity. This leaves room for maintenance that takes a bit longer than expected, and for unscheduled repairs. Again, if less than 20% of the labour time goes into repairs, then it might be more economical to ease back on maintenance.
The goal is to see this number trend upward. The front-line operators should be able to do routine condition testing and routine preventative maintenance. The better they perform these functions, the more time the maintenance staff has for more complex work.
The training should include "how to write up a work order" so it communicates the situation as effectively as possible.
Scheduled maintenance should, ideally, avoid impeding regular production. The machine itself is not available for work, but it should not be required during the maintenance period. Downtime should count only the elapsed time that this machine is unavailable.
In a Lean Manufacturing shop, where work-in-process should flow smoothly from machine to machine, an unscheduled repair could bring the entire factory to a halt. In that case, the elapsed "downtime" is multiplied by the number of other machines kept idle during the repair.
So this percentage should climb significantly and stay high.
This metric is similar to "downtime" in the previous section. In this case, however, we measure the value that the machine(s) should have been contributing during maintenance or repair.
Of all the metrics in this article, the cost of lost production is the most important to management. If a breakdown of one machine stops the entire factory, or even a substantial fraction of it, then the cost of lost production is enormous.
These situations are included in the above metrics, but breaking them out is important in managing the inventory of replacement parts. If a necessary part is out of stock when it is needed, one or more machines will be idle. Typically the time to procure a part will be much longer than the time to install it.
The question is very important: what did it cost to be out-of-stock of a critical part?
As noted earlier, one benefit of scheduling maintenance is that the scheduler should check on the availability of the replacement parts. This can "pull" the part by initiating a purchase order. This can also put a delay into the schedule: "Do not set the date before the part has arrived".
This number should be balanced against the cost of keeping a machine idle because of a parts shortage. A repair-oriented shop needs to maintain a large inventory of replacement parts, since it has no way of knowing when a breakdown will occur.
A goal of Lean Maintenance is to reduce the "waste" of keeping an inventory of replacement parts. This inventory should decrease over time as parts are used, and only re-ordered in time for the next usage.
The typical goal in a maintenance program is to minimize the cost of keeping the equipment working.
You need to know and understand the costs for preventing and correcting equipment breakdowns:
When you evaluate your maintenance program, you should find that as the improved maintenance makes your factory more reliable, you are reaping these savings and benefits:
In short, when you evaluate your factory's Lean Maintenance program, you should find that it has led to higher profits.
In many plants that follow a maintenance strategy beyond "Fix whatever breaks", the maintenance plan is integrated with the manufacturing discipline.
Lean Maintenance tries to avoid wasted effort in maintenance by scheduling condition-based maintenance (test or inspect each machine, then schedule maintenance when it begins to show signs of wear). This approach can reduce the waste of storing an excessive inventory of replacement parts for too long, as well as the waste of time to procure out-of-stock replacement parts.
This article deals with integrating Lean Maintenance with other manufacturing disciplines:
The 5S methodology uses five principles: Sort, Set in order, Shiny-clean, Standardize clean-up, and Sustain. It is often a first step towards Lean Manufacturing and other production methodologies.
The Sort step divides things into the necessary and the useless or out-dated categories; then it discards the useless things! Replacement parts are included in the search for misplaced or out-dated items. This step also looks to improving safety conditions and identifying deferred maintenance.
The "Set in order" step ensures every item is in the right place: the most-frequently used are stored nearest the point of use. This step also lays out the flow of material for efficient work processes. A clear action item for the maintenance team is to ensure that tools and replacement parts are stored where they are convenient for the tasks. On the other hand, they will probably be stored farther away than the daily production tools. Therefore, clear labeling and a strict discipline of "put it back where it belongs" must be followed.
"Shiny-Clean" is a one-time step to make everything clean and set that as the standard.
Standardized clean-up is the ongoing and regular task of keeping the facility clean to meet the "Shiny-Clean" standard. It should be a five-minute task at the end of each shift, with larger tasks on a schedule. Regular inspection or maintenance of equipment by the operators can be added to this daily schedule.
The final ‘S' is "Sustain". This ingrains the 5S behaviours into the factory's culture and into the standard working life of each employee.
To some degree, the 5S view of dirt is like the Lean Maintenance view of equipment breakdown. 5S would ask "Why does this area get dirty? What can we change to prevent dirt from building up"? Preventive cleaning is like preventive maintenance: it keeps the problem from occurring.
The 5S program is very useful if dirt, dust or other contaminants lead to equipment breakdown. It is also excellent at ensuring tools are stored in the right places. This can be especially important if maintenance tools have been stored near the maintenance team's headquarters, rather than in the department where they are used. 5S also helps if rarely-used tools are misplaced or forgotten.
A typically successful 5S program increases overall productivity by about 10%. Several factors come into play:
As well, applying 5S to the maintenance shop affects the productivity of these technicians. It is distracting and counter-productive to leave half-finished repairs sitting on a workbench because replacement parts are not available or not conveniently stored. Either prohibit bringing machines into the maintenance shop unless the replacement parts are available, or store the broken equipment safely out of the way so the technicians can repair items that do have all the parts available.
Lean Manufacturing eliminates wasted activities that add no value for the customer, improves worthwhile activities by eliminating inefficiencies, and smoothes out the workload to eliminate high-speed and high-stress phases.
Are there similarities and synergies between Lean Manufacturing and Lean Maintenance?
Does maintenance add value for the customer? In its most direct form, maintenance does not. The customer values the finished product, not the warm and fuzzy feeling that their supplier's machinery is working smoothly.
However, Lean Maintenance should reduce unscheduled downtime for repairs, leading to on-time production and on-time delivery to customers. Customers do value on-time delivery. One aspect of evaluating a maintenance program is to check for an improvement in customer retention due to timely deliveries.
As well, many machines begin producing defective, or at least, less-than-ideal products, as they near their next maintenance date. Performing maintenance on time, "just" before the machine begins to deviate from its specifications, can lead to lower defect rates and continued high-quality output.
With a focus on condition-based scheduling for maintenance and a reduction in the need for unscheduled repairs, Lean Maintenance does reduce inefficiencies. Foremost is the fact that unscheduled downtime in a Lean Manufacturing shop affects more than just one machine: downtime can disrupt the whole factory.
When scheduling maintenance, the supervisor can also check and ensure that necessary replacements parts and skilled workers will be available at the scheduled time.
As well, if multiple maintenance operations can be performed during one outage, it can be more efficient than staging multiple disruptions. This is especially true if the same parts must be disassembled each time, and if several people can work on different tasks simultaneously.
This may be the most obvious benefit of scheduled maintenance versus unscheduled repairs. In a repair scenario, everyone feels the pressure to fix the machine so production can resume. The lost productivity is expensive in labour costs at other machines idled by the breakdown. Missed deadlines may disrupt customer relationships and lose future contracts. The pressure can be intense.
Lean Manufacturing recognizes that working at high speeds and under pressure often lead to errors. Lean Maintenance has the goal to make scheduled maintenance about 80% of the crew's labour time. This does not reach the Lean Manufacturing ideal of smoothly scheduling all work, but it may be a significant improvement over the ratio of maintenance to repair that your factory currently experiences.
To conclude, the two "Lean" approaches do work well together. It will not be possible to fully smooth and schedule the maintenance side as completely as the manufacturing. Even if you "overspend" on maintenance, Murphy's Law will sometimes cause an unexpected breakdown.
However, the corporate culture espoused by "Lean" works well for both the manufacturing and maintenance operations.
Remember that Lean Manufacturing requires reliable machinery to keep the production processes working on time. Lean Manufacturing absolutely cannot succeed if unscheduled repairs cause widespread delays.
Indiscriminant maintenance, meaning "hire more technicians, keep more spare parts, and schedule lots of maintenance", is an unprofitable response to the need for reliability. A Lean Maintenance program is both effective and cost-effective in improving the reliability of the machines in a Lean Manufacturing environment.
The Six Sigma discipline reduces variation in order to improve quality. A Six Sigma approach to Lean Maintenance uses Six Sigma's DMAIC project cycle:
One might run a Six Sigma project for Lean Maintenance on one machine at a time, rather than trying to analyze a whole factory.
Defining the problem would then be more specific than simply "unscheduled repairs". An example of a more focused problem would be "the gears seize up too often; we have to stop and make repairs".
Measure that specific problem by reviewing the repair and maintenance log for that machine. This may reveal "the gears seize up every two or three months". Each repair takes about two hours, so lost productivity and direct labour can be calculated easily. Let's say that parts are not usually damaged on this machine, so replacement parts are not a factor.
Six Sigma has an extensive problem-solving methodology, using principles from statistics to determine the most likely causes. For this example, the analysis shows that the correct lubricant is used. Operating speeds and temperatures are normal until just before the gears seize, at which time it slows down and heats up. The lubricant is much too thick at the time of failure. No excessive wear is found. A lot of dust is in the air around this machine, and settles on everything. The electric power supply is stable and not the problem. The analyst concludes that excess dust is getting into the lubricant.
Further analysis is required to decide on a solution: to spray a liquid to hold down the dust where the material is being cut; fans or blowers; air filters or dust shrouds? Should we simply schedule changing the lubricant every six weeks, or when the temperature rises past a certain point? These theories are tried and tested before deciding on a solution.
At this point, implementing the solution may require changes in the engineering of the equipment or in the behavior of the operator.
Once implemented, the project is not complete until the effects are measured. In this case, does the lubricant remain at the correct viscosity for six months or a year? Do the gears seize up anyway in three months? This controls the process so the outcome is known and verified.
The Six Sigma approach can be overwhelming in the analysis step. However, it works well for difficult problems. It can be time-consuming, especially in the analysis and final control steps.
However, the DMAIC approach, with appropriate quality tools, is effective in solving the widest range of problems. This approach is the ideal standard procedure for troubleshooting and problem solving, regardless of whether the company officially embraces Six Sigma.
By Oskar Olofsson