Availability x performance x quality

Martin Ashcroft and Thomas R. Cutler report on the key plant floor performance indicator ÔÇö overall equipment effectiveness. The ability to incorporate real time manufacturing data into operational decision-making is important in the quest for operational excellence. Plant floor analytics can be used to drive and sustain lean manufacturing improvements, proactively manage variances from planned goals, and achieve visibility into manufacturing operations to identify strategic improvement opportunities.One of the most useful but misunderstood tools in the improvement process is overall equipment effectiveness (OEE). OEE can be a simple metric to identify the current status of a piece of equipment, or a complex tool that allows manufacturers to understand the effects of a number of variables on the entire manufacturing process. According to Hellen Budaya-Pileski of Shoplogix, a leading provider of real-time integrated software solutions: ÔÇ£OEE provides a framework to track underlying issues and root causes. It also provides a framework for improvements in the manufacturing process.ÔÇØ OEE is a combined measure of three key production parameters, availability, performance and quality. A perfect machine (with an OEE of 100 percent) would operate at maximum speed, with no downtime, without producing defects. There is, of course, no such thing, so OEE compares a machineÔÇÖs effectiveness against the mythical perfect model by combining the losses incurred when it is not available when needed, not running at maximum output and not producing first pass quality output. Hence, OEE = availability x performance x quality.Availability refers to the machine being available for production when scheduled (and only when scheduled). Availability is not measured against the machine running 24 hours a day, unless it is scheduled to do so, so lunch breaks and planned shutdowns are excluded from the efficiency analysis, but setup time, changeovers and adjustments are what the measure is all about. When a process is running it is (or should be) creating value for the customer. Whether due to setup, mechanical failure, lack of raw materials or operators, when the process stops it creates cost with no associated value. By comparing actual run time with scheduled run time, the availability component of OEE determines lost production due to downtime. The performance measure identifies production lost when the machine is running at less than optimal speed, by comparing actual cycle times against the ideal. This takes account of anything that keeps the process from running at its theoretical maximum speed, including component jams, operator inefficiency, and equipment wear. Finally, quality takes account of the time wasted by producing something that does not meet quality standards (rejects and rework). The percentage of good pieces to total pieces made, becomes the quality measurement.An overall OEE figure emerges when the three elements are multiplied together, and a world class OEE is generally accepted to be 85 percent. That doesnÔÇÖt sound too strenuous until you realize how the multiplier works. If all three elements were 90 percent, the overall OEE would be only 72.9 percent (90 percent of 90 percent of 90 percent). In practice the accepted world class standards are not uniform across the three factors, with world class availability being recognized as 90 percent, speed 95 percent and quality 99.5 percent to achieve an overall 85 percent.OEE provides data specifically about the manufacturing process, and as such is but a tool to identify the current state. Where overall equipment effectiveness really comes into its own is when the results of OEE calculations are used to compare the performance of manufacturing cells, an assembly line involving a number of machines, or individual production shifts, as a part of a continuous improvement program. The first hand experience of companies that have implemented OEE technology solutions demonstrates its effectiveness.Benlan, a medical device manufacturer specializing in molded and extruded disposable plastic products for the healthcare market, had issues with its packaging machinery, which experienced frequent and lengthy downtime when the film needed replenishing. The equipment could idle for 10 or even 15 minutes in each replenishment period, waiting for a mechanic to come to perform the operation. Since the implementation of Shoplogix Plantnode, an automated alerting function notifies mechanics by email or on their PDA when the packaging materials are nearly depleted. The subsequent reduction of replenishment downtime has helped to increase production through the machine by between 20 percent and 25 percent. ÔÇ£Plantnode pays for itself within a few months,ÔÇØ said Tom Enns, president of Benlan, ÔÇ£purely on the labor savings associated with the machine replenishment. When you factor in the increase in production, and our ability for improved capacity planning, itÔÇÖs easy to see the value of this product.ÔÇØ The OEE solution, he said, also gives ÔÇ£good intelligence about how many of each type of product we can put through this machine in a given timeframe.ÔÇØAt the other end of the packaging spectrum, Shorewood Packaging, an International Paper company, wanted to improve efficiency on the shop floor, and engage its employees in the process. Shorewood was able to identify machines that were ahead or behind in productivity but needed to find a way to provide more job specific information to machine operators, and give them the ability to set efficiency goals and affect change on a real-time basis. In August 2004, Shorewood implemented Plantnode on all machines in its Midland Avenue facility in Toronto. The company translated key performance metrics into dollar terms, to help machine operators understand the impact they could have on the profit or loss of any given job. Now, employees can access information on estimated make-ready time, job run time, and quantity to run; and compare these job estimates to dynamically updated performance information on their machine and shift. ÔÇ£Tying our performance metrics to the bottom line has increased systematic thinking on the shop floor, sponsored a culture of continuous improvement, empowered our employees, and fully engaged them in their work,ÔÇØ said engineering maintenance manager Neil Pierobon. Fourmark Inc., an Oakville, Ontario-based contract manufacturer of custom injection molding for a variety of industries, made use of the same technology during a period of rapid growth. Increased demand meant Fourmark needed to add more injection molding machines to improve capacity. But the rate of growth meant it began to lose control of its production output. The company needed a way to capture data from machines on the plant floor to understand how they were operating, and how well each shift was performing. Use of technology has identified opportunities for improvement by capturing downtime information related to color changes and mold setups, and scrap rates. This has resulted in a reduction in mold setup times by 30 percent, and has increased overall plant output by five percent to ten percent. The company is also using the proactive alarming functionality of Plantnode, which can signal various individuals in the plant and office if something happens to any of the equipment. We said at the beginning that overall equipment effectiveness was useful but misunderstood. There is a health warning attached to the use of OEE figures, just as with any other kind of measurement. Care should be exercised in the use of results to avoid their joining the ranks of lies, damn lies and statistics, under the sponsorship of political factions within the organization. While it is generally beneficial to improve the overall OEE figure, it is worth remembering that because the final calculation involves three elements which often affect one another, it cannot be assumed that a higher OEE will always align with the companyÔÇÖs strategy and goals. Consider the example of two production shifts in the same organization. At the end of the day shift, figures show overall equipment effectiveness at 81 percent ÔÇô not bad, but not quite world class. The night shift which follows it returns an OEE figure of 86 percent, and they sure like to let everyone know about it. When you look at the breakdown of the figures, however, the day shift achieved 90 percent availability, 90 percent speed and 99.5 percent quality (90 x 90 x 99.5 rounds up to 81 percent). The night shiftÔÇÖs 86 percent comes from a steady 95 percent for each factor. So you have a little work to do to improve availability and speed on the day shift, but if you are a supplier of highly engineered parts to automotive, aerospace or any other quality sensitive customer, your biggest problem is likely to be the reject rate of the night shift, where the cost of scrap might be prohibitive. The moral here is that manufacturers should never be blinded by ÔÇ£pureÔÇØ figures, but use them to identify the issues that prevent them from achieving their objectives. When it comes to calculating an OEE figure for a whole plant (if this is what you want to do), a straight average can be used, where you add the results for each factor separately for each process, then divide by the number of processes, before multiplying the results to get the overall figure. Depending on company strategy, however, there might be more value in producing a weighted average, to focus on quality or efficiency, or some other business priority.