Hacker News Reader

Comment by DustinEchoes | Hacker News Reader

Still not convinced of the safety of lidar. I guess all these cars with cheap lidar sensors on board will generate real world safety data over the next few years.

View on HN

Why not? And cheap is a relative term, from American point of view these sensors may be expensive because they have to buy it from suppliers, from BYD’s perspective it could be home grown given they are by far the most vertically integrated vehicle manufacturer.

See here https://news.ycombinator.com/item?id=46110395

People think you can just slap lidar on and poof, self-driving is solved. There still has to be a software / ML stack. You still have to know what you're doing. You still need a lot of data.

What if the real world safety data over time is... secret retinal damage to millions of walkers and runners, with symptoms attributed to Covid mysteries (and not obviously due to vision), and it takes years more before someone happens to get enough data, and does the right study analysis, and then there's industry with strong incentive not to be on the hook for blinding millions of people?

If the tech industry has taught us anything, it's that big money is still as irresponsible and greedy as ever.

I suppose that one small bit of hope is that one of the most obvious bad actors in general happened to be opposed to Lidar, and might like to screw competitors with a scandal. So the news might come out, after much tragic damage is done.

Lidar is incredibly low power and fast scanning, the retinal risk is probably much less than having to drive when the sun is near the horizon.

However LIDAR safety is currently mostly evaluated on the assumption of a single LIDAR being present. If LIDAR becomes common, with multiple systems per vehicle, the probability of multiple LIDAR beams of different LIDARs hitting your eye at the same time goes up

And that's if scanning never malfunctions.

Everyone is accustomed to cars malfunctioning, in numerous ways.

An intuition from an analogy that should be recognizable to HN...

Everyone is accustomed to data breaches of everything, and thinks it's just something you have to live with. But the engineers in a position to warn that a given system is almost guaranteed to have data breaches... don't warn. And don't even think that it's something to warn about. And if they did warn, they'd be fired or suppressed. And their coworkers would wonder what was wrong with them, torpedoing their career over something that's SOP, and that other engineers will make happen anyway. Any security effort is on reactive mitigation, theatre, CYA, and regulatory capture to escape liability.

I'd like to think that automotive engineers are much more ethical than tech industry, but two things going on:

(1) we're seeing a lot of sketchy tech in cars, like surveillance, and unsafe use of touchscreens;

(2) anything "AI" in a car is presumably getting culture influence from tech industry.

So I wouldn't trust automakers on anything intersecting with tech industry.

There have been reports of lidar damaging camera sensors. If it can damage sensors, it can damage retina. And unlike the sun, it's not visible. Someone could stare straight into it for a good while and not realize.

It's also typically at a wavelength that has lower risk to human eyes (ie near infrared).

Laser eye safety risk is very measurable and well classified.

Reply on news.ycombinator.com