Clean, Inspect, Restore

Basics of total productive maintenance used to help get troublesome packaging line ready for new product rollout.

A company which we will call Omega Foods (not its real name) was expanding its business rapidly, and had announced plans for a major rollout of a new series of products that would be introduced into several existing, as well as a few new, distribution channels. This rollout would require significantly increased production at Omega's Schnitzel plant (not its real name either). After reviewing historical performance, production capacities at the plant's bottleneck, staffing levels, raw materials suppliers, and so on, Schnitzel's plant manager concluded that they could accommodate the added production demand.

Unfortunately, the plant was now into its third month of production preparing for the rollout, and was already nearly a month behind in the inventory levels needed to support Omega's objectives. On review, the bottleneck—originally thought to be the extrusion process—was actually at packaging. The packaging line was significantly over-designed and had an instantaneous capacity rating of 50 units per minute, or more than twice what was needed to support the rollout.

But it was producing only an average of about 10 units per minute per shift, with the actual rate varying wildly from day to day, and never more than 20 units per minute. And, as you might expect, downtime levels were exceptionally high. The prevailing opinion was that the packaging line was very unreliable and the maintenance manager was taking considerable heat over the line's performance. A common refrain was "Couldn't he and his staff just—fix it—right?"

Coincidentally, Omega was having a manufacturing excellence audit done at all its plants, applying principles of reliability and lean manufacturing. At the Schnitzel plant, one of the primary areas for applying these principles was the packaging line.

Adapting FMEA methods
In analyzing the problems at the packaging line, the technique applied was one adapted from failure modes and effects analysis (FMEA) methods, or, perhaps more properly, reliability centered maintenance (RCM) methods. Since insufficient information was available about the causes for production losses, other than anecdotes and "swags" at the nature of the problems, the plant had to be more structured in its analysis. In this case the adaptation of FMEA/RCM methods was to look at the packaging line as a business system, and to define a functional failure of the system as anything which resulted in production losses. A cross-functional team was assembled that consisted of two operators (senior and junior), two maintenance technicians (electrical and mechanical), an engineer, a packaging line vendor representative, and a maintenance supervisor.

The cross-functional team was introduced to reliability and lean manufacturing principles, as well as RCM and total productive maintenance (TPM) principles, during an initial workshop. This was followed with a group exercise where the following questions were posed:

  • What are the major failure modes which result in production losses? Note: these were analyzed one at a time.
  • What is the approximate frequency of occurrence of each? Guesses were OK, but we had to get general agreement among the cross-functional team, and occasionally did some external validation of initial estimates.
  • For each failure, what is the approximate typical consequence or effect on lost production per year in hours? And/or tons? And/or gross profit?
  • Are there any extraordinary repair costs for each of these failure modes?
  • What are the potential causes of these failures? Note that at this point we did not want to do a root cause analysis, but rather to make a list of potential causes for later analysis.
  • Can we detect onset of these failures to help avoid them, or better manage them?

With this information in hand we developed preliminary priorities as to which of these failures to address most urgently, i.e., those resulting in the greatest production loss.

Applying TPM principles
We also began to apply TPM principles by asking the following questions of the cross-functional team relative to each of the significant failure modes:

  • Does any of the equipment on the line require restoration to like-new condition?
  • What operator care and PM practices do we need to establish, e.g., inspection, calibration, condition monitoring, lubrication, or other basic care, to avoid or better manage these problems?
  • What maintenance PM practices do we need to establish, e.g., inspection, predictive maintenance or condition monitoring, calibration, lubrication, or other basic care, to avoid or better manage these problems?
  • Are there any specific maintenance prevention actions that we need to take in terms of modified operating or maintenance practices, design issues, etc.?
  • Do we need to do any additional training of operators or maintenance technicians to help avoid failures?
  • How do we measure our success?

Coincidentally, one simple measure which indicated our success was the number of bags of product shipped compared to the number of bags consumed per month. The difference represented how well the packaging line was performing relative to ideal. The discarded bags were also a very high percentage of the cost of manufacturing the product.

Cleaning, inspecting, and restoring
Before proceeding with any further analysis, we did a "clean, inspect, and restore" effort on the packaging line in light of what we had learned in reviewing the various production failure modes, their consequences, potential causes, and potential actions for operators and maintainers. This effort followed TPM principles, and is similar to detailing a car.

We first had to think through the logistics of the cleaning effort, and assemble our cleaning tools, e.g., vacuum cleaner, rags, solvents, etc. Then the team cleaned the machine superbly, inspecting it for anything it thought might be creating a problem, e.g., loose bolts, miscentered guides, belts off track, dirty grease fittings, etc. The cleaning and inspection was done with the failure modes in mind. In doing this we found many things that could be corrected immediately, often as a team. Some problems required formal work orders for later implementation, and still other things required design modifications, training, better procedures, or changes by suppliers.

As we were cleaning and inspecting the packaging line, we found the following situations, for which a "hit list" was kept by the operators and technicians and consolidated by the maintenance supervisor.

1. The tare weight control had been severely contaminated with product, making it very unreliable. It had not been calibrated in some time, and was not on any routine calibration preventive maintenance (PM). Operators had been regularly bypassing the "interrupt signal" resulting from load cell faults in order to avoid stopping production, and there was a lot of pressure not to stop production. But this bypassing was resulting in over and under fills, routine minor stops, and product buildup, which in turn often slowed or stopped the entire production line.

2. The bag magazine had bent loading racks resulting in improper alignment of the bags and jamming. Photo eyes and reflectors on the bag magazine were also in need of alignment to avoid fault readings and minor stoppages.

3. The two cutter/trimmers for the bag tops were different types, had different settings and different springs, and were gummed up with lacquer from the bags. Apparently improper spares and/or training had led to the use of different parts for the cutters, resulting in different settings to make them work, and no routine PM had been established for cleaning the cutters or replacing them when they were worn. Nor was there any standard calibration setting for the different style bags in use. Each was rigged daily to make it work, and a lot of lube spray was used to keep the cutting edges from sticking. The result was non-squared cuts, improper folds, poor glue lines, and generally a lot of rejected product. But pressure was so high to run the production line that downtime to correct this problem was not allowed.

4. Lacquer rubbing off the bags and accumulating at major wear points was a general problem throughout the production line, resulting in increased friction and routine jamming, but there were no routine operator or maintenance PMs for cleaning the lacquer to avoid this problem.

5. The glue pot was set 10 deg F too high, and the glue nozzles had accumulated dried glue, all of which resulted in improper glue lines and bonding, and rejected product.

6. The mechanical spreading "fingers" were not properly set, and one set was bent and had a bad sleeve bearing, resulting in improper spreading and thence folding, and therefore gluing, and frequent rejected product.

7. Several of the cams that control the advancement of the bags and other settings were caked with grease, resulting in progressively deteriorating settings as the bags moved down the packaging line, increasing the probability of jamming.

8. Several guides were bent and/or misaligned, and bolts were loose or missing, resulting in misplacement of bags for packaging, increasing the probability of jamming.

9. Several chains were loose, resulting in a jerking motion as packaging progressed.

10. Several bearings were worn out, and/or unlubricated, under lubricated, or over lubricated.

11. The palletizer had photocells and reflectors misaligned or broken, limit switches that were not calibrated, loose chains, and a faulty solenoid. Additional influencing factors included poor conveyor tracking and poor knockdown bar and shrink wrap settings.

In addition, further discussion suggested the following:

1. Bag specifications and bag quality were a problem. Marketing had insisted on a highly varied mix of colors, lacquer coatings, thicknesses, and sizes to target specific markets and customers. This was all well and good, except the bag supplier was often unable to meet the bag specifications for each, resulting in spats of jamming, rapid lacquer buildup from rubbing of the bags along the line, inadequate bonding of the glue due to excess lacquer on the bags, etc.

2. Related to bag variety were additional internal factors in that there were no specific settings for the line for each product type. As a result, changeovers and startups after a shutdown tended to be a hit or miss effort. Run it, try it, if it doesn't work, try a new setting until you get something that does work. Of course some operators were better at this than others. And the production plan changed regularly to meet the revised perception of market demand. In one instance we spent two hours starting up a production run, and were able to achieve only 36 bags of product, something we should have been able to do in 10 minutes, not 120 minutes.

3. The hopper feeding the packaging machine appeared to be undersized for the duty now required by the rollout, resulting in inconsistent feed rates.

4. There were considerable differences between shifts as to exactly how each operator ran the packaging line. Given the lack of standards and the poor condition of the equipment, this was not surprising. Each operator was trying to accommodate the faults in the machine along with an extensive number of product types, leading to a wide range of operating practices.

Or, as the saying goes, "Other than this, everything was running pretty good!"

The results
As we were cleaning, we managed to correct many of these problems on the spot, e.g., re-calibrate the tare weight control, align the bag magazine, reset and calibrate all the photocells and reflectors, clean the lacquer buildup, reset the glue pot temperature and clean the nozzles, clean the cams, tighten the chains, and so on. Other efforts required a work order and spare parts, e.g., restoring the cutters and folders to like-new condition and replacing certain bearings. Still other things required additional engineering review, e.g., the hopper size and rates; PM requirements, including new procedures, training, and schedules for PM; and the development of specific settings on the line for packaging various products.

In just two short days—one day of review and analysis and another day of cleaning, inspecting, and restoring—we doubled production output. While it was momentarily painful for the plant and production managers to take the packaging line out of production, it was essential to identify and solve the problems that had been identified.

However, this was just the beginning. We now had to develop and agree to:

  • A specific operator care and PM program, one which involved operators doing basic care activities—tighten, lubricate, clean, calibrate, and monitor—to avoid or detect developing problems. One point to be stressed is that the operators had to own the reliability of the equipment, much the same as we own the reliability of our cars. To do otherwise is like expecting mechanics at the garage to own the reliability of our cars—they can't and won't. Mechanics can help with diagnosis, repairs and restoration, PM, and so on, but we have to own the reliability of our cars.
  • A specific maintenance PM program, one which involved the maintenance techniciansrestoring the assets to like-new condition, assuring consistency and adequacy of spares, doing loop calibrations, planning for shutdowns and startups, doing the more difficult and/or intrusive PM, and generally working with the operators to assure reliability and manufacturing excellence in the packaging line. Note that the two PM programs were actually one program developed and agreed upon by operations and maintenance.
  • A specific set of instructions for the PM program, one which included taking digital pictures of each PM area, adding arrows/circles, instructions, key steps, cautions, and so on. This PM instruction set was to be kept near each machine for ease of use by operators and maintenance technicians.
  • A specific manual of setup instructions for each area of the line for each of the different products packaged in each type of bag, i.e., for each stock keeping unit (SKU). Each area would have specific settings for each SKU, which would be used for all changeovers as the reference point. Each shift would be required to follow the same settings and procedures, or to log and justify the reason for not following the settings. Settings would be upgraded with improvements and changes to SKUs.
  • A training program to train all the operators and maintenance technicians in the proper PM, operation, maintenance, and setups for the line.

All in all we got pretty good results in just a couple of days, and developed a good plan for continuing with the improvement and assuring manufacturing excellence. MT

Ron Moore is managing partner of The RM Group, Inc., 12024 Broadwood Dr., Knoxville, TN 37922, and author of "Making Common Sense Common Practice: Models for Manufacturing Excellence" as well as numerous journal articles on manufacturing excellence. He can be reached at (865) 675-7647.