Total Productive Maintenance / Preventative Maintenance

Konichiwa!!

Hisashiburi da ne (It's been a while)!!! I just took my last finals and now officially a Master of Science in Industrial Engineering. I also had the honor of having Bill Clinton as my Keynote Speaker at Yankee Stadium! You can take a look at it at http://www.youtube.com/watch?v=rSfSk92wDog
It has been a pleasure going to NYU-Poly and being apart of the I2E culture. Now back to the blog.

I would like to go over Total Productive Maintenance (TPM) because it is very important to all organizations and people alike. TPM was developed in Japan as an improvement to Dr. Deming's U.S. born preventative maintenance programs. Nippondenso was the first company to employ it because they were tired of spending so much money on fixing their machines so they trained their operators to recognize potential problems and fix them themselves; thus, creating TPM.

The target of any productive maintenance program is to maximize plant and equipment effectiveness to achieve the optimum life cycle cost and downtime efficiencies.

How can such a program be useful to an organizations and person's life?

The answer is something that you may already be doing but have not really focused on. Your body is a machine and in order to increase its effectiveness while decreasing its downtime you must learn to service it. You can service your body by learning about indications of failure, what to eat to increase output, what to wear for different seasons to prevent sickness, what to look for as indications of sickness, etc. These preventative measures can increase one's life cycle and will do the same for machines.

Before we get into Measures (DMAIC) let's briefly go over Implementation (DMAIC).

Since TPM has three goals (Zero Product Defects, Zero Safety Accidents, and Zero Equipment unplanned failures (Loss)) and uses a learning organization method of implementing this, we will look at some of those:
 1. Create a clear business culture of continuous improvement
 2. Standardize and create a systematic approach to maintenance and a production system
 3. All departments are responsible for quality and productivity, changing from a reactive workplace to a predictive one
 4. Create a transparent multidisciplinary organization with the goal of zero loss
 5. Remember that steps are taken as a journey and not as a race (change is psychologically difficult). Best way to achieve a Kaizen environment is through training and standardization.

Here are the 8 pillars of TPM:
1) Efficient Equipment Utilization - This involves efficient worker utilization and efficient material & energy utilization.
2) Planned Maintenance - It focuses on Increasing Availability of Equipment & reducing Breakdown of Machines.
3) Initial Control - Establish a system that allows for minimal run-up time in new production and use of new equipment.
4) Education & Training - Formation of an autonomous work force, who have the skills and techniques for autonomous maintenance.
5) Autonomous Maintenance (Jishu Hozen) - It means, "Maintaining one's equipment by oneself". There are 7 Steps in & Activities of Jishu Hozen, which we can go over in a different blog if you want it.
6) Quality Maintenance (Hinshitsu Hozen) - Quality Maintenance is establishment of machine conditions that will not allow the occurrence of defects & control of such conditions is required to sustain Zero Defect.
7) Office TPM - To make an efficient working office that eliminates losses.
8) Safety, Hygiene & Environment - The main role of SHE (Safety, Hygiene & Environment) is to create Safe & healthy work place where accidents do not occur, uncover & improve hazardous areas & do activities that preserve environment.

The period of time between P and F, commonly called the P-F interval, is the window of opportunity during which an inspection can possibly detect the pending failure and resolve it through corrective action. This allows equipment owners to inspect to detect, detect to correct, and correct to perfect.

The next thing to any program is how are you going to measure it? (Remember everything you do should be to the DMAIC methodology). I know this next part may seem boring but it will all become abundantly clear when we do a couple examples.

There are some measures that work well with TPM and they are:

Overall Equipment Efficiency (OEE) is measured by A * PE * Q
Mean Time Between Failure (MTBF)= (Sum(start of downtime-start of up-time))/Number of Failures
Mean Time to Repair (MTTR) = Average of Repair time and (MTBF + MTTR = 1)
Availability: It's the proportion of time the item is ready for use = A= MTBF/(MTBF+MTTR)
Failure Rate: It is the average number of failures per time period, which is measured as the number of failures per unit time. =  The number of failures/(Number of units tested * Total length of Time).
Finally, RELIABILITY is R = e^lambda t or 2.718 (natural base logarithm) ^ failure rate * length of time period.

Now for some examples:

Let's say you are the newly appointed director of maintenance and the VP is concerned with the downtime of 48 basketball machines. Maintenance records show that the MTBF is about 80 days on each machine and the MTTR is 1 day. What is your availability, Failure rate, and Reliability numbers to give to the VP to make him/her happy?
Answer: A = 80/(80+1) = .988 or 98.8% of the time; FR = 1/80 = .0125 failure per day per machine
             Reliability = 2.718^-.0125(80) = .368

The curve above represents the Failure Rate curve. It shows during the early period of operation, called the burn-in, debugging, or infant mortality period, the fail rate is high. The is often caused by cheap materials, lack of quality, or just bad workmanship. In the next phase, the failure rate is lower and is stable (Constant Failure Rate). The last stage is the wear out period where failure rates are on the rise. We could even depict the wear out period in the Bell shaped curve under a normal distribution (All we have to know is the probability of remaining life and variance).

Another example is:

If you are the facility manager and you need to make a decision on which type of light bulbs are the best to use between conventional versus energy star, how would you do it?
Answer: We would do a simple comparison.
Number of Units = 200        Electricity Rate = .0113    Hours used per day =  8
                                             Energy Star                                      Conventional Unit            
 Initial Cost per unit                 $3.40                                                       $.60
 Wattage (watts)                       15 Watts                                             60 Watts
 Lifetime (Hours)                     10,000                                                    1,000
                                                Annual and Life Cycle Costs and Savings
Life Cycle Costs                                       Energy Star    Conventional    Savings
       Energy Costs    
Energy Consumption * Rate = 8760*.113 = $987           $3,949            $2,962      
Energy Consumption (kWH)
                  (200*8* 365*15)/1000 =     8,760 Watts     35,040W         26,280
Maintenance Costs
               @$20.00 an hour at 15min   = $0  (20*.15)*200*10 =$6000  $6000
         Purchase Price                                  $680                $120             (560)
  Total per 10,000 hours                          $1,667              $10,069          $8,402
We would save $8,962 per life cycle! ((10,000/8)/365) = 3.42 years so roughly $2,457 annually

The final Example is Total Productive Maintenance in an nutshell:

First I will go over some brief terms and formulas then a brief description and finally will give you the last problem to test your TPM and Preventative Maintenance Management Skills.

Here are the terms and equations you will need to solve most TPM problems. Take these terms and apply them in situations of comparison and goals.


Overall Equipment Efficiency (OEE) is measured by A * PE * Q
Availability - Scheduled time - 30 min break = 450 min. Unscheduled downtime is 60 min so now its 390 min. So Availability is 390 available min/450 scheduled min = 86.7%
PE is performance Efficiency and measured by Rate Efficiency (RE): actual average cycle time is slower than design cycle time because of jams etc and Speed Efficiency (SE): Actual average cycle time is slower than design cycle time because it running at a reduced speed.
In all, Performance = (Parts produced*Ideal Cycle Time)/Available Time
Q = quality rate or yield. What is the percentage of good parts out of the total produced? If 242 units started to produce 230 good units the percentage would be 230/242 = 95%

Example: If a given work center experiences Availability at 86.7% with Performance at 92% and quality at 95% the OEE is 76.6%!!
This is one of the KEY PERFORMANCE INDICATORS in LEAN MANUFACTURING. It is a measurement that I find very useful in both facility management and production planning.


It is important to know that OEE is Heuristic (speedy way to find a good enough solution) function because its not a perfect measure due to some organizations place precedence on quality and also OEE doesn't work well with Asymmetric costs or data. For example 20%*80% = 16% while 50%*50% = 25%. I like this because I think we should be punished for having so much variation in a system. Standardization is a must!


Ok now get mentally prepared for this and see how you might apply this to everyday life as well as your professional one: Here it is...

????Final Preventative Maintenance Question????

Question:

MTBF = 34 hours and at present the machine is only repaired when it fails at an average cost of $50. The company is considering a preventative maintenance program that will cost $30 for each inspection/adjustment and there is 1 shift per day. Should this be done? If so, how often? (Hint: Assume 260 working days per year)

Answer: (Don't look at the answer until you try and calculate the answer for yourself. You should be able to do it with just the information I have given to you in the question and applying the lessons you learned today)


1. Compute average annual cost. Assuming 260 working days per year and 1 shift per day, there are 2,080 hour of available time.
2. If MTBF is 34 hours, then we expect 2,080/34 = 61.2 breakdowns per year.
3. Hence, Annual cost will be (61.2)($50) = $3,060.
4. Now, suppose we inspect every 25 hours or 2,080/25 = 83.2 times per year
5. This results in (83.2)($30) = $2,496
6. Now we should add probability of time between failures to the computation
7. TBF                                        Probability
    27.5                                         .2
    32.5                                         .4
    37.5                                         .3
    42.5                                         .1
8. Using this chart that is given by most manufacturers on equipment and if purchasing new equipment it should be insisted upon. We can come to some conclusions.
Time Between Inspections                              Total Costs (including Inspection and Failure Costs)

 25                                                                  2496
30                                                                   2773
35                                                                    3564
40                                                                    3900
Therefore, the answer is we should start an ambitious preventative maintenance program at inspection every 25 hours!

Let me know if you have any questions regarding today's blog, as I know I threw a lot of needed information at ya. Remember, you can easily replace body with machine and do the same thing.

Ja Matta ne and All the best,

Scott Lager, M.S.


The Next Blog: Inventory Management!

Write me an e-mail or comment on future comments. I already got a couple e-mails saying they are interested in Theory of Constraints. Let me know :)