Judge: Breath test firm must hand over source code An Arizona Judge has approved requests to allow the examination of the software in a new generation of machines that examine blood alcohol levels of motorists. Pima County Superior Court Judge Deborah Bernini ordered CMI, the manufacturer, to hand over the source code of the Intoxilyzer 8000 to representatives of defence …
the breath test is only the first line detection? I can't believe there are places that don't back them up with blood tests as it's the only accurate way to confirm blood alcohol level. A breath test only shows how much alcohol is on your breath, which is then used to determine blood alcohol. It is vulnerable to false positives!
In civilized countries, breath tests are only used as indication that a blood test might be a good idea, the latest being the only acceptable input for the judiciary bone-cruncher. Now most country don't rely on "walk the line" tests either...
These breath tests and, hem, "behavioural" tests are ok for small fines (like, for drunk loitering), NOT for drunk driver, who NEED to receive more than a slap on the wrist.
There has been no peer reviewed studies which has showed that breath tests actually perform a valid measurement of blood alcohol levels. So instead, often by statute, it is mandated that the alcohol level in the breath is multiplied by a factor of 2100 to produce a "blood alcohol level". That "2100" is called the "partition ratio" and can vary widely between individuals as well as over tim ein the same individual.
http://forensic-evidence.com/site/Biol_Evid/Breath_Tests.html
..."arguing the request is a defence ploy designed to get their clients off the hook"
And there's the crux of the matter. US prosecutors are not interested in justice, or it being proven that convictions are unsafe and have ruined innocent lives as a result. Just in getting as many sucessful prosecutions as possible, regardless of how ludicrous the claim.
...hanging chad redux. I agree that the source code should be available for inspection, that's what the vendor of the device should have done in the first place. The problem is that the devices are paid for by the very fines they produce (Radar speeding tickets sometime fall into this category). So, there is a vested interest to have as many "convictions" as possible, even when they are not founded in fact. For all we know, it prints out a sheet that says "drunk" for every 9 out of 10 tests it does. The assumption being that if the policeman has any suspicion he is probably right.
To the vendor: if it is THAT secret, you shouldn't sell it anyway. A chemical analyzer is a pretty standard device. Anyone should be able to make one, compete on cost, and show us the code!
I have worked for the local Law Inforcement as an IT person and I *know* for a fact that (at least in this region of the world):
A) A positive breath test gets you pulled in for a blood test.
B) The devices used by the cops (ie, breath testers, radars, etc) went through rigorous calibrations and are re-calibrated on a regular basis (a workmate spent 3 months on one of these calibration exercises).
While the breath-test will never give a 100% accurate BAC level, it *will* give you an indication of whether the driver should be pulled for further testing, and do *that* with reasonable accuracy.
The problem with using blood tests is that they are taken and delivered much later than breath tests - most cops don't have the training or equipment to draw blood at the scene. The alcohol concentration decays with time, and not necessarily the same for everyone or on a linear fashion. So people can argue that the test assumes too high a decay rate and that they were at only 0.06 three hours earlier, not 0.18 .
Texas now is trying to bypass the right to refuse and is just doing both the breath test and blood test. so basically what they are now saying is that rights of the people no longer matter but being this is the state that the "dubya" is from goes to show. they are also implementing street cameras as well though they are not lawful here they use them under the civil code and issue you a civil fine which you dont have to pay under penalty of law but they put a lein on your house or car.
The handheld device is used to indicate whether it's worth pulling the driver in for two tests on a proper big ass machine. A positive result on the handheld unit itself will never secure prosecution as it's evidence is not concrete {does not give exact levels of alcohol present, just has traffic light indication}
So if in the US they are trying to prove their point using JUST this machine then too right the cases should be {wrongly, in the sense of guilt, rightly, in the sense of doing things correctly} thrown out. Just a case of guilty people getting off because the pigs are too lazy to build a proper case.
PH because even she's not that stupid and I'm so pleased that newer lame PH icon has died :D
The gizmo has a sensor and other elements as well as the processing.
The proof of its effectiveness should be in measurements of breath alcohol against a "gold standard" - if it gives consistent and precise results then it's working.
My suspicion is that the defence are after trying to prove that errors in the source code mean that it MIGHT occaisionally misreport a result. And if it could, then can it be shown that it didn't misreport for the defendant who swears blind he'd only had a small lemonade and was weaving all over the road because a wasp got in through the window....
Or the defence are up to the usual mix of lawyerish tricks.
"We need to breathalyse you"
"Can't you take a blood sample"
"Well, if you insist"
*click*
*thud*
*extract*
"Well, he's sober but at least when he comes to and drives off we'll get him for driving without due care and attention due to concussion"
Also, surely it should be called Sauced code?
if i'm not mistaken, surely if the only evidence they have against these people is a breath test and not a blood test, all they need to do is prove that the breath test gives false positives and they get off, right?
ok, defense lawyer takes out an approved breath test machine in the court room, does a breath test which gives a result of 0 showing no alcohol at all. now defense lawyer breathes in puts a drop of vodka on their tongue and blows in to machine again. machine claims they are drunk - case dismissed.
not sure if they would be allowed to do this in the court room in front of the judge (not sure if there are rules about alcohol consumption in court), so it might need to be done elsewhere and filmed then the video submitted as evidence, but doing it in the court room would make the point in the most effective way.
> In civilized countries, breath tests are only used as indication that a blood test might be a good idea
here (England) conviction on the basis of breath test only is clearly mandated by statute and has been for a long time. But not, I believe, using the results of a hand held portable analyser - the hand-helds only give an indication to pull the motorist in for test with a rather larger and suposedly more reliable and accurate analyser at the police station. However, contaminated mouthpieces are something of a problem: policeman swabbing mouthpiece with alcoloh before handing it to motorist, ostensibly to make sure it's clean but you may imagine that there could be a different motivation.
If you don't pay the fine they put a warrant out for your arrest. Yeah I know it is a civil offense but they have manipulated the law to make the penalties criminal. And they don't even have to notify you by registered mail. Regular mail is ok.
And quit blaming Dubya. The Democrats have been pushing the same thing for years.
Neoc, Breathalysers are only accurate if the victim doesn't burp. And as far a calibration is concerned, I work with a lot of calibrated equipment in the Semiconductor industry. If the calibration procedure does not match the usage it is invalid. Even if it is traceable.
So why should the source code be made available? nobody ever asked for circuit diagrams or schematics of the analysers. Surely a calibration certificate that the device measures breath alcohol content sufficiently accurately should be enough. All seems rather spurious to me. Still i guess at least the lawyers'll do well out of it all, and maybe a few drunks'll get with dui or dwi or whatever.
Heh, while they're at it I wonder if they checked whether the cops had been properly trained to use the new devices? or whether the drunks had received adequate training in blowing into the machine...
People should ask for the circuit diagrams too. It's not good enough to calibrate these things on standardised samples and then argue that that 'proves' they are accurate. For all we know the software has arithmetic overflow or underflow when converting the a/d converter values into parts per thousand, it may have provable race conditions, dependencies on date/time or any one of dozens of typical software-only faults. Similar voltage-level dependencies could easily exist in the analogue circuitry - only a detailed design review by honest third parties can give good assurances to the contrary.
The fact that they won't hand the code over strongly suggest that it's never been quality audited or properly reviewed and may well indeed have dozens of stupid bugs in it.
I don't want to see drunken drivers get off, but neither should they be subject to randomness in their convictions.
they ask for the code for red-light cameras and speed cameras. Some company is selling a piece of hardware that's purely and completely designed to screw you, mr(s). average citizen drving down the road and if they can jigger the results with funny code you are going to be nailed to a cross and burned (or the legal equivalent).
If they have nothing to fear, they have nothing to hide (where have we heard that before)...
In any event, if you are liable for damages or punishment, you have a legal right to see the evidence against you and how it was gathered and that it wasn't "shaped" along the way.
p.s. if a later blood or urine test proves you are over the limit, you need to be fried anyway!
ight to refuse what here? Could Texans refuse to be breathalysed in the past or have I got this wrong?
BTW In the UK refusal to provide a specimen is treated the same as providing a failed specimen and gets you a 12 month ban.
Same thing herer in California, except your auto rates go up to
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
