|
|
The Hype about OEE
March 31, 2008
In response to last month's post OEE is Like Sailing, a reader questioned why all the sudden hype about a technology that is so old. I think that the answer is that it's not about the technology, it's about the business. And sometimes the business performance is so low, that technology isn't the starting place for a solution.
More than 10 years ago, I participated in benchmarking that convinced me that packagers could implement OEE on a uniform basis to optimize operations in plants around the world. One European-based global beverage processor which I visited prominently displayed up-to-date OEE's for every line in the company throughout their engineering offices. Managing OEE was part of the culture.
But many packagers, especially in the US, are reluctant to apply any new technology until some business crisis drives them to it. Packagers today are facing crisis. Consumer packaged goods plants are trying to improve their performance metrics to keep operations from being sent to co-packers or, worse yet, to plants in Mexico or China. Pharmaceutical companies are feeling pressure from many sources and are now focusing on previously ignored areas such as packaging line performance.
Pharmaceutical company packaging OEE's are reported to be among the lowest in the industry. Experience teaches that those feeling the greatest amount of pain are the ones most likely to successfully drive change. Perhaps that is why three giant pharmaceutical companies reported on their efforts in this area at ARC's Forum on Winning Strategies and Best Practices for Global Manufacturers.
Bristol-Myers Squibb described development of what was described as a fully automated paperless manufacturing operational philosophy. The manufacturing execution system (MES) architecture included integration schemes for documentation, training, laboratory management, calibration, building management and shop floor systems. Shop floor integration resulted in creation of electronic batch records via S88 and S95 compliant integration of distributed control systems (DCS), a plant data historian, and an SAP enterprise resource planning system (ERP). Although packaging systems aren't typically integrated through a DCS architecture, the packaging events are considered part of the batch record, so packaging data is currently integrated manually.
Pfizer described a similar program that began with a focus on integrating batch analysis and laboratory information management (LIMS) records but shifted emphasis to packaging line Overall Equipment Effectiveness (OEE) as a result of management concern for performance. The resulting Packaging Excellence Project was implemented in Latin-American packaging operations where paper-based OEE systems had been in use. Rather than implementing Pfizer's Pfind-IT system with automatic data collection, a lite version of the system was implemented that uses manual data entry through 3 interface screens. Data is collected hourly for the OEE analysis.
Eli Lilly was the third pharmaceutical company that reported on the use of OEE in packaging operations. In their plants, data is collected manually on 75 – 80% of the lines with the remaining using some form of automatic data collection.
These three reports confirm what many of us know: that automatic collection of data from packaging lines to calculate OEE is difficult. But as OEE expert Paul Zepf pointed out, you really don't need automatic data collection until OEE's exceed 75%. This is one of those profound statements that should be so obvious that most of us will miss it. I did.
Until a line is routinely operating at an OEE of 75% or higher, which most aren't, OEE isn't a technology problem for engineers at all - but it is a huge opportunity for those running the business.
TrackBack
TrackBack URL for this entry:
http://ontheedgeblog.com/blog-mt1-mt/mt-tb.fcgi/73
 |
Comments
Just a clarification on the paragraph below.
"These three reports confirm what many of us know: that automatic collection of data from packaging lines to calculate OEE is difficult. But as OEE expert Paul Zepf pointed out, you really don't need automatic data collection until OEE's exceed 75%. This is one of those profound statements that should be so obvious that most of us will miss it. I did.
Until a line is routinely operating at an OEE of 75% or higher, which most aren't, OEE isn't a technology problem for engineers at all -but it is a huge opportunity for those running the business."
At the ARC conference I unveiled the OEE Roadmap. To simplify things I said if your OEE is lower than 55% then good ball park values can give a good picture of the low hanging fruit. From 55 to 75 it still can be done because there is still low hanging fruit but it requires a little more insite and little more data. Above 75% your best guesses will be more wrong than right and now require a deeper dive into data and analytics. The right data from a good data acquisition system is always helpful and simplifies work regardless of the OEE level. Unfortunately, most data acquisition systems have the wrong OEE definitions and the data is massive and convoluted and unfortunately at low OEE is just as good as your best guess from quick line reviews. I did not want to get into talking about data acquisition software and wanted to simplify my approach to companies having nothing set up. Most companies need to understand their production lines first before being able to select a good data acquisition system. One of the ways to get going was to start on a simple roadmap which starts you going. Once going and improvements progress more data and insite is required and a data and analytical system needs to be put in just before it is required so that the improvement program can continue without stalling out.
To start you only need about two dozen pieces of data based on a week of production. Multiple SKU runs require aggregate grouping but can be done. A good data acquisition system needs to have the proper OEE definitions, packaging knowledge embedded into it and has to be validated. Only one package that I know of fits that specification and therefore is helpful anywhere along the road to success and can expand to meet your needs now and into the future right up to World Class.
I hope that clarifies a few things. The idea is to start today with or without tools (your choice) but you will eventually have to follow up with getting the tools you need just before you need them in order to keep improving. Smart companies justify the data acquisition system in the low hanging fruit phase so that they can implement the system at the best optune time.
If you are not improving every day and have no method of checking that progress you will slide into the land of has been companies.
Posted by: Paul Zepf on March 31, 2008
OEE is the most powerful metric I've seen for shop floor performance. Apply it by machine center first. When you break 75% at the machine level, measure the OEE of each routing -- e.g. measure corrugate-print-DC-glue as a single shop floor routing. You'll find that 75% OEE on each individual machines net 20% or less OEE when you consider the entire routing. Bring the routing OEE to 50% and you'll be printing money instead of boxes. Our industry focuses exclusively on individual machine performance but we really produce over a series of machines -- rarely just one -- and OEE is an excellent metric for understanding how inefficient our routings are.
Posted by: Ken Rohleder on March 31, 2008
Let Keith Know What You Think: We supply pharmaceutical industry with packaging and are feeling the effects of the examples you describe. The drive for defects reduction in our products are coming from all sides, and a big thing now is that customers are talking about WW application of "Japanese" quality levels in their buy specifications. This of course means that there are no longer any "minor" defects. At least some of this is being driven by efforts to improve OEE on their pack lines.
Posted by: Wesley Prais on May 1, 2008
|
|
 |
| About Keith Campbell |
| Leaders learn from the past while
looking to the future - and bring both to bear on the here
and now. This is the philosophy that has steered Keith Campbell's
30+ years in manufacturing. It has worked for him in operations,
maintenance, engineering, R&D, education, consulting and
professional organizations--and now he's putting it to work
for you--taking you to the edge of his thoughts on packaging
operations. |
|
Keith's
Travel Calendar