Man Tests If Tesla Autopilot Will Crash Into Wall Painted to Look Like Road https://futurism.com/tesla-wall-autopilot
In the water test, Rober has the Tesla driving down the center of the road, straddling the double yellow line. Autopilot will not do this, and the internal shots of the car crop out the screen. He almost certainly manually drove the car through the water and into the dummy.
One person tried to reproduce Rober's Wile E. Coyote test using FSD. FSD v12 failed to stop, but FSD v13 detected the barrier and stopped in time.[2]
Lidar would probably improve safety, but Rober's video doesn't prove anything. He decided on an outcome before he made the video.
Let's get back to my main point, that Tesla's not having Lidar is stupid and I don't trust a self-driving car that can't adequately detect solid objects in it's environment
I'm not defending any of those replies to Rober. In fact I find it quite annoying when dogmatic, sneery people happen to share my views. But the content of those replies does not change the content of Rober's videos, nor does it change the content of the video showing FSD passing the test.
> Let's get back to my main point, that Tesla's not having Lidar is stupid and I don't trust a self-driving car that can't adequately detect solid objects in it's environment
In the video I linked to, the self-driving car did adequately detect solid objects in its environment. My main point is that your main point is based on a video that used non-self driving software engaged seconds before collision, edited and published to make people think it was FSD engaged much farther back from a standstill. And at least one other test (the water test) didn't even use autopilot, just manual driving. I don't know why Rober did that, but he did, and it tanks his credibility.
Again, I'm not arguing against lidar. I already said that lidar would probably improve safety. But Rober's video does not show that, as he didn't use Tesla's FSD software. The person who did showed that it stopped successfully.
In a world where lidar greatly improves safety, we would see the latest version of FSD go through the Wile E. Coyote barrier. That didn't happen, so we probably don't live in that world. In a world where lidar improves safety, though not as much, we'd see FSD stop successfully. And in a world where lidar doesn't improve safety (weird I know, but there could be issues with sensor fusion or lidar training data), we'd also see FSD stop successfully. Right now we don't know which of those worlds we live in. And we won't know until someone (probably Tesla) launches a vision-only robo taxi service. Then we can compare accident rates to get an idea of how much lidar improves safety. And if Tesla doesn't have a robo taxi service within the next year, that indicates that cameras alone aren't safe enough to run a robo taxi service.
I followed Mark Rober on X to learn more about him and possibly understand more about his Tesla tests. Maybe he's a Musk/Tesla hater like Thunderf00t, I don't know. (yes, I'm on X - for entertainment purposes only)