At 09:28 AM 03/07/2003 -0800, Major Variola (ret) wrote:
At 12:52 AM 3/7/03 -0600, [EMAIL PROTECTED] wrote:
>A tiny fuel cell that detects the alcoholic breath of a drink-driver
>and calls the police has been developed by a team of engineers

Would you buy one if you're drunk? Would you put one in your trunk?


Who's the target market for this sort of thing?
Engineers can build things for the existential pleasure of it,
but usually they're trying to solve problems for people,
and it's not clear what the business requirement is here.
Did someone fund them?  Who?  Why?

Doing the technical part of detecting alcohol vapor is cool,
but doing the systems integration to make it call the police
makes a large number of assumptions about the occupants of the car
and the legality of the actions they're about to perform
and the probability of false positives and false negatives
and the willingness of the police to be called about it.
(Police, for instance, don't like false alarms from burglar alarms.)
Validating those assumptions is part of the engineering job,
just like validating the effect of opening all the car windows
before you get in is.  Newspaper clippings usually don't do a good
enough job on details to let you estimate whether the engineering
was done well (except of course when things fail spectacularly.)

Building a device that can call any pre-programmed number
is a much different problem - it's almost identical technically,
but applications include selling to parents for their kids' cars
(and be sure to include a speakerphone in the communications part.)
        (Bobby!  The machine says you're drunk!  Are you ok?
        I'm fine, ma, I'm just driving Alice and Carol home.)
or if you're trying to sell it to people who are habitual drunks,
having it programmed to call a taxi makes more sense.

There may be some captive market for selling to people on probation,
who might accept it as an alternative to not being allowed to drive at all,
but that's clearly a niche market, not an install-on-all-new-cars market.



Reply via email to