Subject: Uncertainty Debate RE4 Date: Wed, 24 Feb 1999 10:07:29 -0500 (EST) From: Henrik Nielsen To: Greg Gogates Greetings, I agree with the points made by Howard Castrup. The fact that an uncertainty budget cannot be developed by the average lab technician does not mean that it is unnecessary or does not add value. In the attached contribution Jack Gale argues that because 95 % of the people buying calibrations are not interested in uncertainty, it should not be a requirement. I think you will find that the same 95 % would skip calibration all together, was it not for some outside requirement (such as ISO 9000) mandating they send their equipment to calibration on a regular basis. I think we all have the basic, conceptual understanding that some calibration is necessary and adds value. So we do think it makes sense that instruments and standards should be calibrated on a regular basis. Up until now we have based calibration requirements on rules of thumb that sometimes makes us calibrate more than what is necessary, sometimes less than what is necessary, but rarely, if ever, exactly what is necessary. Uncertainty budgeting is the only tool I know of that, used properly both by the user of an instrument and by the laboratory calibrating it, can provide a data-based set of calibration requrements, that will help us do the correct amount of calibration to the appropriate uncertainty without making it a question of faith or arbitrary rules of thumb. In most cases I have seen, the calibration requirements gets reduced (as in less calibration), but more focused on what really matters for the performance of the instrument or standard, when calibration requirements are based on uncertainty budgets rather than faith in rules of thumb. In other words, uncertainty budgets used appropriately saves you money. The fact that the average technician cannot design and manufacture a car or an airplane or a computer without input from people with significant theoretical education, does not make these products useless or less valuable than chuck grinders. If we want progress and prosperity, we need to move from anecdotal to data based decision making and in the world of calibration, that means using uncertainty budgets to decide what needs calibrating, how well and how often. Best regards, Henrik S. Nielsen HN Metrology Consulting, Inc. 5230 Nob Lane Indianapolis, IN 46226 Phone and Fax: (317) 377 0378 E-mail: hsnielsen@worldnet.att.net Web: http://home.att.net/~hsnielsen > -----Original Message----- > From: owner-iso25@quality.org [mailto:owner-iso25@quality.org]On Behalf > Of Greg Gogates > Sent: Tuesday, February 23, 1999 8:17 PM > To: iso25@quality.org > Subject: Uncertainty Debate > > > Date: Tue, 23 Feb 1999 18:02:49 -0500 > From: Jack Gale > To: Greg Gogates > Subject: Uncertainty Debate > > To Greg, Bruce, Howard, et al : > > I'm leaning heavily towards the Bruce camp on this issue. We have over > 1200 customers for our calibration service and only a handful could even > begin to grasp the importance of uncertainties. We are dutifully > marching towards G25 (many customers never heard of this either) and > calculating the uncertainties for our standards. Maybe 2% of our > customer base would appreciate it, even less would know what to do with > the information. My customers represent a smart cross section of > high-tech, biomed, nuclear and automotive companies, so let's not > discount the source. > > Bold statement: "Ninety-five (SWAG) percent of the calibration dollars > spent in this country are spent by people with no knowledge or interest > in uncertainties." The metrology community knows, but the large base of > users (who eventually pay all the bills) don't care. Based on the looks > I get when G25 and uncertainties are brought up in discussion, I believe > this to be true. > > We're attempting to sell filet mignon to a ground chuck customer base. > Ground chuck is fine because most people are eating hamburgers. Let them > eat chuck! It's nice that some people do filet mignon, Bentley > automobiles and Armani suits, but burgers, Fords and Dockers do for the > most of us. > > In the most real world (commercial), the customer would rather we spent > our time to improve turnaround and pricing than uncertainty. This may > seem as heresy to people with more academic pursuits in metrology, but a > huge market exists for "tell me my equipment is okay for another year" > and "get it back to me yesterday". > > I'm not saying no to uncertainty, just having it as the only choice. > > Historical Note : Early Ford Model T's were offered in many colors. It > was only after improvements in the production line and the development > of a fast drying black paint thet they were restricted to the single > color. You may also note that GM caught up with Ford and passed them in > the 1930's as the nation's leading auto producer because they offered > what the customers wanted. Let's not repeat the past. > > > -- > ******************************************************************* > Jack Gale, ASQ-CQE The opinions expressed > ESSCO Calibration Lab herein are the author's and > (800) 325 - 2201 do not reflect the opinion > jackgale@esscolab.com of ESSCO Calibration Lab. > ******************************************************************* >