Researchers say a breathalyzer has flaws, casting doubt on countless convictions

https://www.zdnet.com/article/draeger-breathalyzer-breath-test-convictions/

by uptown • 8 years ago

133 146 8 years ago

nerdponx• 8 years ago

Remind me again why a device paid for with public funds and used to carry out essential public functions of the government should be anything but open source.

sigstoat2• 8 years ago

the government buys some of very nearly everything and uses them to carry out essential public functions. you've got to have a better criteria than that.

logfromblammo• 8 years ago

The defense in a criminal case has the right to examine witnesses that presented evidence against them. When the witness is a machine, such an examination must necessarily include inspection of the operating protocols, technical specifications, calibration records, and source code.

Forcing the defense to ask for it in discovery for every single case seems somewhat counterproductive. If your code is being used to put people in prison, you don't get to keep your trade secrets privilege.

jpdb• 8 years ago

Because the software/IP isn't owned by the government.

I'd personally rather the US government pick the best solution, not the most open source one.

I would hope the best solution is open source, but the gov't shouldn't be forced to sacrifice quality.

However, I can agree that any Software/IP developed via public funds should definitely be open source.

DennisP5• 8 years ago

Define "best." My definition would be something that's not only accurate in most conditions, but lets the accused mount a meaningful defense.

Comprehensive testing by an independent lab gets us partway, if all the results are available to the defense, but it's better if we can both do black box testing and examine the code.

thatcat• 8 years ago

How should the public verify the government has chosen the best solution if the design of the solution isn't readily available? Depending on your perspective and definition of best, faulty designs that give more favorable results could deliberately be chosen to increase revenue / crime stats.

jononor• 8 years ago

Trustworthyness is (or maybe should) be an criteria for things being used to provide data for legal matters. Being able to independently verify the device is critical for this, and open-source makes this easier.

craftyguy2• 8 years ago

I'd personally rather the US government pick the best solution, not the most open source one.

I'd personally rather the US government consider trustworthiness, which is only afforded in open source projects, to be an important criteria for "best solution."

dbrgn• 8 years ago

In Switzerland, breathalizers may be used to justify pulling people out during traffic controls, but only a blood test is valid proof of too much blood alcohol before court. Is that different in the US? After all, even ripe fruit can result in a breath test indicating too much alcohol.

skc• 8 years ago

There are now breathalyzers that are near enough as accurate as blood tests that are admissible in court these days.

But they are huge machines kept at the police station after you've been hauled in after failing the normal road side breathalyzer test.

alkonaut2• 8 years ago

And that’s the type of machine being questioned in the article, not the roadside ones?

Should be clearer - and if those bulky machines are replaced with blood tests (assuming they are much more reliable) I don’t see the problem. Having doubt in the test used to bring someone to a second test is fine. The second test is used for conviction so there can’t be any doubt.

MertsA• 8 years ago

There are now breathalyzers that are near enough as accurate as blood tests

But in the general case this is quite simply not possible. It might be very accurate in the controlled case of a subject who hasn't consumed anything in the past hour but these breathalyzers are not being used in a nice controlled environment. The concentration of alcohol in a breath can vary wildly compared to the concentration in the blood. If someone used e.g. mouthwash a few minutes before being tested it will indicate a high BAC even if it's actually zero.

Unless the police are going to start forcing people to brush their teeth, clean their tongue, and use some form of non alcoholic mouthwash before testing you can't come close to the sort of accuracy that an actual blood test would give. In real world conditions there are too many situations where the concentration of alcohol in your breath is substantially different than the concentration of alcohol in your blood.

APCarr• 8 years ago

I use a cheap breath tester to see if I'm in ketosis when losing weight (burning fat creates ketones, which trigger breath tests).

Jus' sayin'.

raarts• 8 years ago

What's the precision of that device?

VectorLock• 8 years ago

How does that compare to using test strips?

2RTZZSro• 8 years ago

You should be skeptical of the output of a ketone measurement device. You need to measure blood concentration of beta-hydroxybutyrate to accurately determine your ketosis situation.

rascul6• 8 years ago

The biggest flaw is that it doesn't actually measure drunkenness or impairment.

pc862• 8 years ago

If you're going to be arresting people for DUI, you need an objective measure of impairment. Whether or not someone can walk in a straight line or say the middle third of the alphabet should be plenty for probably cause but not enough to arrest someone.

maxerickson2• 8 years ago

BAC isn't an objective measure of impairment, it's a measure of BAC.

But short of an actual objective test, arresting people for driving after consuming alcohol is plenty reasonable to me.

alkonaut2• 8 years ago

I’d rather have an objective measurement of blood alcohol than a subjective measurement of drunkenness. People can’t judge their own capability or drunkenness so even if you ace a reaction/driving/balance/whatever-test with a high blood alcohol level, there is nothing to say that you could have judged that ability.

That’s why the law isn’t against driving drunk or impaired it’s a law against driving with alcohol in your blood.

notheguyouthink11• 8 years ago

Also, is "drunkenness" not something measured? I thought the effects on reaction time due to BAC were well understood. Are they not?

Decision making is more difficult to judge of course, but reaction time is massively important. So if we have a quantifiable measurement, who cares about vague things like drunkenness?

Though, if something like decision making goes south far before reaction time does, then I could totally see an argument being made against using BAC as a measurement. Yet, I don't believe that's the case.

dragonwriter9• 8 years ago

That’s why the law isn’t against driving drunk or impaired it’s a law against driving with alcohol in your blood.

It's often a law against doing either one, but if you are above they legal BAC limit that's much easier to prove beyond a reasonable doubt than is impairment due to alcohol.

lttlrck• 8 years ago

You'd still need to be stopped, which would still take an officer observing erratic behaviour, and he'd still make you walk in a straight line and ask you questions to check if you are slurring your speech. And then you'd still need to be tested to see if you are over the legal limit.

So maybe you mean that the law shouldn't be based on BAC?

merpnderp• 8 years ago

Every time I've been stopped, the "erratic behavior" was it being late at night and me being out on the road. I was stone cold sober on every stop, and at night I always drive particularly careful as the cops are much more likely to pay attention to you in particular.

At one "rolling a stop sign" stop, the cop basically admitted I hadn't rolled the sign, they just wanted to make sure I hadn't been drinking - "Yeah, I just wanted to make sure you hadn't been drinking, have a good night."

rascul6• 8 years ago

With stronger reckless driving laws then we can get rid of DUI and cell phone laws. If you're driving recklessly and putting other people in danger, it shouldn't matter how you got there. Drinking, prescription drugs, lack of sleep, cell phone usage, eating, reading, and watching the coast as you drive by are all examples of things that can have equally bad results. Yet we focus almost completely on convicting people based on some arbitrary measurement and sobriety tests designed to make you fail. We should instead convict people based on endangering others.

cascom2• 8 years ago

Sorry - in my town the police pull over drivers late at night trying to get DUI arrests - the “reason” for the stop is almost always a minor lane violation - then a mandatory breathlizer if you pass no ticket is issued, and you’re on you way - if not you’re going to jail.

Point is the pretext for the stop is nonsense...

thaumaturgy8• 8 years ago

It doesn't look like this is an error that will cause a totally sober person to fail a breathalyzer test. If you're worried about the accuracy of the machine, you can also submit to a blood test: https://duilawyerlosangeles.com/dui-lawyer-advice-breathalyz...

This is one issue where I'm firmly on the side of the police until there's evidence that faulty breathalyzers are putting totally innocent people in jail. If you choose to get into the driver's seat of a car, you should not be anywhere near the legal BAC limit. By the time an officer asks you to submit to a BAC test, they've already observed enough behavior to make them believe that you may be a danger to others.

Plan ahead so that you don't have to worry about leaving your car somewhere. Call an Uber or a Lyft. Get a friend to drive you home. Moderate your drinking. Take some time, go for a long walk. On popular holidays in many places, you can even call AAA for a free ride home, even if you're not a member.

Just please don't drive home on the hope that a defense lawyer can argue on your behalf that breathalyzers aren't reliable.

Clubber3• 8 years ago

When someone blows a 0.08 when it really is a 0.07, that would be putting an innocent person in jail.

simion3142• 8 years ago

Don't you have the right to ask for blood tests where you live?

lifeisstillgood3• 8 years ago

If you got into the drivers seat at 0.07 you are doing something wrong.

Unless breathalysers detect alcohol when none exists, then I think knowing they are "10% eitherway" is a good caution.

thaumaturgy8• 8 years ago

In case it wasn't clear from my comment, I'm less willing to argue for the innocence of a person who was at .07, or .075, or .0799, because of the potentially serious consequences of their decisions and the number of convenient alternatives they had available.

That's my personal position on this. It's consistent with my own moral framework. I get that there's a valid counterargument from a position of law-and-order morality.

But I've also had friends who were first responders to a horrifying drunk driver collision. Trying to keep innocent people from bleeding out on the side of the road messes you up for a while. It's not a victimless crime.

You can get from .08 to .06 by walking around for an hour. If you think you might be close enough to borderline, go take a walk.

DanBC• 8 years ago

If someone stops drinking, then gets in a car, then an hour later blows a 0.07 they were probably over when they started driving.

DennisP5• 8 years ago

That's excellent advice for personal behavior. However, the justice system should enforce the law as it stands, as accurately as reasonably possible. If the legal limit is too high, we should fix that in the legislature, not use faulty equipment to punish people who didn't break the law; someone who's a bit under the legal limit is in fact "totally innocent" according to the law.

I once was given a BAC after being stopped for a faulty tail light on a Saturday night. I'd had a single glass of wine two hours before, and at that point was completely sober, wide awake, and paying attention to my driving. The officer didn't mention any other reason for stopping me, and seemed disappointed when my BAC test was clear.

grepthisab• 8 years ago

I had a similar experience. Went out, had a single drink over about a three hour period, drove home, got pulled over for speeding. I think given the time of evening they just assumed I had been drinking. They gave me a BAC and officer also seemed disappointed it was clear. If I had had more than two or three drinks over the same period of time (which would have made me still well under legal limit), I may have requested a blood test out of an abundance of caution and to not rely on a possibly-inaccurate BAC to determine the (I’m told) very expensive consequences of getting a DUI.

matwood• 8 years ago

The officer didn't mention any other reason for stopping me, and seemed disappointed when my BAC test was clear.

This is way more common than it should be. Dash cams have finally helped this situation somewhat, but when I used to work late in college I got pulled over way more than I should have for BS reasons. "You touched the white line", "Nope". "You were swerving", "No again".

Finally a judge started throwing out all DUI cases without a dash cam and the police who dragged their feet for years after being mandated to install dash cams suddenly got it done.

No one should ever be driving after drinking, but the police also need to follow the law.

dzhiurgis• 8 years ago

“Nice” thing about blood draw is that it’s going to take some time to get to a clinic, letting you loose 0.1 promile due to metabolism (or gain if you actually had a drink just before leaving). Sometimes you can even probably find ways to drag time...

Don’t drink and drive tho.

jfaat• 8 years ago

In court they will just extrapolate from the blood results what your BAC would have been when pulled over, so this isn’t a silver bullet. Not drinking before driving definitely is though!

sandworm101• 8 years ago

Take this from a lawyer: Even if you are drunk, talk to a lawyer before the blood test. There are scenarios where, without drinking more, your blood alcohol can increase between the time you are pulled out of the car and the time your blood is drawn. Mix alcohol and food and there can be significant delays between drinking, breath alcohol measurements, blood alcohol measurement, and actual intoxication. Remember that the crime involves having a blood/breath alcohol level WHILE DRIVING. "Drinking and driving" alone doesn't make you guilty.

I get very worried whenever "evidence" is taken far away from the offense. I've even heard about people who, after being arrested/released based on breath alcohol, are allowed to continue drinking and then have their blood drawn.

If you are stopped and are given a breathalyzer, pass or fail, do not drink anything for at least 24 hours. You never know what the cops are thinking. They may be back. They may test you again in 15 minutes. Don't go back to the party. If the cops test you, your night is over.

jrjarrett• 8 years ago

Remember that the crime involves having a blood/breath alcohol level WHILE DRIVING.

Yep. This. I was on jury duty last year for a case involving this exact scenario. The defense lawyer was able to successfully argue the case "to a reasonable doubt" based on this basis.

jonknee• 8 years ago

By the time an officer asks you to submit to a BAC test, they've already observed enough behavior to make them believe that you may be a danger to others.

That's the goal, but cops also sometimes just like to exercise their power and arrest people who don't show signs of impairment. Or use that as an excuse to search the car and try to find a speck of weed. I certainly don't trust an officer to accurately gauge something so nebulous. Hopefully in a few decades we'll all be in self driving vehicles and DUI will be a thing of the past.

vivekd• 8 years ago

As a lawyer, I support your personal advice that no one should be anywhere near the legal limit while behind the wheel. I advocate for 0 drinks before driving.

That said I want to correct a few things:

1. In many jurisdictions, including mine, you do not have the option to submit to a blood test. If the police demand a breath test, and you refuse, you will be charged for failure to blow, the penalty for which is, in many cases, the same as being charged with drunk driving. You can't say to the police "can't I just get a blood test instead" to get out of this failure to blow charge.

2. The intoxylizer breath testing devices are heavily guarded, the government and the company that makes them have refused to release the source code or the machines themselves for testing. The one time the court did order the source code to be turned over the company claimed they lost it.

https://tampaduiattorney.wordpress.com/2011/07/26/cmi-lost-t...

Under such circumstances, it is unrealistic to think that real objective and meaningful tests will be done to establish the validity of these machines.

3. I'm not the one telling people that there is a legal limit and they should be below it. I'm telling people not to drink at all. It's the government that made up this stupidity about legal limits and it seems to be the least we can ask for as citizens for the government to stick to the very stupid and ill devised rules they come up with.

If you don't want people to be anywhere near the legal limit when driving then make that the rule and tell them that. Don't make up this mumbo jumbo about legal limits and pretend you have the ability to test for it when you don't because that'll lead many people to inadvertently break the law.

supergeek1333• 8 years ago

Exactly, there is quite a bit of news stories every year about "here is how many drinks a person of this height and weight can have before being .08" or whatever.

Just don't drink and drive, that's the only sure fire way to not get accused. Especially for a limit you can't test yourself without spending ~$100USD to get a personal tester which they don't guarantee the accuracy of.

We're trying to enforce laws on an uneven playing ground (a limit the average citizen can't test against until it's too late).

Balgair• 8 years ago

FYI for other readers, in CO and MI, the limit is 0.05. Be careful out there.

If the police demand a breath test, and you refuse, you will be charged for failure to blow, the penalty for which is, in many cases, the same as being charged with drunk driving.

Question for a lawyer then: How does this affect international travel. I know that you can't go to Canada if you have a DUI on your record, and other countries can be even more scrutinious. Is a Failure to Blow considered the same as a DUI to countries like those?

dragonwriter9• 8 years ago

It's the government that made up this stupidity about legal limits

Breaking the legal limit is usually sufficient but not required for DUI; DUI laws existed before legal limits and addressed being under the influence. Legal limits were added on top of the base rule to create a bright line where there would be clarity of guilt, but AFAIK in most places they did not replace the base rule, just added on top of it.

camgunz5• 8 years ago

The BAC limit in most states is basically two drinks. That's moderate drinking, but if you get pulled over that's a DUI. If you have one drink but your breath is too hot, that's a DUI. If you have two drinks, walk for 6 hours and get behind the wheel, that's also a DUI because the time for your BAC to go down is more or less "all night". The legal limit is very, very low, to the point where reasonable behavior is illegal.

Basically the rule is "if you've had any alcohol in the last 24 hours, don't drive". Most people think you can have one or two and drive home, but you can't.

djrogers3• 8 years ago

If you have two drinks, walk for 6 hours and get behind the wheel, that's also a DUI because the time for your BAC to go down is more or less "all night".

Completely untrue - I've been a 'subject' in police department intoxication training, and have watched the BAC of myself and my fellow 'subjects' drop rather predicatbly over the course of an hour or two from 'going to jail' to 'probably shouldn't have driven but not over the limit'.

The legal limit is very, very low, to the point where reasonable behavior is illegal.

Also untrue for most people - yes, if you are a high-function alcoholic you may be more able to perform basic tasks while at .08 or .1, but most people are significantly reduced in capcity when actually tested (not just 'I feel fine').

dragonwriter9• 8 years ago

The BAC limit in most states is basically two drinks.

Very roughly true.

If you have two drinks, walk for 6 hours and get behind the wheel, that's also a DUI because the time for your BAC to go down is more or less "all night".

No, it's in the neighborhood of 1 hour to drop 0.01% from most sources I can find.

EDIT:

The legal limit is very, very low, to the point where reasonable behavior is illegal.

No, slamming two drinks and hopping immediately behind the wheel is not even approximately reasonable behavior.

toss1• 8 years ago

I agree that I'm all against anyone even getting near the wheel after drinking, and the police should have every good tool at their disposal to prevent it.

"By the time an officer asks you to submit to a BAC test, they've already observed enough behavior to make them believe that you may be a danger to others."

Not true at all.

I know of numerous instances where people were required to blow into the breathalyzer with no prior observations related to substance impairment, or even a field sobriety test.

I personally had an instance where I was stopped for 'suspicion of speeding' where the officer got no high reading but thought I was going fast (possible; I'd seen an odd reflection off to the side of the road and braked a bit), I had been driving for hours and had consumed zero drinks for days, but he told me to sit in his cruiser and do the test. Seemed disappointed.

So a bad testing device could be a serious issue.

stochastic_monk• 8 years ago

This is a feature, not a bug.

In fact, many jurisdictions have requirements that, above a breathalyzer-reported BAC threshold, different minimum mandatory punishments are levied. Sure, a sober person won't get a result that's above the legal limit most of the time, but someone who's done some drinking and is within appropriate parameters is often labeled as DUI, and could potentially register as above this higher, mandatory-jailtime, mandatory-felony situation.

I've known of cases where the breathalyzer in question has been shown to be broken and decommissioned, yet the people whose numbers were so wildly inflated that they couldn't possibly be accurate have still been convicted using these numbers.

In one case in particular, the judge ruled that the evidence on the unreliability of the machines in general (and this machine in particular) was not admissible for court.

It's about punishment, not deterrence.

DennisP5• 8 years ago

The police department says they "approved the instrument that best fit our business needs." It's a bit disturbing that they think of themselves as a business, especially since arguably their business need is to get as many convictions as possible.

instaheat• 8 years ago

I got a DWI in 2009, pre Uber/Lyft. I don't excuse my actions and I got what I deserved.

That being said, I would never consent to a breathalyzer or field sobriety test today sober or not. I see no issue with flexing our rights and really get tired of the "just comply" narrative.

frandroid• 8 years ago

As opposed to pre taxis?

dsfyu404ed3• 8 years ago

This is news to exactly nobody who ever read the manufacturer specs of the sensors used in devices that detect alcohol for law enforcement.

They're running industrial process control sensors at the limits of their range and in widely changing environment. It should come as no surprise that they're hard to calibrate accurately and likewise deliver variable results in a lot of cases

In a factory environment equipment doesn't get calibrated in a 70deg car and used in 20deg winter air to measure 90deg human breath. In law enforcement that's exactly what happens.

Ankle bracelets are even worse.

The fact of the matter is that most hardware/software solutions sold to law enforcement are only a small step above the scam bomb-detectors that got sold to the Iraq government (Google it if you're unfamiliar with the story).

If NASA were writing the calibration code, the results weren't being interpreted by laymen and their interpretation wasn't being used to destroy lives I'd feel differently.

arbuge• 8 years ago

This is not really news. For example in Texas, DWI lawyers run a lucrative racket combating breathalyzer test results using the known flaws in the machine and its application. Blood tests are considered more definitive and harder to defend against.

pavel_lishin• 8 years ago

Why is it a racket?

sli• 8 years ago

There is at least one model where, if the officer covers the outlet tube, the reading doubles extremely reliably. I saw this tested in a short documentary some years back.

If you're going to get breathalyzed, will you notice if the officer is doing something shady? Could you even know, without knowledge of the various machines?

I'm also quite miffed that a cop can give you a DUI even if your reported (by the machine) BAC is under the legal limit, 0.00 included.

kkielhofner• 8 years ago

It should be know that in many jurisdictions (at least within the US) the officer's belief that you're impaired is usually enough for arrest - regardless of the outcome of any field sobriety tests or machine results.

On one hand I understand this approach due to the dizzying array of factors that cause impaired driving - from alcohol to prescription drugs to designer drugs that may not have definitive tests available. Seeing as sleep deprived drivers often perform as poorly as impaired drivers on various tests that should be included as well.

On the other, this is still a situation where an officer's opinion with no supporting evidence (or conflicting evidence) can lead to arrest.

Sure, you may "win" in front of a judge eventually but this is after you get handcuffed, spend the night in jail, have an arrest on your record, etc. Not to mention having your mugshot posted online by various unscrupulous sites running an extortion racket (demanding payment to remove). You'll soon discover that to many people simply being arrested = guilty.

In my case my "win" came 14 months later when the judge watched the dashcam footage, saw that my driving, speech, behavior, etc clearly wasn't impaired and threw the case out. Total time in court was less than 20 minutes.

Interestingly we watched the dash cam footage after the officer testified. At one point my attorney asked why the officer thought I was impaired. He responded with "I'm Irish Catholic and I know what drunk people look like". The entire courtroom (myself included) laughed.

craftyguy2• 8 years ago

He responded with "I'm Irish Catholic and I know what drunk people look like".

Perhaps the officer was impaired.

lisper2• 8 years ago

I drink and drive all the time, which is to say, I will drive to a restaurant, enjoy dinner and drinks, wait, and then drive home.

In order to be sure that I've waited long enough to be safe, I bought a BACtrack S35 breathalyzer to keep in my car. In many, many years of testing myself, I have never seen a reading above 0.03, at which point I feel pretty impaired. If I'm over 0.02, I don't drive. I don't think I could stay conscious long enough to get myself up to the 0.08 legal limit.

All of which makes me wonder: can I trust this gadget to be properly calibrated? And how would I find out? The only way I can think of is to approach a police officer while drunk and ask them to breahalyze me so I can compare readings. Somehow, the opportunity to do that has never arisen.

It seems to me like a no-brainer that any establishment that serves alcohol ought to be required by law to have a properly calibrated breathalyzer on the premises for its customers to use. But of course this won't happen because the truth is we don't really want to know. We have drunk-driving theatre just like we have security theatre, and both seem to be a stable equilibrium.

Alas.

supergeek1333• 8 years ago

BACTrac actually recommends you send it to them on a regular basis for calibration

lisper2• 8 years ago

Sure. But the results I get today are consistent with the ones I got when the unit was brand new. In fact, it's so consistent I can predict with very high accuracy what my reading will be just based on how much I've had to drink and how I feel. So if the unit is inaccurate now, it was inaccurate when it was new.

Also, we have two units, one in my car, one in my wife's car. I've tested myself back-to-back with both units (at home) several times and the readings are always consistent.

kevin_thibedeau• 8 years ago

Why would it even be possible to disable sensors critical to getting a credible estimate of blood alcohol? It all sounds like a convenient way to boost revenue with false positives while hiding behind the techno-wizardry.

DrScump• 8 years ago

Years ago, on "L.A. Law", a story arc had one of the lawyers stopped for weaving. His partner advised him, "breath or unine -- not blood!". This was in CA, which gives the driver his/her choice of the three methods.

Are there cases where blood disfavors the suspect?

Researchers say a breathalyzer has flaws, casting doubt on countless convictions

Related Stories