|
The production tester
we developed for Passy-Muir has a lot of people asking, "What's
that wild-looking purple thing in the lab?" The stunning
anodized device looks like the high-tech nut cracker you might see in a
science fiction film. It is actually a quality control device we designed
for one of our clients that measures tolerances for tracheostomy speaking
valves. Click
to see an image of the precision testing device. 
Before a valve is
released to stock, it must be tested to assure that it meets production
requirements. Passy-Muir tests 100% of their valves before they are
released. The Passy-Muir tracheostomy and ventilator speaking valve
contains a thin silicone diaphragm that flexes, allowing air to pass
through the plastic body.
Measurements with the
precision measuring device indicate whether or not the valve will work
properly when used by a patient. Years ago, after Passy-Muir determined
the critical tolerance parameters for valve production, they hired an
automation design firm to build precision measuring instrument. At the
time, the device cost almost $70,000.
Not long ago, Passy
asked Omnica to re-think the existing machine. The original device was
built with a now-obsolete programmable controller, and if it broke down,
valve production at Passy-Muir's manufacturing facility would stop.
OMNIview interviewed
Andy Scherer, our senior mechanical engineer and head of the development
effort, to tell us how the new design evolved.
What
exactly did Passy-Muir want Omnica to do?
We have been working with them
for a long time, almost 18 years, developing their speaking valves. She
(Patricia Passy, the CEO) originally wanted us to reverse-engineer the
original tester, only make a more robust version that was both less
expensive and less likely to break down. At least make one that was as
accurate and could be easily repaired.
What
did you want to change about the original design?
When we got serious about the re-engineering, we saw that the existing
machine had some drawbacks. For one, it was an unnecessarily complex
instrument. There were a lot of precision electronics in it. They used a
stepper motor, drive screws, and controls for the force gauge. Also, the
thing was huge. It's 4 feet tall and it took up a lot of floor space. I
couldn't see us re-engineering an elephant that would still cost a lot and
continue to be a maintenance nightmare.
Describe
your thought process when you made the decision to completely re-think the
whole package.
Rex and I figured we could do better. I knew we could make it smaller and
simpler. It (the existing tester) goes through the same five-step
procedure every time, so why do you need a computer to do that? We
designed a mostly manual process with minimal electronics. This tester
does everything the other one does, faster, and with about a quarter of
the parts. It's accurate, and the results are consistently reproducible.
We incorporated a precision off-the-shelf electronic dial indicator and an
SPC printer. The printer tape gives you a hard-copy record showing that
the measurements are within tolerance.
Tell
me about the unusual design. It doesn't look anything like the original
machine.
We decided that the new tester didn't have to be ugly or boring. Tim Payne
(Industrial Design) came up with the curvy organic shape. Mark Mossberg (a
CAD/CAM specialist) programmed the CAM cutting path, and we machined it
here (in-house). Patricia figured the anodizing would match the
purple valve they are currently producing, and that it would jazz-up the
production area. It did! It's definitely different, and the whole thing
has a footprint the size of a clock radio.
Does
it do everything that the existing device does?
Yes, in about a third of the time, and at 20 per cent of the price. It's
easy to set up. Just level it and enter the tolerance range for the type
valve to be tested.
Could
you use it to crack nuts?
Kind of expensive to do that, but if you sped up the cycle and dropped the
weight a little faster. . . (Pauses for a moment) With the
precision electronic force gauge you could tell exactly how much force
you'd need to fracture a Filbert.
|
