03.18Predicting and Preventing
My firm, Integral Concepts, Inc., and I provide consulting and training in quantitative statistical methods for optimizing product designs and manufacturing processes. It is sad that almost all of our projects involve REACTING to major problems that have surfaced which are jeopardizing production, customer satisfaction, and profitability.
We much prefer to work with our clients PROACTIVELY by using methods during product and process design and development that will ensure that the product will perform adequately both initially and over time. Methods such as Design of Experiments and Reliability Analysis are especially useful since they are based on models from physical measurements from production or prototype parts.
Even when a good, robust design is achieved, changes in the manufacturing process (raw materials or manufacturing conditions) often go undetected until a field concern arises. Thus, a potentially very small problem that is easy to correct becomes a major and very expensive issue. Basic Statistical Process Control (SPC) methods are invaluable to detect potentially harmful process changes – but they are often not utilized. When SPC is implemented, it is usually misapplied so that the benefits are often not realized.
Furthermore, the time spent dealing with preventable problems is a huge distraction to management and technical personnel. As a result, it seems that fire-fighting consumes an inordinate amount of time and prevents strategic efforts that is truly value-added work for the company. Some questions I wonder about….
What will it take for companies to pay more attention to PREVENTION rather than REACTION to problems?
Why is there so little effort on statistical/quantitative aspects of quality and reliability?
Why were the Japanese so much more receptive than Americans to Deming’s teachings about reducing variability so that our products perform consistently?
When will U.S. manufacturers realize that the highest levels of quality and reliability are necessary to compete in the global economy and that honest and serious efforts are needed (instead of flavor of the decade watered down quality programs?).
Steve Wachs
Thanks for the great post, Steve.
I share your frustration. It is so hard to get companies to look at the big picture.
Kirill Liberman
August 18th, 2008 at 12:20 pm
Steve,
The last time I spoke to Kirill he showed me a something called SPCIQ that looked very interesting. Kirill basically described SPCIQ was a way to quickly deploy SPC with out the usual training expense and learning curve. He also emphasized the updated “math” that was designed into SPCIQ. Have you looked at or reviewed SPCIQ? What are your thoughts about it? How effective do you think it can be? What do you think about the “modern math” claim? Is Minitab the only way to go?
Mike
August 26th, 2008 at 4:22 pm
Hi Mike,
Kirill and I collaborated on the development of SPCIQ – so Yes, I’m quite familiar with it! A few comments:
SPCIQ is a great tool to get started using SPC methods correctly and without the frequent over-simplifications. Using flowcharts and the integrated Minitab work instructions – one can “do the right thing” without a lot of training. That said, it does not replace the need for SOME personnel to have some substantial training in SPC by a qualified trainer/consultant.
Regarding “modern math” – many production processes are not conducive to the assumptions that are required to use standard X-bar and R charts. As a result, the charts don’t perform adequately in these cases.
Other charts are required and SPCIQ addresses WHEN and HOW they should be used. Many (or most) quality personnel have not even heard of these newer charts despite their need for many situations.
Another big concern is assuming normality of data for process capability. The common use of formulas that assume normality is incorrect and misleading and provides a false sense of security in most cases.
The SPCIQ product does not cut any corners…Fortunately, Kirill was able to help organize the many options and decisions into something that is coherent and easy to follow.
Finally, MINITAB is one option…but there are many other options especially for real time data collection and chart creation…BUT – many programs out there allow you to do silly things so it’s a good idea to know something about SPC before using the software.
August 26th, 2008 at 5:36 pm
Thanks, Steve. Kirill mentioned that SPCIQ was a joint effort. I guess he meant you. Digging around on the website and looking at some of his literature I see that your company is mentioned.
Kirill also pointed out testing data for normality and a couple of other steps (sample size/frequency and possible differences between locations) that I thought were interesting. We don’t do this at my job or at my previous employer. You probably run into this a lot.
The reason I asked about Minitab is that it is very expensive. I work for a relatively small company and the current business climate is making spending money very difficult. I like SPCIQ, but when I add Minitab cost to it, it makes it a harder sell to my management. I looked at some Excel programs, but SPCIQ is not compatible with them. Any plans to make an Excel version of SPCIQ that is more affordable?
Mike
August 28th, 2008 at 12:50 pm
Hi Mike,
Yes, Steve is the one that contributed the statistical know-how and depth to SPC-IQ. He is the statistical methods guru.
Steve and I worked very hard to make sure that SPC-IQ was robust and eliminated all of the common mistakes. We also worked very hard to make it accessible to and usable by everyone, regardless of their level of statistical process control training or know-how. Furthermore, our objective was to also make application and utilization of statistical process control a business process that can be integrated into an organization’s business management system or quality management system. However, as we discussed and Steve appropriately notes, every company should have someone that actually understands what SPC-IQ does. In fact, once a company uses SPC-IQ they realize this because they finally begin to see the real potential benefits of using SPC properly.
Regarding Minitab, I have had other small and mid size companies mention that adding Minitab’s cost to SPC-IQ make it less affordable/attractive. Steve and I have discussed this and may come up with a lower cost alternative at some point so stay tuned.
Kirill Liberman
August 28th, 2008 at 8:20 pm
Hi Mike,
I think Kirill answered your question. If not let me know!
Best Regards,
Steve
September 2nd, 2008 at 8:58 am
Nice post. I was checking constantly this blog and I’m impressed! Extremely useful info specifically the last part 🙂 I care for such information a lot. I was seeking this particular info for a very long time. Thank you and good luck.
December 11th, 2012 at 6:33 pm
Thanks you, John. Please come back soon.
Kirill Liberman
December 13th, 2012 at 8:26 am