I'm not going to knock Tesla for trying a primarily visual-only approach, but I will knock them for insisting that it will work out to level 4 autonomous driving with just software.
I predict that these kinds of unexpected edge cases will keep popping up for several years. It certainly will be interesting to see how robust their image processing algorithms get as time goes on, but I wouldn't hold my breath for reliable full self driving for a while.
I would hold it against them. I think we should start with as many sensors as is necessary to get it working robustly. Over time we can then strip one or the other sensor as we see what we can do with software improvements.
First get something working, then reduce costs, not the other way around.
Just planning ahead. Musk knows that soon he'll be on Starlink 5, completely blocking out the sun and moon. Why bother with programming a workaround for an issue with a better solution coming!
I had this happen on my commute to work, but from the sun. It's really smoky here, and the sun was breaking through the smoke just enough and at the right angle to look like a yellow light. It was pretty amusing to see a bunch of yellow traffic lights stream by as I was driving.
ayane_m|4 years ago
I predict that these kinds of unexpected edge cases will keep popping up for several years. It certainly will be interesting to see how robust their image processing algorithms get as time goes on, but I wouldn't hold my breath for reliable full self driving for a while.
carlmr|4 years ago
First get something working, then reduce costs, not the other way around.
BenjiWiebe|4 years ago
HWR_14|4 years ago
Tempest1981|4 years ago
https://news.ycombinator.com/item?id=27934539
kevin42|4 years ago
high_byte|4 years ago
simonh|4 years ago