Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?
Software Government The Courts Your Rights Online News

DUI Defendant Wins Source Code to Breathalyzer 638

MyrddinBach writes "CNet's Police Blotter column looks into a Minnesota drunk driving defendant case with a twist. The defendant says he needs the source code to the Intoxilyzer 5000EN to fight the charges in court. Apparently the company has agreed to turn over the code to the defense. 'A judge granted the defendant's request, but Michael Campion, Minnesota's commissioner in charge of public safety, opposed it. Minnesota quickly asked an appeals court to intervene, which it declined to do. Then the state appealed a second time. What became central to the dispute was whether the source code was owned by the state or CMI, the maker of the Intoxilyzer.'"
This discussion has been archived. No new comments can be posted.

DUI Defendant Wins Source Code to Breathalyzer

Comments Filter:
  • Additional info (Score:3, Informative)

    by y2imm ( 700704 ) on Thursday August 09, 2007 @05:54PM (#20175513)
    "Wetzel was arrested at his home on Feb. 25 after allegedly rear-ending a car with his pickup truck and then driving off. He faces five gross misdemeanor charges, including causing bodily harm and driving while intoxicated." FREDERICK MELO Pioneer Press

  • by Sloppy ( 14984 ) on Thursday August 09, 2007 @05:58PM (#20175561) Homepage Journal

    ..has nothing to do with this case, and little to do with who holds the copyright. What if he does find flaws, and others have already been convicted using output from the same machine? Suddenly, all those past cases come back up.

    I guess the lesson here is: the source should already have been public and heavily scrutinized. I don't want my government spending my tax money and wasting time in court, to get convictions based on evidence from mysterious unaudited machines. Why? Because sooner or later, some defendant is going to want the mystery peeled back. Some defendant is eventually going to want a fair trial. Might as well give that fair trial to the first one, so that a bunch of expensive shit doesn't have to get re-done (or so that a bunch of guilty people don't end up walking free, simply because the cops used a defective machine that ended up collecting untrustworthy "evidence").

    Keep mysteries out of court, from the start. Don't let a big list of convictions that depend on them, build up. The chances of the device being defective are probably pretty low, but you know there's gotta be some prosecutors with pits in their stomachs.

  • by scanlanj ( 1140555 ) on Thursday August 09, 2007 @06:02PM (#20175613)
    The previous machine, the Breathalyzer, went out of use when it was proven (by defense attorneys) that it was susceptible to RFI. The "new" machine, the Intoxilyzer5000 (and this was 25 years ago) was microprocessor based. It had an RFI detection circuit which was supposed to invalidate results if RFI was present. Other known issues are burping and chewing tobacco. Trouble is, the RFI detector was a comparator driving a login input. Without the software, you can't prove the box's performance from a white box perspective. That's trouble when you're relying on a machine vs. videotaped evidence of impairment.
  • Re:Sigh (Score:3, Informative)

    by futuresheep ( 531366 ) on Thursday August 09, 2007 @06:13PM (#20175757) Journal
    no matter how sober you think you are, those things are damn near never wrong.

    Read this before you think you can be sure about that. They can be wrong, especially the older ones.

    Link []

    Or the the machines may not have been tested properly:

    Link []

  • by neapolitan ( 1100101 ) on Thursday August 09, 2007 @06:16PM (#20175803)
    A low-carb diet (e.g. Atkins diet) can indeed make you "ketotic" and raise your breath acetone level.

    From your college chemistry course acetone has a C=O bond, while alcohol is a C-OH bond.

    Cheap breathalyzers will use a chemical reaction to detect the alcohol in your breath -- often potassium dichromate (these are the ones that go from red to green with alcohol).

    More advanced models (such as the ones the police would use, would use essentially spectroscopy to try to measure the resonant absorbance of the C-OH bond. This would not be fooled by acetone, which has a much different absorbance of the C=O (approximately 1700 cm-1 IIRC). There are also variants of this method.

    If you are ever innocent and accused, get a blood test, which really is a quantitative direct measurement and can be confirmed, with very little chance of being fooled.

    If you are not innocent, **IN THEORY** the easiest way to lower your reading is to silently hyperventilate prior to blowing. This would prevent equilibration of the alcohol in your bloodstream with the air in your lungs that you just breathed in and out. It is far from perfect, and I would strongly advise to never drive drunk, nor rely on this method.

    Additional references: vels_produce_inaccurate_Breathalyzer_results []
  • by Anonymous Coward on Thursday August 09, 2007 @06:19PM (#20175857)

    Back on-topic.. don't people who get caught with a breathalyzer (is what they're more known as over here) get taken to the station for a more thorough and accurate, possibly blood, test to determine the blood alcohol level, before going through the steps of fining?

    no, not at all. as a matter of fact, quite the opposite. in most states you will never be informed that you can in fact have a blood sample taken. you will blow on a machine like this and that will be all she wrote for you. had i know this, i would have asked for the blood test. there are many reasons why. first off, when you ask for a blood test (you usually have to provide (call) the doctor to take the blood sample, yourself, and pay for the doctor) you have about an hour, give or take, before the doctor gets there. this little bit of time might mean the difference between passing and failing. second, if you take the blood test, there is evidence that you can use in court in your defense -- whereas, if you only do a breath test, the only evidence that exists if it goes to court is a print out that the police have with a number on it.
    i blew on a draeger alcotest 7110 mkiii, and let me tell you, i got totally pwned. the guy that blew in front of me was way way over the limit and smelt of pure scotch. sure, they swapped the mouthpiece before i blew, but not the tube itself. well, this unit, inbetween tests, will draw air through the top of the machine to clear out the chambers for the next test. so, basically, when i blew right after him, i am convinced that there was still residual alcohol inside of the tube which in turn got me arrested and a dui on my record.
      whats funny is, at the time, the draeger alcotest 7110 mkiii was being contested in 4 states. i was going to fight it, based on those grounds, until i saw a case from a county in my same state. this guy blew the legal limit, and went to court to challenge the results. you see, with most breathalyzers the state /has/ to calibrate them every so often, and, if they cannot produce the documents proving the last time it was calibrated, you walk. that guy tried to get by on that paired with the residual alcohol stance. see, the draeger is actually quite advanced, it does the two most popular tests. infrared spectroscopy and electro chemical cell. but, inbetween tests, the machine does a self diagnostic. well, in this poor guys case, the state couldn't produce the documents showing the last time it was calibrated, but, the judge said, since the unit preforms a self test every time, that was good enough. sent him to jail for 4 years, first offense, driving at exactly the legal limit.
    let me finish with this. for all of you that don't know, the breathalyzers dont tell you your blood alcohol content, they tell you your BREATH alcohol content. this number is then multiplied by some outrageous number like 1200, and thats how they estimate your blood alcohol content. this is why many people that dont fit the 'average' in terms of body mass are often falsely convicted.
  • by nate nice ( 672391 ) on Thursday August 09, 2007 @06:23PM (#20175959) Journal
    We're already seeing a ton of people with absolutely no legal background commenting on legal things. Here's a tip: the law doesn't work like you probably think it does. The law is rational or reasonable. It's a jumbled mess of subjective orders and expressions that lawyers can mold into defense or complaint.

    Looking at the source code could very well be the basis for a very solid defense that beats whatever state statues and ordinances the defendant is suspected of violating. We have no idea what's in the code so why not? Crappy programmers probably wrote the software and it probably wouldn't be hard to find something that doesn't function right or doesn't map just right to a state issued requirement for the system.

    If this gun reading is the states main piece of leverage it's because this device conforms to some strict requirements defined by the state. So maybe looking at the code will show that it doesn't and that the machine is in fact illegal.

    Lastly, if you ever get pulled over for something like this, don't talk. That is when they ask if you've been drinking, always say no. What does "drinking" mean? Well, it's not up to you to define this at that time. Let your lawyer handle it. Never tell a cop you might be breaking the law. Because once you've admitted that you have been drinking, they can ask you a whole bunch of other questions that can only hurt you. How much? For how long? Where at? Where are you going? With who?

    Here's a sample:

    Officer: Have you been drinking?

    You: No.

    Officer: I smell alcohol.

    You: I haven't been drinking, officer.

    He'll still ask you to get out and do his little tests. But you've never admitted to anything. This can help a lot down the road. In short, never say anything you don't have to.
  • Re:What about (Score:1, Informative)

    by LunaticTippy ( 872397 ) on Thursday August 09, 2007 @06:38PM (#20176131)
    So let's say the cost is $20 and 1 hour for the drinker. Multiply that by a hundred million drinkers and multiply that by 20 times a year and what do you get? $40 billion dollars and 2 billion hours. Now how many lives does that save? It might be negative, might be zero, might be positive. Lets say it is positive. Lets say 1000 lives are saved every year, which I find highly doubtful. Remember that most traffic accidents are caused by sober drivers crashing into other sober drivers, and that drunk drivers are already breaking the law. You're willing to spend 40 million dollars and 2 million hours to save a single life. That's idiotic. If you want to spend money and time to save lives, you can save a lot more while spending a lot less in a myriad of ways.
  • by pyite ( 140350 ) on Thursday August 09, 2007 @06:39PM (#20176137)
    America has some problems, but I don' think we're to the point yet where they can arrest you on "suspicion of drunk driving" then use that to "forcibly take a blood sample".

    Yes, that is close to the reality.

    In the case of drunk driving, most states have adopted the law that if you are driving a vehicle, you have then given consent to submit to the approved test to find out if you're driving under the influence of alcohol. When you are stopped and you're not sure of what your alcohol level is, you cannot refuse to take a breathalyzer test. As soon as you got your drivers license, you gave consent in advance to do this. If you refuse, you will find yourself in bigger trouble than you would have by submitting to the test. This implied consent is automatic in the case of anyone who drives a vehicle. From: sent.html [].

    So, you've agreed to it in advance by having a driver's license. You get to pick your poison in terms of what kind of test it is.

  • Re:What about (Score:5, Informative)

    by Albanach ( 527650 ) on Thursday August 09, 2007 @06:48PM (#20176267) Homepage

    Having spent time in North Italy (Lots of mountain roads), I can say that I saw many people out to lunch split a bottle of wine between 2 or 3 people, and drive back to work. In all my time there, I didn't see one wreck. Not one.
    Then you were either lucky or weren't looking for them. Italy has one of the worst accident rates in Western Europe and drink driving is currently a hot political topic following a sixteen year old girl being hit and killed by a driver three times the limit. More here from the Herald Tribune []
  • by neapolitan ( 1100101 ) on Thursday August 09, 2007 @07:26PM (#20176697)
    Yep -- we have a varied audience on Slashdot though, so it is necessary to simplify things a bit. When reading esoterics on kernel hacking, I appreciate a simplified overview. You can't be an expert in every subject.

    I was a chemistry major; I have seen an OH peak, and know the range, and am intimately familiar with spectroscopy... The points you raise are well taken -- yes, the local environment would depend somewhat. However, when running a typical IR spec, how much do these things affect the absorbance? Would a typical IR spec vary with lab conditions / if your AC is on/off, or if the room was hot or cold? The human body would actually provide quite a controlled environment (36 to 38 degrees, high humidity) for measurement. Thus ambient temperature would be relatively constant, provided the person can breathe at a consistent rate. I don't think this would affect readings much, and would be very easy to empirically determine.

    Acetone itself would not affect it much on the levels that a ketotic person could produce. Again, I said *much*, not theoretically or at all. I think your estimates (one -> six beers) are way off.

    The proof is in the data -- none of us can say that the presence of acetone would do this without trying it, and I acknowledge the improbable possibility.

    Do you want to see for yourself -- look at the relative contribution to the spectrum: frared/oxygen.html []

    That is why if you are truly innocent, demand a (unfortunately invasive) blood test and put the issue to rest.
  • by Omeger ( 939765 ) on Thursday August 09, 2007 @07:43PM (#20176855) Journal
    Actually it's a crime to not take a drunk driving test in most states if asked, which would hurt your record just as bad since it implies guilt.
  • by HomelessInLaJolla ( 1026842 ) * <> on Thursday August 09, 2007 @07:50PM (#20176913) Homepage Journal

    However, when running a typical IR spec, how much do these things affect the absorbance?
    The same sample run in the same infrared spectrophotometer a mere five minutes apart can have an ~3000 cm^(-1) absorbance peak area which varies at least 20%. I'm not saying this is normal. Normal variance for this peak, for the exact same sample at five minute intervals is, in my experience, around 10%. The same sample analyzed a day apart will be, honestly, unpredictable. Calibration based on this peak is scientifically impossible if one truly desires at least two significant digit precision.

    The human body would actually provide quite a controlled environment (36 to 38 degrees, high humidity) for measurement.
    Since the humidity of the calibration sample cannot possibly be tuned to match the humidity from the suspect human's lungs this is an enormous factor. Water vapor plays a huge part in the size and shape of the ~3000 cm^(-1) peak.

    Keep in mind, though, that the Intoxilyzer in question works on two lines. ~3000 and ~1650. There is one calibration line for each of these, but the calibration curve is a single point curve: origin and set calibration point. That's hardly acceptable for readings with two significant digits. On top of that many intoxilyzers add an electrochemical detector which works on any primary or secondary alcohol and possibly some aldehydes, depending upon ambient temperature and humidity. So what you have are two very poorly tuned calibration curves and an electrochemical cell reading--those _THREE_ readings are then fed into an internal algorithm to produce the final reading.

    Imagine buying two low quality current meters, one from your hardware store, one from Radio Shack (choose their cheapest model), and then making your own using parts (resistors and capacitors) that you bought from a mail order catalog. Now use all three of those to measure the current on a particular trace on your motherboard. Now average all three of those readings together so that it fits the mobo manufacturer's spec of "+3.2V".

    Now add in the fact that the line at 3000 is affected by water, primary alcohols, secondary alcohols, tertiary alcohols, and 1',2',and 3' amines. There's _NO_ specificity. There's no way of showing that any given signal was solely ethanol, or if it was 1:1 ethanol/ethyl amine, or if it was 1:1:10:20:40 ethanol:ethyl amine:cologne:your wife's perfume:rubber cement from the trooper's thumb.

    Now add in the fact that the line at ~1650 is affected by primary alcohols, _all_ aldehydes, and _all_ ketones. The same "how much of the signal is really JUST from ethyl alcohol" applies.

    Now add in the fact that the EC current reading is from any reducible organic which flows over it. Theoretically the EC is somewhat alcohol specific but double bonds and amines will probably contribute to the current reading.

    Now factor in that BAC is only empirically related to breath alcohol level and the ratios change given any individual's particular body chemistry.

    Now factor in that the source code for the instrument has to take all of these measurements of dubious quality and spit out one single number.

    Garbage in, garbage out.

    The proof is in the data -- none of us can say that the presence of acetone would do this without trying it, and I acknowledge the improbable possibility.
    A former co-worker was a state trooper for a while. He exposed to me just how much the officers know how to manipulate the tests. His quote was,"You'd be surprised what you can find in the lunchroom refrigerator whose vapors will be over the legal limit."
  • Re:What about (Score:4, Informative)

    by terrymr ( 316118 ) <> on Thursday August 09, 2007 @07:55PM (#20176961)
    Alcohol related is a world away from alcohol caused - most of those stats refer to accidents where somebody present had had a drink in the last 24 hours.
  • by spikedvodka ( 188722 ) on Thursday August 09, 2007 @09:17PM (#20177593)
    Check your driver's license... and the local laws. for example:
    Florida's Driver's Licenses states right on the front "Operation of a motor vehicle constitutes consent to any sobriety test required by law"

    Maine law is that "Under Implied Consent, you automatically agree to a chemical test (blood, breath, or urine) at any time authorities have probable cause to administer it. If you refuse to take such a test for alcohol or drugs, your driver's license will be immediately suspended. The suspension could be for a period of up to six years. Because it is an administrative suspension, no court action is necessary. In addition, testimony from the arresting officer regarding your driving performance can result in an OUI conviction even without the BAC test! f you are found guilty of OUI based on the police officer's testimony, your refusal to take a test will be considered as an aggravating factor by the judge and another suspension, as well as mandatory jail time, will be tacked on. So by refusing, you will have a much harsher penalty than if you'd taken the test.

    Remember a test can protect you. If you are not legally intoxicated, the test will show it. " []
    actual text of the law: 29-Asec2411.html [] but if you "Operate a motor vehicle while under the influence of intoxicants, AND failed to submit to a test at the request of a law enforcement officer" you are guilty of Criminal OUI.. then "A law enforcement officer may arrest, without a warrant, a person the officer has probable cause to believe has operated a motor vehicle while under the influence of intoxicants if the arrest occurs within a period following the offense reasonably likely to result in the obtaining of probative evidence of blood-alcohol level or drug concentration"

    and then of course the meat-and-potatoes... if you don't take the test, the fine is higher... I'd copy/paste, but you can click the link just as well as I can
  • by Idarubicin ( 579475 ) on Thursday August 09, 2007 @10:09PM (#20177989) Journal

    It's the truth. 0.08 is below any significant level of impairment under normal driving conditions....

    First of all, you need to cite some sort of source for a statement like that. (A review by Fell and Voas [] reports that reducing the legal limit from 0.10 to 0.08 reduced alcohol-related crashes, injuries, and fatalities by between 5% and 16% in the United States; they report further statistically significant reductions in fatalities in jurisdictions that have moved to a limit of 0.05.)

    Second - as other posters have noted - how prepared are you to deal with a surprise abnormal condition?

    Third, nice weasel word--below any 'significant' level of impairment? What does that mean?

    Fourth, I should hope that the limit would be below the level of significant impairment under any condition. There's no compelling reason why anyone should have to drive with any alcohol in their blood; any limit ought sensibly to include a margin of safety.

  • by Grym ( 725290 ) * on Thursday August 09, 2007 @10:22PM (#20178059)

    let me finish with this. for all of you that don't know, the breathalyzers dont tell you your blood alcohol content, they tell you your BREATH alcohol content. this number is then multiplied by some outrageous number like 1200, and thats how they estimate your blood alcohol content. this is why many people that dont fit the 'average' in terms of body mass are often falsely convicted.

    Actually, it's more complicated than that. Most of the breathalyzer machines actually measure the absorption of light sent through a chemical that changes color when oxidized by alcohol. From that absorption number they are able to estimate the the concentration of alcohol in the sample in the machine and, from that, the concentration in your breath. The concentration of alcohol in your breath correlates with the concentration of alcohol in your blood because the alveolar membrane in your lungs is extremely thin and highly vascularized (it must be this way to permit gas exchange for respiration). The alveolar membrane is of a similar thickness in people of all body types and would probably not be a source of error. The only possible exception to this might be someone affected with fibrosis, which would actually lower the measured amount.

    That being said, this process that the breathalyzers go through requires a lot of assumptions which often are not true in all cases. In fact, I've heard of Chemistry/Biology grad students making quite a bit of money by offering their expertise to challenge (often successfully) these assumptions in court.


  • by Eivind ( 15695 ) <> on Friday August 10, 2007 @08:45AM (#20181349) Homepage
    Actually, punishments are for 3 reasons. But you're rigth, none of them are for "revenge" or "anger".

    First, as you say, locking someone away obviously protects the rest of society from the consequences of their actions for the duration they're locked down. If this was the only concern though, all punishments would be jail forever, which ain't the case.

    Secondly, we hope that being punished for a certain action, makes you less likely to do it again. For many notorious criminals this is quite doubtful, for other people it may work better; if you get a ticket for parking in the wrong spot you may well be more careful with parking for a while thereafter.

    Third though, and probably more important than the second, we hope that the knowledge that you *MAY* be punished if you do certain things will keep you from doing them. This certainly works to some degree, peope *do* infact commit more crimes when they think, for whatever reason, that the chances of getting punished for it are slim to none.

Where it is a duty to worship the sun it is pretty sure to be a crime to examine the laws of heat. -- Christopher Morley