Applying principles of lean to a calibration lab?

[ScottBP]

The competition among all the commercial calibration labs is turn-around time, and it seems that many principles of lean manufacturing could be applied to a calibration lab to speed up the turn-around time.

The question is, how would you go about applying the principles of lean to a calibration lab? I know most manufacturers have an in-house cal lab or at least an instrument and control (I&C) shop, so I'm pretty sure the Kaizen gurus hit up the cal lab or I&C shop at the same time they're evaluating the rest of the company.

But the lab managers of most 3rd party cal labs have never worked in the manufacturing industry, where lean production control skills are learned and ingrained. I've been told that calibration software is needed to take care of these tasks.

We are looking at utilizing Fluke MET/CAL to speed up our electrical calibrations, because it can automate a lot of electronic calibrations by controlling the calibrators via IEEE-4888 interfaces, and then collecting data from the instruments being tested via communications ports. (We are currently using a manual data collection system.) Automating calibrations makes it mistake-proof, and by automatically flagging typos, incorrect decimal places and out of tolerance readings, the number of QA mistakes are dramatically decreased. Besides that, being computerized, it whacks the paper shuffle off at the knees.

But MET/CAL is geared mainly towards electrical calibrations, and we are a multi-disciplined cal lab. You can't connect gage blocks, torque wrenches, glass thermometers or analog pressure gauges to a computer, so for these things, calibration data would still need to be entered manually. For that, there are numerous other softwares like Blue Mountain's Calibration Manager, IndySoft's Gage Insite, etc., which we are also looking at.

What other things could we look at "leaning out" at a cal lab? Should we start out with a process map that gets all down into the nitty-gritty of every step along the way from the time we receive a piece of equipment from a customer to the time it's back in their hands? Anybody else have experience with leaning out a cal lab?

[gard2372]

I'm not sure of the capabilities, but have you tried SIMCO or GageTRax I'm not affiliated with them and I'm not sure if this even helps your situation.

[AndyN]

From my experience of Lean principles, and lab operations, there are many tools and techniques that can be applied in a laboratory which would speed the turn around time. Have you studied any of the tools of Lean, like 5 S, or spaghetti diagrams to evaluate 'travel' etc.?

Andy

[sonflowerinwales]

I came from a manufacturing background into a lab, and the most important thing I did was to motivate the staff. To say they were laid back would be an understatement! I then made sure boooking in, putting equipment in the lab was a easy, (thus quick) as possible. After cal, I encouraged enthusiasm in getting the product out the door. We were limited to a weekly run, so in on a Tuesday, deliver on a Tuesday. The thorn in my side was UKAS , they say in 17025 something like "no commercial pressure should be made", effectively blocking the "get your finger out" form of encouragement. We used specially written gauge control software, but data for all the kit was manually input.
Paul

[Steve Prevette]

Originally Posted by ScottBP

What other things could we look at "leaning out" at a cal lab? Should we start out with a process map that gets all down into the nitty-gritty of every step along the way from the time we receive a piece of equipment from a customer to the time it's back in their hands? Anybody else have experience with leaning out a cal lab?

Yes, I would strongly recommend laying out the process map with all that nitty-gritty. And in addition, collect cycle time data for each of those stages from actual operations and either queue times waiting for the next step or inventories waiting for the next step. Then you can apply Theory of Constraints to the data - work to expand the choke points - and get the cycle times down. My experience in doing this (before I knew what "lean" or even "theory of constraints" was) was at a US Navy torpedo maintenance facility.

[Hershal]

Actually, some of the military is looking at 6S and lean at least for the engineering centers.....they have tons of statistical information so the transition should be a bit easier for them.....commercila labs will find the effort more challenging as the statistical information may not be as easily distilled unless you have a great track record of detailed uncertainty calculations, hence the efforts may take longer, but it is doable.

Hope this helps.

Hershal

[gard2372]

Originally Posted by sonflowerinwales

I came from a manufacturing background into a lab, and the most important thing I did was to motivate the staff. To say they were laid back would be an understatement! I then made sure boooking in, putting equipment in the lab was a easy, (thus quick) as possible. After cal, I encouraged enthusiasm in getting the product out the door. We were limited to a weekly run, so in on a Tuesday, deliver on a Tuesday. The thorn in my side was UKAS , they say in 17025 something like "no commercial pressure should be made", effectively blocking the "get your finger out" form of encouragement. We used specially written gauge control software, but data for all the kit was manually input.
Paul

I'm thinking the "no commercial pressure should be made" represents the effects of turnaround scheduling conflicting with the integrity of the calibration results. Kind of like the old adage one should not inspect something one has just worked on. Perhaps if you revised your QMS to state new turnaround processes for calibrations and allow an acceptable time period that would allow for "motivating" your highly skilled techs. Therefore the pressure will be coming from inhouse versus comercially.

[ScottBP]

Thanks for the quick responses... I've been gathering info on 5S, Kaizen, etc. all along, and am in the process of coming up with some sort of game plan to get the ideas across to management, so we can "re-balance" (or re-organize) our cal lab. Of course, you can't rush a good calibration itself, but my line of thought is that if you take care of the little things, for instance knocking down the paper shuffle with a better automated system or making repair parts more readily accessible, then the work flow is smoother, and when work flow is smoother, there is less stress to put out work in the quoted turn-around time, and less stress means fewer mistakes, fewer mistakes means lower uncertainties, etc.


Memorable calibration quotes:
"Until a weight set is calibrated, it's just a box of metal."
"Without data, you're just a person with an opinion."
"Without traceable data, you're a person with a questionable opinion."
"In God we trust, all others bring data."