I've always been skeptical of skipping LIDAR, but a recent study at Cornell found that with binocular vision they could be almost as accurate as LIDAR, but with much cheaper hardware.
https://www.therobotreport.com/researchers-back-teslas-non-l...
https://www.therobotreport.com/researchers-back-teslas-non-l...
It's fairly clear machine vision is going to surpass LIDAR in the near future (taking capability and cost into account), doesn't take a robotics expert to see that.
It's just a matter of time before AI can drive better than humans even if using similar vision. Especially in aggregate (e.g., factoring in distracted drivers).
I am not considering here situations where either system is impaired since that will lead to varied degrees of failure with both.
And you're oversimplifying a little, it goes far beyond just avoiding obstacles. You need to be able to anticipate, learn and adapt to completely new situations, understand consequences and things that aren't strictly related to what happens on the road. That's something even most mammals fail to do. Our computers are not there and they won't be for some time.
Garbage in garbage out. Don't skimp on your sensors.
I'm not convinced absolute speed is that important of a variable, when comparing to something much smaller dealing with closer distances, but some birds actually do reach highway speeds and beyond: https://www.worldatlas.com/articles/the-fastest-birds-in-the...
But the fastest relative to size might be the hummingbird, which can fly 30mph backwards: https://www.ponderweasel.com/what-bird-can-fly-backwards/
The Anna's hummingbird has been clocked at up to 50mph in a dive, or 385 body lengths per second. https://www.wired.co.uk/article/the-hummingbird-thats-faster...
How is it considered unacceptable? It has been acceptable for humans to drive in every country since 100+ years.
Anyway, the biggest problems with humans driving that cause accidents are distractions like smartphone usage, sleepiness, excessive speeding, tailgating, road rage, drunk driving, old age, being new to driving, being tired, medical problems like strokes, seizures, diabetic comas etc. which automated driving will solve.
VSLAM, visual SLAM, which is what you have to employ if relying largely on vision-only, is so far cruder than that.
I'm sure Tesla has some sensor fusion that also employs GPS and short-range radar/ultrasonar to help.
You can get some velocity and positional data through GPS, but it will have quite a lot of latency and is very coarse (at least the commercially accessible part).
SLAM is much, much easier to do with LIDAR since that will give you a point cloud 3D model. With VSLAM, you have to infer the depth data and build it up over time, and it's a much harder problem to solve reliably. OTOH, LIDAR's Achilles heel is, however, heavy rain/sleet/snow. Waymo has shown off using deep learning to filter the bulk of the noise out, but it looks like it's still early days there.
Vision isn't great for sleet/snow either, to be fair.
[1]: simultaneous localization and mapping
I look forward to the day when cars have giraffe-necks protruding from their roofs, mounted with stereo cameras...
But until then, the researchers actually just confirmed that front-mount cameras are only about a third as accurate as LIDAR.
And as a bonus, if the car flies off the road into the river, it can use a fishing pole style winch to pull your vehicle out of the water and place it back on the road.
The picture and the words in the article show cameras next to the rear view mirror.
""" Using two inexpensive cameras on either side of a vehicle’s windshield, Cornell researchers have discovered they can detect objects with nearly LiDAR’s accuracy and at a fraction of the cost."""
That being said, existing self-driving cars already have some cameras at that level without displaying the sort of accuracy the researchers found, which relates to the methodology issues other commenters have noted.
In any conceivable situation where you'd use this, it would actually be better to just use LIDAR.
Why don't you read the paper a bit -
https://arxiv.org/pdf/1812.07179.pdf
And then work out why they used the words "birds eye" that you found so upsetting.
The research paper, in a nutshell, was about transforming stereovisual images to psuedo-birds eye data, like roof-mounted LIDAR. This, theoretically, would result in about a 350% improvement over current vision-based methods...but still only 50% of that of existing LIDAR systems. In exchange, you just have to 1) convert the stereovisual data to a 3d point cloud, 2) transform the point cloud to birds-eye perspective, and 3) run the LIDAR systems on the point cloud, and 4) do that all in milli-seconds.