FSD able to compete with lidar! Mark Rober Debunk Video - Heavy Rain Test - 2026 Tesla Model Y HW4 FSD

[freyguy]

The vehicle tested is a new model y but it has same tech as the Cybertruck so thought I’d share how well FSD is able to perform in substantial rainfall. And shoutout to the cool firefighters who help make this test possible. Every day I feel safer and safer not just as a person in a vehicle but also to pedestrians in the steeets.

Sponsored

 

[no%X#XMVk65v#cq]

Have driven in rain, can confirm. Pull up the cameras, can't see shit. But cybertruck interprets that camera data differently than our brains, and still sees things we can't.

 

[yohst]

Kewl!

Now because this test send a crap load of water in the road, it feels like a tough test but that might be wrong. For one, this was in broad daylight, in a real scenario with heavy cloud cover, there would be much less dayslight and hence the cameras would not have as good contour contrast. Additionally, this was more like a rain “curtain”, a longer traject with water coming down where the car is in full “rain” for longer than a second or so would make it more challenging.

Would have been nice if you shared video from inside the cabin going through the “rain”.

Still impressive results of just camera based FSD, and seems like you guys had fun!

 

[Docedward]

I drove in the recent downpours in Nashville area- FSD simply shuts down and you have to take over- so, unless the person is present in the 20 feet between me taking over from FSD driving- I’d still hit them. No?

Also, agree about the daytime light is not realistic. I love the fact that it sees the dummy and stops but would love for it to give me a choice to allow it to override me hitting the dummy as well. I guess there’s been too many false positives so it’s not available.

 

[freyguy]

Kewl!

Now because this test send a crap load of water in the road, it feels like a tough test but that might be wrong. For one, this was in broad daylight, in a real scenario with heavy cloud cover, there would be much less dayslight and hence the cameras would not have as good contour contrast. Additionally, this was more like a rain “curtain”, a longer traject with water coming down where the car is in full “rain” for longer than a second or so would make it more challenging.

Would have been nice if you shared video from inside the cabin going through the “rain”.

Still impressive results of just camera based FSD, and seems like you guys had fun!

The main point of this video was to validate mark robers claims of FSD.

a few weeks back he put Tesla camera system against lidar in a “non sponsored “ video. This test the Tesla failed but mark was using autopilot and not FSD with latest hardware.

so the person making this video just re enacted mark robers test where Tesla was basically shitted on.

 

[Docedward]

Agree. Mark Rober’s money grab for LiDAR was a FUBAR for his career and a “Me Too” hate on Tesla. And this was a MUCH better test while Mark’s did not even approach reality and his failed slight of hand with Autopilot and FSD was criminal.

 

[freyguy]

Agree. Mark Rober’s money grab for LiDAR was a FUBAR for his career and a “Me Too” hate on Tesla. And this was a MUCH better test while Mark’s did not even approach reality and his failed sleight of hand with Autopilot and FSD was criminal.

exactly!!!! I’m just glad there are others out here with resources who find ways to price or disprove a claim.

with marks reach of followers you can’t tell me he had no bad press towards teslas. He made these cars out to look like murder on the road.

very disappointed in mark but glad others can show us the TRUTH!!!

 

[tingmo13]

'Tesla FSD in China' YT videos comparing other Chinese EVs sophisticated tech proves it's more than Lidar or Radar. Tesla FSD is doing amazing with less parts.

 

[yohst]

The main point of this video was to validate mark robers claims of FSD.

a few weeks back he put Tesla camera system against lidar in a “non sponsored “ video. This test the Tesla failed but mark was using autopilot and not FSD with latest hardware.

so the person making this video just re enacted mark robers test where Tesla was basically shitted on.

Got it, hadnt seen the robers video, usually don’t click on clickbait

 

[freyguy]

Got it, hadnt seen the robers video, usually don’t click on clickbait

Me neither but I been a long follower of him and his engineering, but that video left me with a sour taste for his content going forward.

 

[bluerice]

Not sure this tests anything other than Tesla reacting to a wall of water. You should've REMOVED the dummy in another set of tests to see if reactions were different.

 

[Docedward]

That’s exactly what I was thinking. Remove the dummy and see if it slows and stops still

 

[Sjohnson20]

didn’t he get caught faking some of that glitter bomb video? That’s all I know him for and it’s been a long time since I watched that.

 

[freyguy]

didn’t he get caught faking some of that glitter bomb video? That’s all I know him for and it’s been a long time since I watched that.

This is news to me… I’m about to go down a rabbit hole soon it seems because if that’s the case I’m really dissapointed.

 

[Dazureus]

Not sure this tests anything other than Tesla reacting to a wall of water. You should've REMOVED the dummy in another set of tests to see if reactions were different.

Yeah, I wasn't seeing a person visualization on the screen, but it wasn't the clearest picture. I think FSD was stopping because the road cues were obscured with the water.

Sponsored