THE REAL WORLD: OEE
UNDERSTANDING AND OPTIMIZING PACKAGING LINE PERFORMANCE USING OEE (OVERALL EQUIPMENT EFFECTIVENESS).
By Paul Zepf
One of the world’s largest consumer products companies was recently having serious production difficulties with a well-known hair care product and decided to improve their production line. Spillage of the product was the major problem; since it was mildly corrosive and required time to thoroughly clean up after a spill. The bottle has an over cap that blends into the body of the straight walled oval bottle. The number of people on the line has been reduced from five to four people, with workers positioned at the unscrambler, filler, capper/case packer and case erector. They rotated every two hours.
The production line was relatively new (less than three years old) and had been marginally efficient from the start, but recent performance decreases were cause for concern.
The case packer and case sealer were becoming more problematic, with the case sealer’s performance deteriorating especially rapidly.
The company set up metrics to determine the problems and their extent. The plant concentrated on recording manual downtime data, while a consultant concentrated on disturbance frequency analysis, Mean Time Between Failures or stoppages (MTBF) and Mean Time To Repair or correct a jam or fault (MTTR) – which aren’t common tools used in the packaging industry.
A review of the before downtime data (shown below) indicated that the case packer and case sealer were the main contributors to the losses. They also contributed to the inability to maintain a reasonable schedule. The OEE value as a benchmark showed that this line had a very low performance compared to sister plants and other plants running similar products. The consultant’s data appears to be contradicting the downtime data and a lively discussion resulted. Eventually, the consultant’s approach was dismissed since the techniques were foreign and incomprehensible to the plant.
If OEE = Availability X (1-Scrap) X (1-Rate Loss)
Then Availability = OEE / ((1-Scrap) X (1-Rate Loss))
Therefore Availability is 56.15 percent. If comparable companies had production lines with an OEE of 70 percent what does this reveal about the present production line?
If the industry average MTBF is 15 minutes, what does this say about the problematic line?
In reviewing the before information engineering decided to purchase a new case packer and case sealer. Do you agree with engineering? Before reading the solution, review the before data and formulate your own conclusions about what you would do.
The consultant’s approach initially caused some concern but in the end resulted in questioning the validity of the data. It was discovered that:
The manual downtime data collection missed at least 40 percent of all downtime events. In particular, the data missed events that covered less than one minute. The recorded downtime did not reconcile with the calculated OEE using actual produced product. It was later found that the operators were fudging the numbers to get them to reconcile.The downtime is only an indication of lost time at a point where the time was lost not necessarily at the point of the cause.
Engineering, maintenance and operators had different points of view on the causes and solutions. Few people really understood the entire production process and inter-relationships. All agreed the OEE was correct and representative of the overall effect of operations. Scrap and rate losses were poorly documented and therefore the OEE was probably less.
The consultant’s data and approach projected the real causes of downtime and said that the case packer and case sealer were not major contributing elements. There were also minor issues related to set up and machine abuse.
If the industry average MTBF for a machine is 15 minutes, it says that the case packer exhibits a problematic profile, but does not say that it is the problem. Actually, any MTBF of less than about 30 minutes is very poor. Using 30 minutes as a yardstick indicates that all machines display effects from causes.
Engineering would be wrong in deciding to purchase a new case packer and case sealer as the problem is back in the capper. When one looks at disturbances and downtime and they do not agree it is usually the case that the upstream machines have more disturbances that migrate downstream to impact downstream operations.
THE CAUSES WERE:
1. An input material problem. The overcap was on the plus side of the tolerance and overhung the body of the bottle, which was on the minus side of its tolerance. This catch point caught the bottle and pulled it up in the dividing bars of the drop packer. In some cases the case packer did not jam but sent the pulled bottle–which was now on top of the open case–and was crushed in the case sealer.
The solution was to have the overcap on the minus side and the bottle body on the plus side as well as taper athe case packer dividing bar’s sharp edges.
2. The capper sorter was missorting and had difficulty in delivering a steady stream of caps. The operator had a long pole that solved this problem. The capper sorter was modified to reduce inverted caps and fat caps were eliminated by going to the minus side of the cap OD tolerances.
3. The unscrambler issues were set up by bent components and hammer marks from the misuse and improper set up of the machine. The solution was training, machine re-alignment, knobs for adjustments and the removal of hammers.
4. The case sealer could not deal with a bottle lying on-top of the case and the vast majority of the time no operator was close enough to the case sealer to monitor it. On the infeed of the case sealer a vision system was installed to look for pulled up bottles and proper case load and a case reject before the case sealer.
All these changes plus training and set up sheets cost the company under $150,000 dollars and were completed in less than six months. The After data reveals that the payback took only two months. No capital was required and no disruption of production resulted during these changes.
This example demonstrates that most production lines have nagging little issues that can cause significant problems. Since companies do not have the proper metrics to find causes and undertake effective and efficient action, it’s advisable to question the abnormal, rather than consider it to be normal, anticipating problems before things take a turn for the worse.
Paul Zepf, co-founder and director of engineering for Zarpac Inc., has over 34 years of packaging production experience.
|
OEE AND THE BOTTOM LINE Translating a customer’s needs and having the agility, flexibility and infrastructure to deliver quality goods earn packaging production process profitability. This requires a vision and solid planning. There are various levels of tools that can be helpful to a production manager to assist in helping identify areas of opportunity and to point the way towards improvement. They are in order of simple to complex: 1. Package run cycle (PRC) charting - (Quick Picture) There are various forms of PRCs, from the simple, to the overall output per period in sync with the state of each machine in the production process. For lines that run poorly, this is the best view to make quick overall decisions to pick up the low hanging fruit. 2. Utilization, speed, wastage, materials, labor, planning and scheduling - (Pieces of the Picture) Utilization refers to times or comparisons using time. There are various forms of utilization such as rate, element, system, change out, run and asset. Speed refers to rates of production in products per minute or cases per minute, but there can be up to seven different types of speeds and each will alter your perspective of the line and its design. Hierarchy of Speeds: 1. Design speed (DS or s) Machine Cycle Rate Thus, DS>c>Dm>D>ds>de>y per unit of time The intent of Rate Utilization is to measure the degree to which the full rate potential (Ideal Rate) of the unit operation is being utilized based on the rate (Target Rate) at which the organization has chosen to target the process to operate. This is also called the speed factor. Wastage is a measure of scrap caused by the production process itself. There is also the scrap peripheral to the production process that needs to be understood and accounted for separately. Some people use the word yield for wastage or scrap. Yield is basically what came out compared to what went in (or output over input.) For single product production lines for products such as paper, converting, ejection molding and blow molding this is simple, but on packaging lines, the number of inputs from bottles, caps, labels, cartons, etc. causes the problem to become more complex. In a lot of cases the losses from non-critical inputs can exceed the losses from the critical input. Labor is the direct labor to run the production process and the indirect labor in supporting the production process. Today labor is stressed too much–especially in automated lines. In automated lines, the level of labor is so low that reducing it could reduce the reliability of the production process. Why chase labor for five or six figure dollars when wastage may be in the six to eight figure dollars? Planning and scheduling are difficult activities made more complicated by demanding and changing customers. Nevertheless, if the production process is not put into control and has high OEE values, then there is little hope of keeping a schedule. Scheduling is made up of a group of activities in which the timing and sequence of production need to occur to ensure the requirements of the customer are met. It is the interplay of timing, machinery and labor. The intent of the Schedule Utilization measure is to indicate the percentage of available Calendar Time that is used for routine production, maintenance and improvement activities. These areas give us a picture through a particular filter to view aspects of the production process and how those areas are functioning by themselves even though they are inter-related. 3. UEE (utilized equipment effectiveness), OEE, Availability, Process Reliability, Performance Index, Capacity Utilization - (The Picture) These measures give the overall big picture number of the reality of the production process. In and of themselves they are not helpful in improving the production process. They only give an overall value of the reality and assist in benchmarking these values against a target or other similar production processes to see improvements. They trigger the decision to improve if values are below specific pre-determined and calculated levels. Once the decision to improve is given, then the production personnel need to dive deeper into the pieces of the pictures and maybe even deeper then that to find the root causes and develop an ROI and action plan. Availability is intended to measure the ratio between uptime and total ‘operating’ time. Simply put, availability is the fraction of time the machine is in an operable or operating state to the total production time period defined. An operable state is the condition that allows the machine or system to function at its achievable run speed (Ideal Rate) in a manner that produces an outcome or an assemblage of inputs within the stated specifications and conformance to customer's needs. The bottom line is to buy a machine or system that can contribute or support high availability or high MTBF and low MTTR that yields the lowest TCO (Total Cost of Ownership), the best Value and the lowest Life Cycle Cost. 4. OEE = Availability x Quality x Performance Process Reliability (PR) is the more correct form of OEE for Packaging Production Processes. The intent of the OEE or Process Reliability is to measure the extent to which scheduled time invested in production, maintenance and improvement activities is converted into net production. Scrap is only calculated in input materials or products removed from the production line during running. Scrap removed as a result of a downtime is not used in OEE but is kept to determine overall wastage. Rate fluctuation is determined from only the running period of production. The overall Production Loss Summary chart is the most effective intermediate tool to use in overseeing the overall nature of a given production process. It can pinpoint areas for focus and improvement and quantify them. This is the simplest and most powerful tool to use, but most packaging companies are unaware of it and how to use it. 5. Capacity Utilization (CU) – CU is the measure to indicate the degree to which a purchased and installed asset is utilized. In other words, the CU calculation determines how close a given system is to maximum production capacity. Benchmarks are: 1. Comparisons of performance against internal baselines or targets. One cannot use benchmarks unless they are apples to apples comparison and the terms and definitions used to get whatever metric is identical. Focus on establishing the best practices to produce the highest quality at the lowest produced cost. And what is good business practice? It is only using those tools that enhance an operation in becoming the lowest cost producer for the highest quality product within a given expertise and customer need. Everything else is a fad. Accounting and report data are usually not the information and knowledge required to improve the production process. Data is useless by itself. Only data automatically converted to information at worst, and knowledge at best, is the key to becoming world class. Paul Zepf, co-founder and director of engineering for Zarpac Inc., has over 34 years of packaging production experience. |
