Mark Rober just set up one of the most interesting self-driving tests of 2025, and he did it by imitating Looney Tunes. The former NASA engineer and current YouTube mad scientist recreated the classic gag where Wile E. Coyote paints a tunnel onto a wall to fool the Road Runner.
Only this time, the test subject wasn’t a cartoon bird… it was a self-driving Tesla Model Y.
The result? A full-speed, 40 MPH impact straight into the wall. Watch the video and tell us what you think!
I saw the video pop up in my Youtube recommended, but didn’t bother watching because I just assumed that any cars tested would be using LIDAR and thus would ignore the fake road just fine. I had no idea Tesla a) was still using basic cameras for this and b) actually had sophisticated enough “self driving” capabilities that this could be tested on them safely.
They are not still using cameras but removed LIDAR and radar from their cars during the chip shortage 2020/21. The story they were telling was “humans don’t have LIDAR but can drive cars as well, so the cars also only need ‘eyes’ like humans”.
Small correction here: they never had LIDAR. Cars with LIDAR have big racks on top with a spinny thing measuring the surroundings. Teslas had radar but removed during the chip shortage (and disabled it on existing cars) and acted like it was an improvement. The radar was used for distance keeping on cars and could actually detect the car in front of the car by bouncing signals off the ground, it was really slick.
Spinny thing is just when you mount one ontop. It doesn’t have to be. The example in the video appears to use a forward facing cone LIDAR. Presumably in addition to other sensors.
Humans cannot, in fact, drive cars well. Humans kill tens of thousands of other humans with cars every year in the US alone.
Yup, cameras and humans share various exploits. Self-driving is going to work better than humans once every car has it and communicates with each other, allowing for minimal gaps even at high speeds, once roads are all very standardized and in a database, and-
Wait, that’s trains
Fucking build more electrified high-speed rail and forget tech bros’ shitty promises
I was getting mildly outraged and ready to comment how you were re-deriving the train at first. Well played.
They really had me in the first half, not gonna lie
Trains don’t go from my driveway to my destination exactly when I feel like going there, while carrying all my luggage.
I get that it’s fun to be smug on the Internet, but private vehicles aren’t going away any time soon.
It’s not a binary decision between all cars and no cars. If trains and public transit have enough capacity and convenience to make most trips feasible by them, car infrastructure will no longer have to be added (in fact can be converted into bus and bike lanes) while shortening trip duration (less cars = less jams) and improving safety.
Also, you barely have luggage for most trips. 99% of my trips are made with luggage I can carry to the nearest stop and board the bus with.
Yeah it’s not a binary decision, but trains are almost never the answer for a lot of people. If I’m going less than a couple hours, then I’m driving that distance. If I’m going much further than that, I’m flying. If I need to move a ton of stuff, I’m either taking my car or renting a uhaul. If I’m taking a lot of people, I’m taking my car. Trains never enter the picture unless I’m looking for variety in my mode of transport.
And trains do not shorten the trip duratiion, not without absolutely kneecapping the roads. And over long distances, they’re absolutely slow compared to planes. In the short distance, they’re slow compared to cars.
Depends on where you live. In most of Europe, trains are frequent and direct between city centers.
My parents tend to prefer the car for the 3-hour trip (also 3 hours by train and bus) to Grandma’s when at least 3 people go because it’s cheaper. A higher toll on the highway could change the threshold, and we’d go more comfortably. Politicians can smoothly adjust the number of people for which public transport wins out with taxes and investments. You’re more likely to cling to the car and they’ve accounted for that in their models, maybe making you switch for a specific kind of trip is not worth the investment. There are lots of factors, such as political alignment, culture, wealth distribution, existing infrastructure etc. that make some jurisdictions able to move the threshold faster than others. Still, the majority of people using cars is unsustainable for lots of reasons:
so there is an obligation to eventually push the threshold in favor of public transit for most trips.
It definitely makes sense in Europe. It really doesn’t in the sparse US west, but it might in the US east where it already exists due to population density.
In the west, my neighborhood is larger than many European cities. That isn’t hyperbole either, it’s 30,000 acres/121km2. Not completely developed yet, but that is the full size of the area.
I would love light rail or high speed between cities but it would take a century and Trillions to do and then I’d still have to drive to a station. There’s just no way I’d be able to walk like I did in Germany. I do wish, it was amazing.
Very well stated. Thank you.
I have a van load of tools to transport most of the trips I make.
So you’ll keep using it. And enjoy narrow but way less jammed streets. Maybe you’ll be incentivized/required to join the self-driving network, but in decades, not years, after positioning markers have been added to every road in the last repaving, while infrastructure funds have been directed towards making the city traversible for non-drivers.
See, the problem is, where I live already has great public transport, including electric commuter trains and buses, lots of cycle lanes etc.
And, despite that, traffic is still shit, because there’s a massive chunk of the population who need to be on the road for whatever reason.
That should tell you just how vastly complicated driving is.
There are plenty of uncomplicated things that humans do poorly, too.
Walking up and down stairs, not that I’d know.
And the really dumb thing is that lots of modern non-selfdriving cars now have lidar sensors to help the humans not crash into things. Musk apparently wants the AI to be working at a disadvantage.
He just wants people to buy his junk, and doesn’t care how many people would have to die as collateral damage.
That statement of him is not entirely wrong. But we humans have a very powerful bio computer that is perfectly tuned to process those visual inputs in realtime. Until a comparable performance is possible, removing LIDAR is very stupid.
Besides that, in the fog and rain tests a human likely would have killed a kid anyway, and why settle for human limitations when you could be safer?
We absolutely should also have lidar or analogous tech as part of a solution here, even if cameras did manage to get to human level safety.
The child dummy was clearly visible through the water in the rain test. Tesla’s systems just suck.
Also competent drivers generally know to slow down during rain. Hell I was literally taught to drive some roads like its a speedway and even I drop below the speed limit during the rain if visuals are bad enough, especially first rain pulls oils out of the road makes it slippery and may cause hydroplaning.
And let me just add, Musk ordered the LIDAR removed against the engineers better judgement.
Did he have lidar removed? I thought it was radar. I didn’t think any Tesla vehicle ever took on the cost of lidar.
This is true. There never was LIDAR.
I’ll add that every other self driving car company has a pretty good safety record, specifically because they do use LIDAR and RADAR so they can see better than humans.
IIRC Musk said it would rely on AI using the footage from all the Teslas and it’s better than LiDAR. That idiot was proven wrong once again.
They tested a LiDAR rigged car, and it stopped just like you predicted. As of 2021, Tesla uses only cameras for FSD, and not even radar (which my stupid fine Toyota truck has).
They tested the idea safely by building the wall out of styrofoam, or at least that’s what it looks like when it blows apart :)
Front-facing radar is the bare minimum needed to pass the test given (fake-road wall). Many vehicles use it for adaptive cruise control, and radar is even faster than either cameras or lidar for figuring out the range to an object. 1000 Hz measuring distance to an object is enough to find both the relative velocity and the acceleration of another object. This provides enough time to apply the brakes safely when approaching a vehicle or obstacle
LIDAR is even better, and also more compute intensive and expensive to install.
I think Tesla was very short-sighted in removing radar sensors, certainly. If they hadn’t, they could’ve spent more of their energy on making the FSD cars better instead of just making them sufficiently safe with insufficient sensors
Forget lidar, they don’t even have mature tech like radar for emergency braking. Edit: +even
Why do you think lidar is not mature? It is radar, except it uses light and can get much more resolution than an RF radar. Or was that a joke… That was probably a joke… if it was then nm.
Lidar is mature but its automotive application is not. Radar is basic by now in comparison.
That comment just missed the word ‘even’ - as in they don’t even have radar, and that’s on regular non-self-driving cars, and lidar would be a step above that.
Not a joke, lidar struggles with fog/rain, which radar can handle much better.
You know what struggles even more in fog and rain? Camera only systems 😄
And ?most? humans. 😉
As far as I’ve seen, any system would be additive. If it has lidar, it would also have cameras and radar. So that you get the best of all the technologies (e.g cameras are the only only of the three that can follow lane markings)