Hacker News Reader

Comment by rswail | Hacker News Reader

The problem is not drunk drivers but incapacitated drivers, whether due to alcohol, drugs, lack of sleep, distractions (mobile phones), etc.

The fix is not arbitrary alcohol blood levels, or extending the "War on Drugs", but with standardized, enforced, in-vehicle continuous testing that ensures that the driver is awake, responsive, and in control of the vehicle.

That is something where R&D in AI/ML/software to analyze and detect incapacity could/should be directed.

It can be entirely on/in-vehicle, with access to the data only available to law enforcement (and, perhaps, insurance companies), if and when it is needed to determine liability (not as a general "carte blanche" for the data), in the same way that aircraft black boxes are only analyzed for investigation after an incident.

View on HN

I have a suspicion that most accidents involve the worst 10% of drivers.

I'd like to see research into ways to identify those drivers and encourage them to improve their driving skill.

For example, one way might be to make a register of driving skill public information, and then insurance companies can jack their prices up for those people unless they go on and pass (government funded) extra training courses.

Such a register could simply be an open database of insurance claims, or police crash records showing the names of people involved.

Here's a starting point for you - "Estimating likelihood of future crashes for crash-prone drivers" at https://www.sciencedirect.com/science/article/pii/S209575641... .

> At-fault crash-prone drivers are usually considered as the high risk group for possible future incidents or crashes. In Louisiana, 34% of crashes are repeatedly committed by the at-fault crash-prone drivers who represent only 5% of the total licensed drivers in the state. ... From the findings, it is recommended that crash-prone drivers should be targeted for special safety programs regularly through education and regulations.

>Such a register could simply be an open database of insurance claims, or police crash records showing the names of people involved.

In the US, these are CLUE and A-PLUS

https://risk.lexisnexis.com/products/clue-auto

https://www.verisk.com/insurance/products/a-plus-commercial/

The fix is to make it so you aren't left stranded in life without a car.

Car dependency is just as bad a disease as any.

Cars give you the freedom to go anywhere, but take away the freedom to do anything without them.

Hi. I'm a wheelchair user. It's only reasonably recently that the buses here had a restraint system that can lock my chair in place added, before that I needed to hold on to a bar next to me, which destroyed my shoulders.

Uber decimated the taxi industry, the pandemic finished them off, and now taxis need to be sent from one town over, and those drivers don't accept my government subsidy card.

At my fittest, I could push my chair about 10km per day.

I need a car.

You are a wheelchair user, yes, but you don't really address the parent. The parent is saying that we should make society more accomodating, which you acknowledge with mention of restraint systems on public transport. I take it that's what they're asking, so you don't "need" a car, but have viable options everywhere.

Sounds like well designed transit and taxis would serve you to some extent. As would microcars or an electric third wheel.

None of these things prevent you from having a car, and driving in areas that prioritise them is safer, faster, and more pleasant for the people that need cars.

Car dependency does eliminate transport for all of the people with disabities that prevent them from driving though.

I need a car

And this should make you angry. No one should be reliant on a car. Everyone should have access to a robust public transportation system. Even if that means you personally get a subsidized van to take you where you need to go. Subsidized so it costs the same as a bus ride.

Why do you think implementing excessive amounts of bureaucracy is a better solution than just letting people do their own thing?

If you need a car you should push even harder so that fewer people who don't really need a car drive. More space for you on the road.

I live in London which has all-but banned cars in the city centre (they're not banned, but there is an expensive congestion charge which means that in practice most traffic in the centre is buses, taxis, or very wealthy people).

However, there is an exception for anyone with a "blue badge" holder (for those with disabilities which prevent easy mobility - wheelchair users and others). These people not only get free car access to the city centre (which now has a lot less traffic because nobody else is driving there), but also get to park in places that most people are not allowed to, etc.

> there is an expensive congestion charge

Pales in comparison to the cost of parking. An afternoon in the Q-Park at Westminster is £45, 3 times the price of the congestion charge.

Yes, but again this doesn’t apply to blue badge holders who park for free in all sorts of places where you’re not usually allowed to park.

10k is a respectfull number on an ordinary wheel chair. I know two wheel chair users who use an armbicycle, like an extra motor assisted handcrank that is attached on their wheelchair. They easily do about 15km per day (considering they have no car and live 8km from work).

That you need a car is not a given but as you say a result from stupid decisions you can't control.

The problem is neither incapacitated drivers, drunk drivers, drivers without seat belts, cars with bad breaks, car without headlights or any other contributing sources to car crashes. The problem is the car crash. Everything else is just contributing risk that a car crash might occur or that people will get serious injured if there is a car crash. In-vehicle continuous testing will not remove all risks to car crashes, and with drugs as an example, amphetamine makes people both awake and responsive but it also carries a risk of delusions and paranoia which is linked to high risks.

AI/ML/software is know for its high false positive rates, so if it was enforced it would lead to many drivers unable to use the car because of a false positive hit. If it was enforced, then the high risk reduction wouldn't occur before the crash and society in general will want to impose laws to fix that. Car crashes effects more than just the driver, and it not reasonable that so many people die from car crashes each year.

Reply on news.ycombinator.com