Hacker News Reader

Comment by hirundo | Hacker News Reader

Why use less reliable proxies like breathalizers to gauge impairment when we now have the technology to measure it far more directly: convert traffic cop cars into mobile virtual reality driving simulators.

Put a gaming laptop in the trunk, and the DWI suspect behind the wheel with a VR headset on and let them drive through a couple of minutes of reasonably challenging simulated traffic. Let software come up with an objective score and store the simulated session as evidence if they fail.

Not only does this more directly measure impairment, it catches any cause of it, not just a particular substance. Virtual driving impairment will correlate with real driving impairment far better than parts per million of a molecule in the bloodstream.

View on HN

You're describing a fancy field sobriety test. There's a reason why those who fail field sobriety tests for alcohol are then asked for a breath sample (or when they refuse, a blood test)

A criminal defense lawyer would tell their clients to say in court they intentionally failed the test. (Burden of proof still lies with the prosecution) You can't be convicted for failing a test. You can only be convicted of being under the influence.

I read something (a legit study IIRC) that breathalyzers are not that accurate and if you don't hold them right can increase the BAC.

And NO you CAN be convicted for failing the test. The evidence is just a paper reciept from the breathalyzer.

The reason you choose the blood test is so you have evidence for later.

Also you will sober up a little while you wait.

To be clear, I mean no one is being convicted of failing a field sobriety test with no breathalyzer or blood BAC results.

I believe there was a case in Colorado where a medical Marijuana user had enough THC in their blood to be considered impaired but they were able to pass a field sobriety test so they weren't convicted of a DUI. The problem is that the amount of THC you need in your blood to impair you is different for every person so it's hard to develop a test for it. A driving simulator makes sense to me, though once you have a bunch of adrenaline kicking in and are totally focused on not going to jail, you may perform better than when you are relaxed in the comfort of your own car.

It's an interesting idea, but the critical issue with it is that it's not illegal to be bad at driving. That could change, but that's how it is now.

Ah, but it's illegal to drive without a license and the test is a de facto driving test so you could make it so that if you fail it you won't get arrested but you get your license suspended on the spot.

Currently simulations aren't valid driving tests, so that entire section of law would have to change. Moreover, does any state do on-demand driving tests? Typically you set the appointment, not the state. Additionally Texas (I'm sure other states are similar) doesn't even "suspend on the spot" for DWIs (you have 40 days or more from the incident before the suspension is effective).

It'll illegal to drive dangerously though, surely. It's a test that removes the license to drive from those who can't drive properly, whether due to old age, eyesight, drugs, tiredness. Only self driving cars for you.

The tests for obtaining a license allow for multiple attempts, and are administered by officers trained for that job, and aren't based on a video game installed in a cruiser.

> It'll illegal to drive dangerously though, surely.

Yes, and for all but the worst offenses (say, 30+ miles over the speed limit) the penalty is merely a fine and points against your license.

If someone drives as if they are drunk, all the time, then I don't think they should have a license as they are clearly a danger to public health. On a similar note, the driving test for a license could include a run in the simulator to both test driving skills and establish a baseline for the test. Then the impairment could be measured as relative to that baseline.

I think it's a good idea.

> If someone drives as if they are drunk, all the time

This doesn't test "all the time". It only tests at one point in time.

States have entire sets of due process for removing someone's license due to medical reasons, etc. It's not on the spot based on a video game running from a laptop in an officer's trunk.

Very difficult to prove that the simulation is an accurate and easy to grasp representation that holds up in a court.

You could put the judge and/or jury in VR headsets for the playback. They would see just what the defendant saw, in four dimensions. That's amazingly high fidelity, and the chain of evidence can be demonstrated with a checksum. Compared to judging the inner workings of a breathalyzer, that pretty good.

All that proves is they failed a simulation, not that they were under the influence. Burden of proof lies with the prosecution.

Seems like we should care more about driving impaired than about driving under the influence. The former is what we're trying to prevent by policing the later. You can be impaired without being under the influence, and under the influence without being significantly impaired.

I don't disagree, but I've had moments when I was a terrible driver: I was tired, I was pissed off, I was stressed, I had just received bad news, I was quitting nicotine, etc.

I've done that too, and I should have been busted for that too. Is it really better to risk other lives due to lack of sleep than from too many drinks? For some stranger's little kid in another car's back seat, impaired is impaired.

By that standard, that stranger could be arrested for careless driving because the child in their back seat was screaming for a juice box and they were allowing themselves to be distracted.

Failure to control a vehicle is a violation in the places I've lived. It applies regardless of circumstances. Ice patch, animal runs in front, screaming toddler, whatever the cause, you can still be arrested for it.

In which case you need no VR setup, as the event happened prior to the officer interaction. Most of what you described is a ticket, not an arrestable offense.

Aside from the other issues, how do you expect your average police officer to handle that thing? You know they're not exactly the brightest people, right?

You tell the cop there's a problem with the simulator. He sits down. You point out that other driver just ran a red light. He gives chase on the simulator. You walk away. Sorted.

I have no idea legally how this would work, but is 'proving' impairment using your proposed method enough?

I would think the missing piece would still be linking impairment to a substance or condition that the driver knew about, but drove anyway. ...which is why a simple sobriety test is typically given before a Breathalyzer.

We should care more about getting incapable drivers to stop driving than detecting insobriety. We should want safety police, not morality police.

And strictly on the morality angle, I don't think driving drunk is more culpable than driving over tired or over angry. I find those things are about as easy to detect in myself as drunkenness.

producer of Black Mirror bookmarks this comment

"You suck at video games, enjoy your DUI"

Honestly this sounds unreasonable.

Reply on news.ycombinator.com