6 ms·

What the hell have we been working on if not that? What good is an autonomous driving system that is not able to consider what happens next?!?!?!?!????

asdfasgasdgasdg 7y ago
Presumably that? I think the author is just saying that it's going to get better. He is not privy to what his competitors have been working on, in any case, except whatever has been announced publicly. (Unless he is doing industrial espionage.) Maybe it's just that Voyage's cars weren't predicting up until now?
- loopz 7y ago
  I would be careful in assuming competence in such a new dev area. Think: rules vs predicting vs rise in false positives Heard of control systems waiting for seconds before acting on new information? So don't assume flawless, perfect technology.
  - asdfasgasdgasdg 7y ago
    I don't think there's any implication that the technology is perfect in my comment. But the idea that e.g. Waymo's software is not making some attempt at predicting future states, even implicitly, is absurd. This video[1] was released two years ago and references prediction at the provided timestamp. [1]: https://youtu.be/B8R148hFxPw?t=77 https://youtu.be/B8R148hFxPw?t=77
- olivercameron 7y ago
  We've been working on the prediction problem for some time. We should have a blog post coming out about our approach (it's pretty neat and in-line with this post) soon.
  - asdfasgasdgasdg 7y ago
    That's great to hear! I think upon fully reading the blog post it's clear that you meant the next leap forward it solving prediction, to the degree that perception has supposedly been solved. It will be interesting to see how y'all do. I for one want my self driving car yesterday.
sgt101 7y ago
Well "even my Tesla" has a collision prediction system (spoiler; it's paranoid) so - yup, you're right
ogre_magi 7y ago
Did you know that in English, you can end a question with just a single question mark?
fxtentacle 7y ago
Not crashing into fences. It's still an ongoing area of research in computer vision on how to detect repetitive objects made out of thin parts. Radar cannot reliably detect them due to being thin and optical methods are not yet advanced enough.

asdfasgasdgasdg 7y ago

Presumably that? I think the author is just saying that it's going to get better. He is not privy to what his competitors have been working on, in any case, except whatever has been announced publicly. (Unless he is doing industrial espionage.) Maybe it's just that Voyage's cars weren't predicting up until now?

loopz 7y ago
I would be careful in assuming competence in such a new dev area. Think: rules vs predicting vs rise in false positives Heard of control systems waiting for seconds before acting on new information? So don't assume flawless, perfect technology.
- asdfasgasdgasdg 7y ago
  I don't think there's any implication that the technology is perfect in my comment. But the idea that e.g. Waymo's software is not making some attempt at predicting future states, even implicitly, is absurd. This video[1] was released two years ago and references prediction at the provided timestamp. [1]: https://youtu.be/B8R148hFxPw?t=77 https://youtu.be/B8R148hFxPw?t=77
olivercameron 7y ago
We've been working on the prediction problem for some time. We should have a blog post coming out about our approach (it's pretty neat and in-line with this post) soon.
- asdfasgasdgasdg 7y ago
  That's great to hear! I think upon fully reading the blog post it's clear that you meant the next leap forward it solving prediction, to the degree that perception has supposedly been solved. It will be interesting to see how y'all do. I for one want my self driving car yesterday.

sgt101 7y ago

Well "even my Tesla" has a collision prediction system (spoiler; it's paranoid) so - yup, you're right

ogre_magi 7y ago

Did you know that in English, you can end a question with just a single question mark?

fxtentacle 7y ago

Not crashing into fences. It's still an ongoing area of research in computer vision on how to detect repetitive objects made out of thin parts. Radar cannot reliably detect them due to being thin and optical methods are not yet advanced enough.