Hacker News Reader

Comment by akira2501 | Hacker News Reader

The fact that they don't highlight any night testing is troubling, and that they have a single closed course facility in a place where it never snows.

I often wonder if these companies are truly trying to revolutionize driving, or just trying to put a couple of "Johnny Cabs" in the southwest and call it a day. Their strategy really does seem geared towards the latter outcome.

View on HN

Recently there have been Waymo vehicles out on the streets in the Seattle area. I even saw one driving around the day it was snowing here.

I suspect we’ll start seeing more training and testing in places with less nice weather as time goes on.

Night isn't especially concerning for vehicles that have more sensors than simply cameras; headlights and infrared cameras exist. Rather, dealing with winter conditions seem like the reason that these will be confined to the southwest.

> Night isn't especially concerning for vehicles that have more sensors than simply cameras

This is precisely the type of presumptions I'd like to see tested. I mean, if you're going to go to the trouble of all this and have a 118 acre road course, it seems the height of hubris to just say "well, the sensors will probably be better than cameras at night."

It's amazing to me that on Hacker News people people have this puritanical embarrassment over obvious technical questions.

Oh this is definitely false!

Night is very concerning and is a problem for computer vision systems.

During the day illumination is fairly consistent. Sure the sun moves around, but it's not a spotlight. At night, illumination varies a lot. A detector that works during the day may not work at all at night. This is no joke, there are papers that show dramatic performance losses for tasks like pedestrian detection at night.

Cameras are worse at night. They need to be more sensitive which dramatically increases their noise. They may need to have longer exposures leading to blur, which is of course made worse if you are moving the camera.

Headlights also don't provide the same visibility so your reactions must be faster. Reflections are a big problem too. Oncoming headlights are also an issue.

Nighttime testing will be critical.

>Cameras are worse at night. They need to be more sensitive which dramatically increases their noise.

This is solved by having multiple sensors. Nighttime is really not a problem, especially given the broad spectrum and LIDAR.

It could just be that night testing doesn't look as good in a blog post.

Waymo has been testing in other cities, including ones with snow, for years.

One example: https://arstechnica.com/cars/2017/10/waymo-starts-testing-in...

These speech recognition systems don't really work that well when there's a lot of speakers present, talking over one another. Are these researchers even trying to revolutionize speech recognition? it seems their only goal is to make speech recognition work with one speaker in a silent room and call it a day.

What does that have to do with self driving cars? These are meant to be used on roads in varied conditions. Your comment isn’t applicable.

Some conditions don't exist in many places. Someone can have a self-driving car and use it for life in Arizona and it doesn't matter that it wouldn't work in Alaska. It's still progress and still useful to a subset of people. Hell, even a car that works 9 months out of the year in the Northeast would be useful. Don't see too many bikes out when it's -5 degrees but no one is claiming bikes aren't useful.

To be fair I have seen a large number of people say that bikes aren't useful because they can't perform a supercommute or can't be used when it's below 0.

I think the safety margin between a speaker and a car are two entirely different things. If your smart speaker fails to work for you, this does not offer any potential to harm me.

I think it's worth putting them into different classes, don't you/

Research and all tech advancements are done in increments, there's no other way about it. I see no point in tackling rough weather conditions when the basics aren't even finished. Of course for the product to actually hit the roads it needs to be held to a very high standard but that's beside the point. You were criticizing their progress, even though we know they don't have a final product yet and they're probably some ways off.

It's like criticizing the Wright Brothers, saying stuff like "meh, they barely were off the ground, who cares"

Reply on news.ycombinator.com