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!
According to Ol’ Elon the robo-taxi service has been a couple months away since 2017 or so. I can’t imagine it’s much closer now than then.
It’s right at the end of the tunnel they’re diggin in CA
I think Rober just showed us why. Mowing down kids in weather is an unacceptable amount of risk.
Unacceptable risk to you. I’m guessing Elon is fully prepared to take the risk and minimise the consequences.
That’s what insurance is for after all
Don’t need to have insurance when anyone hit by a Tesla was a Hamas sympathizer. Then you’re not killing people, you’re killing terrorists for protesting Elon. You can get a presidential pardon for that. /a (maybe? We’re in strange times)
Even Leon Skum has to deal with insurance companies, for now.
Honestly all the fails with the kid dummy were a way bigger deal than the wall test. The kid ones will happen a hundred times more than the wall scenario.
Some sort of radar or lidar should 100% be required on autonomous cars.
I fully agree, but sadly, investors likely care more about their cars hitting walls than hitting kids. Killing a kid or pedestrian in the US is often a very cheap fine. When my uncle was run over on a sidewalk next to his son, the police ruled it an accident and the city refused to do anything. Same thing happened when my friend was ran over in a bike lane… So killing humans is probably cheaper than hitting a wall.
Interesting that in the most consumerist nation on earth, objects have more value than people.
I think insurances will require that is it comes to self driving at least here in Europe.
EU leading the world in consumer protection laws yet again
All these years, I always thought all self driving cars used LiDAR or something to see in 3D/through fog. How was this allowed on the roads for so long?
They originally the model S had front facing radar and ultrasonic sensors all round, the car combined the information to corroborate it’s visual interpretation.
According to reports years ago the radar saved Tesla’s from multiple pileups when it detected crashes multiple cars ahead (that the driver couldn’t see).
Elmo in his infinite ego demanded both the radar and ultrasonics be removed, since he could drive with out that input so the car should be able to… also it is cheaper.Exactly, my previous car (BMW) once saved me in the fog by emergency braking for something I wasn’t able to see yet. My current car (Tesla) shuts down almost all safety features when the camera’s can’t see anything, so I doubt it will help me in such situations. The only time my Tesla works well is in perfect conditions, but I don’t live in California.
Exactly, my previous car (BMW) once saved me in the fog by emergency braking for something I wasn’t able to see yet.
If you were driving at a speed at which the low visibility would have gotten you into into an accident due to some obstable you weren’t able to see yet, you were driving too fast. Simple, isn’t it?
While true, it’s still nice that super-human senses are looking out for the driver on their behalf. Also it’s nice if super-human senses allow for braking earlier and closer to graceful rather than standing hard on the brakes because of late notice.
Fog is one example, but sudden blinding glare could be another situation that could be mitigated by things like radar and lidar. Human driver may unexpectedly be blinded and operating at unsafe speed without any way of knowing that glare was coming in advance.
These things will make people more complacent and lazy, and will absolutely lead to worse drivers and more collisions
Just like government hand outs… Prohibiting accidents is communism, dyind on the grill of a SUV is a patriotic duty… /s
It can be a huge help, depending on the human factor.
If it’s a ‘oh, take your hands off, it’s fine, take your eyes away, it’s fine’, then I could see that the systems replace human weakness but add their own, failing to reach a good “best of both worlds”.
If it’s one of the systems that watches the driver’s eyes and nags if they take their eyes or hands off the task of driving while also encouraging good lane positioning and sufficient, yet perhaps uncomfortable braking in an emergency situation. Enough assistance to aid safety, still annoying enough to make people not rely solely upon them.
Challenge is that’s not a very appealing promise of value. “Our system improves safety by using all this ADAS, but is annoying enough to keep you engaged!”.
As you say, it’s nice if there is an additional assistant, also for e.g. health emergencies.
That said: Driving assistants should only ever be that: assistants. They are not a replacement for safe and controlled driving. I know I’ve been an arsehole on some occasions when I had my driver’s license fresh, and I got lucky that I didn’t have any accidents until I learned to calm down and drive with respect for other people and animals. Just throwing that in here to say I don’t consider myself a saint. But anything “self driving” should be forbidden everywhere, unless it’s on rails that the vehicle can not reasonably escape even if it wanted to (i.e. trains).
Definitely a component of these safety systems needs to be actually effective driver monitoring. You have cars now doing gaze tracking, and tracking things like whether the person seems drowsy. Even while driving unassisted they will nag you if it can’t confirm your attention (I would get dinged sometimes on steep ramps because my arms would block the cameras while turning the wheel, it frankly trained me to reposition hands earlier just to not get the nag).
I used the lane centering to help my kid get used to the sense of correct positioning in the lane. Of course turning it off to make them do it manually, but kind of like training wheels when the kid was tending to push it almost over the passenger line.
I’d be very curious to know how much cheaper it is. Sure, there’s R&D to integrate that with everything, but that cost is split across all units sold. It feels like the actual sensors, at this scale, can’t add a significant amount to the final price.
Back when Elon made avoiding LiDAR a core part of his professional personality, it was fairly expensive. But as any tech genius can tell ya, component prices drop rapidly for electronics.
Now, radar is dirty cheap. Everything has radar. Radar was removed from Teslas. A radar sensor for my truck is $75, probably much less at scale orders.
LiDAR sensors cost anywhere from $500-$1,500 for a vehicle of this type, near as I can tell (this type being Level 2 autonomy rather than something like a Waymo. A well-kitted out self-driving vehicle has 4 LiDAR sensors).
Here is the LiDAR module currently used on the Mercedes S-Class, it’s $400 used: https://www.ebay.com/itm/285816360464
It’s a hideously small cost-savings in 2025 for a luxury vehicle like a Tesla. Any rational company would’ve reversed course after the first stationary-object-strike fatality. Tesla is not a rational company.
I think it’s all about the timeline. Tesla gambled on cameras before AI models became usable (the company most certainly committed itself to the camera sensors a few years before it became public). By the time automated driving models became usable, Tesla had tons of camera data to capitalize on, but presumably not the corresponding radar data (or not in a consistent manner), so rebuilding a multi-sensor dataset for AI training was probably not very appealing in terms of cost and time to market.
tesla uses cameras only, i think waymo uses lidar.
Most non Tesla brands that have some sort of self-driving functionality use lidar and/or radar. I’ve got a BMW iX and as far as I know it uses cameras, radar, lidar, and ultrasonic sensors.
It’s the only sensible approach. Not just is the notion that “humans use just their eyes too” completely wrong (otherwise how would be able to tell that something is off with the car “with our butt”?), computers are not even remotely close to our understanding and rapid interpretation of the world around us or cooperation beyond of what’s pre-programmed, which is necessary to deal with unforeseen circumstances. Cars must offset this somehow, and the simplest way to do so is with vast sensor suites that give them as much information as possible. Of course many humans also utterly fail at cooperation and defensive driving, but that’s another problem.
I remember reading that tesla only uses cameras for it’s self driving. My 2018 Honda uses radar for the adaptive cruise so the technology exists, musk is just an idiot.
Does it? My 2023 model throws a shit fit if it’s cold and I assume the camera covers are iced over.
It probably has cameras as well, for lane guidance etc.
My Mazda complains if the windscreen is dirty for the same reason.
Radar would not detect a Styrofoam wall either the return from Styrofoam is extremely low. Radar also can not distinguish elevation differences very well so an overhead road sign can be mistaken for a stopped vehicle or a stopped vehicle mistaken for an overhead road sign.
Radar doesn’t detect stopped objects at high speed. It’d hit the wall too on radar alone.
This has to be solved by vision and or lidar.
Unless your car is traveling faster than the speed of light, radar will detect objects in front of it. But yeah, I was trying to imply that for a complex system like self driving musk is a buffoon for relying on a single system instead of creating a more robust package of sensors.
They get filtered out and the car will not act on it because there is so much noise from stationary objects all around you. The car essentially wouldn’t drive at all if it didn’t filter them out.
At high speeds, the radar in all cars is used to detect moving objects and the change in velocity of those objects.
Radar will not prevent running into this wall at 40mph.
People can downvote me all they want, but that doesn’t change anything.
Only vison and / or lidar would stop for that wall at 40mph.
Edit: aside from clarity on the above this is the expected outcomes
Radar in cars today: hit the wall
Vision: probably all hit the wall but could be sufficiently programmed to not if they trained on it.
Lidar: would not hit the wall.
You’re partially right, stationary objects on the side of the road will have a different Doppler shift than a stationary object in front of the vehicle, items on the side of the road can be filtered. Cheap radars with low sampling rates will not be able to distinguish as their Doppler bins are fairly large.
They do.
But “all self driving cars” are practically only from waymo.
Level 4 Autonomy is the point at which it’s not required that a human can intercede at any moment, and as such has to be actively paying attention and be sober.
Tesla is not there yet.On the other hand, this is an active attack against the technology.
Mirrors or any super-absorber (possibly vantablack or similar) would fuck up LIDAR. Which is a good reason for diversifying the Sensors.On the other hand I can understand Tesla going “Humans use visible light only, in principle that has to be sufficient for a self driving car as well”, because, in principle I agree. In practice… well, while this seems much more click-bait than an actual issue for a self-driving taxi, diversifying your Input chain makes a lot of sense in my book. On the other hand, if it would cost me 20k more down the road, and Cameras would reach the same safety, I’d be a bit pissed.
On the other hand I can understand Tesla going “Humans use visible light only, in principle that has to be sufficient for a self driving car as well”, because, in principle I agree.
The whole idea is they should be safer than us at driving. It only takes fog (or a painted wall) to conclude that won’t be achieved with cameras only.
On the other hand, if it would cost me 20k more down the road, and Cameras would reach the same safety,
You had a lot of hands in this paragraph. 😀
I’m exceptionally doubtful that the related costs were anywhere near this number, and it’s inconceivable to me that cameras only could ever be as safe as having a variety of inputs.
Musk’s ethos is clear, both in business and government. He will make whatever short term decisions his greed and the ketamine tell him to make, and fuck whatever happens down the road. Let’s not work so hard to sanewash him like the media has Trump.
The whole idea is they should be safer than us at driving. It only takes fog (or a painted wall) to conclude that won’t be achieved with cameras only.
Well, I do still think that cameras could reach “superhuman” levels of safety.
(very dense) Fog makes the cameras useless, A self driving car would have to slow way down / shut itself off. If they are part of a variety of inputs they drop out as well, reducing the available information. How would you handle that then? If that would have to drop out/slow down as much,you gain nothing again/e: my original interpretation is obviously wrong, you get the additional information whenever the environment permits.
And for the painted wall. Cameras should be able to detect that. It’s just that Tesla presumably hasn’t implemented defenses against active attacks yet.You had a lot of hands in this paragraph. 😀
I like to keep spares on me.I’m exceptionally doubtful that the related costs were anywhere near this number.
cost has been developing rapidly. Pretty sure several years ago (about when tesla first started announcing to be ready in a year or two) it was in the tens of thousands. But you’re right, more current estimations seem to be more in the range of $500-2000 per unit, and 0-4 units per car.
it’s inconceivable to me that cameras only could ever be as safe as having a variety of inputs.
Well, diverse sensors always reduce the chance of confident misinterpretation.
But they also mean you can’t “do one thing, and do it well”, as now you have to do 2-4 things (camera, lidar, radar, sonar) well. If one were to get to the point where you have either one really good data-source, or four really shitty ones, it becomes conceivable to me.From what I remember there is distressingly little oversight for allowing self-driving-cars on the road, as long as the Company is willing to be on the hook for accidents.
vantablack Lidar won’t be effected by vantablack out side of a lab experiment. It picks up contamination very quickly and can’t be effectively cleaned.
Money.
they generally do
That’s why they called it the “Coyote” dataset.
They should just program it to drive through the painted tunnel but when another driver comes behind you they crash into it.
OMFG someone test to see if Teslas stop to eat free bird seed.
Meep meep!
So many Acme products we need to test, I saw one guy already successfully tested an Olmec statue.
Or a badly painted sign that says “Free charging”
Somebody with better animation skills than me make a cartoon where Wile E. Coyote is hunting cybertrucks using his old tricks and every single one of them works in his favor.
I’d buy that for a dollar!
+1…a classic!
Anyone with half a brain could tell you plain cameras is a non-starter. This is nearly a Juicero level blunder. Tesla is not a serious car company nor tech company. If markets were rational it would have been the end for Tesla.
If markets were rational, CEO compensation would never have grown so high, and there’d be no billionaires either.
Austin should just pull the permits until all the taxis have lidar installed and tested. Or write a bill that fines the manufacturer $100 billion for any self driving car that kills a person and puts the proceeds 50% to the family and 50% to infrastructure. One of the first rules of robotics was always about not harming humans.
One of the first rules of robotics was always about not harming humans.
Notably, roomba vacuum cleaners use cameras instead of lidar that other robot vacuums use. I bought a high end roomba a couple months ago and it was crap at navigating my home, while my old xiaomi with a lidar works perfectly fine. Needless to say i returned the roomba.
Very surprised Mark isn’t… Super supportive of musk and Tesla.
He owns a Tesla and is rather wealthy at this point. Not to mention that he’s Mormon. I’d expect him to be very conservative and all in on the grift.
What a world we’re living in!
Observing a technical deficiency in a robotics platform requires political considerations. Even when a car drives into a fucking wall at 40MPH on camera, people are asking about the camera man’s political party affiliation and not what’s wrong with the car.
Wild!
Unfortunately when the vehicle in question is created by a company owned by a man operating a government agency, it’s a valid question. He could have just never made the video, but making one that directly opposes the narrative of people you’d expect the “camera man’s” political affiliation to be seems unusual.
Mark is a smart guy, I’m sure he walks great big circles around anything political, at least publicly.
His audience is everybody, aligning publicly with any kind of political flow is generally a bad idea if you want that to stay that way, because the only thing you’ll likely achieve is shrinking your potential audience.
I would also be careful with the assumption that all conservatives agree with what’s currently happening.
I’d expect him to be very conservative
We still don’t know for sure. That video will likely become one of, if not his top-grossing videos. The topic and timeliness are absolute fire.
I give him some credit, though. It’s a dicey time to throw Musk under the self-driving bus while showing that alternatives don’t have the same problem.
Rober is definitely a businessman out to make money and is very self-promoting and will accept just about anybody as a sponsor, but I can’t think of anything he’s done that’s been out-and-out deceitful or political. And he really does have some engineering chops.
I think he’s a good voice for this b3cause he’s been so intentionally apolotical, and even my right-wing family likes his stuff.
Though my YouTube crazy engineer of choice is Stuff Made Here. He spends months between videos, but the stuff he makes is awesome, and he shows off a lot more of the actual creative process. And his fabrication tool collection is insane for a home shop.
I love that one of the largest YouTubers is the one that did this. Surely, somebody near our federal government will throw a hissy fit if he hears about this but Mark’s audience is ginormous
Honestly I think Mark should be more scared of Disney coming after him for mapping out their space mountain ride.
He probably just made Disney admissions and security even more annoying for everyone else.
Judging by the fact that he has an imagineer-video out (effectively) at the same time as the space-mountain mapping, I’d expect that Disney was fully aware of what he was doing, and the whole sneaky-thing was just to make it more appealing to viewers.
It got fucking wile e coyoted
The scientists in Ireland calling their data set to prevent this exact fucking thing “Coyote” sent me over the moon.
Clearly we don’t need lidar!
Direct video link https://www.youtube.com/watch?v=IQJL3htsDyQ
“But humans can do it with their eyes!” - says the man not selling a human brain to go with the optical sensors
“But humans can do it with their eyes!”
That’s the best part, they kinda can’t.
There are videos from before they pulled the sensors of some pretty cool stuff where teslas slammed the breaks before anything visibly happened, based on lidar sensors sensing trouble a couple cars up the road, completely blocked to vision.super cool safety tech, and then they pulled it…
one example here https://www.youtube.com/watch?v=BIcC2ZMePKI
Pretty sure that wasn’t even lidar. It was radar which is even cheaper and pretty much every other new car has if they don’t have lidar.
The thing is, yes humans can do it with their eyes. But even with the giant amount of progressing power from the brain they are still not great at it.
So of the ultimate goal is to the minimum/cheapest to be almost as good as human then yes, optical sensors only are enough.
Of the goal is to prevent deaths and significantly reduce the number of accidents compared to then lidar is the best option.
Very interesting!
What’s the payoff period, I wonder, assuming everyone could afford optical only before everyone could afford better tech.
Don’t give him ideas!!!
On the internet, nobody knows you’re just a brain in a
jarcar.Check out moneybags over here who can afford a jar car.
Please don’t vandalize their JarCar it has my mom’s brain tissue in it.
“But humans can do it with their eyes!”
The thing is, RADAR can see things humans can’t. There was a whole article a while back about a Model X that avoided an otherwise unavoidable accident by bouncing radar under the car in front of it and seeing that car slam on the brakes.
I will point out that if you (or your camera-only driver assist) can’t stop without hitting the car in front of you when they slam on the breaks, then you’re driving too close to them… You really shouldn’t ever put yourself in a position where the person in front of you could cause you to unavoidably hit them.
That said… Yeah, radar/lidar are far better than camera alone and there’s no good reason not to include them in the sensor suite unless you value profits over lives.
And I will point out that if the car in front of you isn’t paying attention and rams a stopped car in the middle of the road, you are fucked no matter what.
Not if you have the following distance to stop, but point taken: a crash decelerates you faster than breaks can and typical following distances are assuming breaking distance, not hard sudden halts.
So increase your following distance. It also has the benefit that it makes it easier to see what’s ahead of the car in front of you.
There’s pretty much no accident that’s unavoidable (barring someone else plowing into you) if you drive defensively enough (assuming good traction and good breaks, but obviously you should increase your following or decrease your speed to compensate for that as well)
Not if you have the following distance to stop
Maintaining a stopping distance like that is nigh impossible in a dense urban area. You’d be constantly cut off and causing tons of traffic.
Really? I do it pretty frequently without issue…
dude is living proof brains are optional
The day I heard that was the day I realized he’s a fucking idiot and I wanted nothing to do with his cars/tech.
Judging by how things have turned out…damn was that a good decision lmao
They pulled the RADAR from mine just before I took delivery, unbeknownst to me at the time. I received no sort of notification.
Absolutely the same timing for me. That was a big, oh that’s his logic?, moment.
I’m kinda confident that even RADAR + cameras was good enough, but they started shipping cars without it and even shutting off the RADAR in existing cars.
The main negative about LiDAR is the cost, but that’s quickly going down.
Why did they shut off radar?
They don’t want to install it in the newer cars, and they don’t want to make two software versions I guess. A backup would’ve been great though…
Thank you!
I tried watching it and it forces a horrible dubbing over it so I didn’t want to watch it. Apparently only way to chage it is to change my whole youtube account language
for the youtube website interface click on the gear wheel, and you can select the audiotrack you want
Yes, but not possible on mobile
I don’t know if it’s somehow not available to everyone, but I am able to change the audio track on mobile.
The rain test was far more concerning because it’s much more realistic of a scenario. Both a normal person and the lidar would’ve seen the kid and stopped, but the cameras and image processing just isn’t good enough to make out a person in the rain. That’s bad. The test portrays it as a person in the middle of a straight road, but I don’t see why the same thing wouldn’t happen at a crosswalk or other place where pedestrians are often in the path of a vehicle. If an autonomous system cannot make out pedestrians in the rain reliably, that alone should be enough to prevent these vehicles from being legal.
Who owns the White House right now?
The question there would be does Austin have crosswalks that don’t have red lights. Many places put a light at every cross walk, but not all. Most beaches don’t have them at every crosswalk, they just have laws that if someone is in or entering the crosswalk you have to stop for the pedestrians. (They would all be at risk from what you are saying).
Yes, there are mid-block crosswalks in some of the walkable parts of Austin. There are also roundabouts with yield signs and crosswalks and no lights.
That will cause huge issues possibly. Do you live near there? We need to get this information to the public in those areas. Even if it is raining. Do not cross without checking over and over. We need to ban them from being there, but we need to protect the people first. 1 life may overturn the law, but 1 life shouldn’t be lost. It’s better we figure out an alternative
I don’t know the answer to your question, but I’ll add that I’ve seen major cities that have overhead yellow flashing light boxes that mean “you must stop if there is a pedestrian crossing the road”
That should at least slow them down, but yeah it could be a real threat there as well.
Not every pedestrian follows the rules of the lights though. And not every pedestrian makes it across the road in time before the light changes colors from red to green.
I didn’t say anything about whether it was adequate. The fact is it is going live. Trying to find weak spots and dangerous areas and point them out to people is all we can do at this stage.
Meep meep.
This is like the crash on a San Francisco bridge that happened because of a Tesla that went into a tunnel and it wasn’t sure what to do since it went from bright daylight to darkness. In this case the Tesla just suddenly merged lanes and then immediately stopped and caused a multi car pile up.
You’d think they have cameras with higher dynamic range and faster auto exposure in their cars by now. Nope, still penny pinching.
If only elon hadn’t insisted on not using lidar or anything other than just visible light cameras
Yeah, pulling radar from the cars was the beginning of the end. Early teslas had radar, and that was what led to all of the “car sees something three vehicles ahead and brakes to avoid a pileup that hasn’t even started yet” type of collision avoidance videos. First, pulling radar was a cost cutting thing. Then Elon demanded that they pull out the lidar too, and that’s when their crash numbers skyrocketed.
They never had lidar, in addition to radar they removed the ultrasound sensors for parking, which is stupid because they cost like $2 and for parking they’re much better than cameras. Same for the rain sensor. Why use a $1 rain sensor that always works reliably all the time in any visibility when you can do that with cameras and complex algorithms?..
It’s been about 7 years of model 3 on the market, maybe 8, and the rain detection still doesn’t work reliably. Or the traffic sign recognition (in Europe). My car fortunately still has the ultrasound sensors. Phantom braking is still an issue, too. Thank God for stocks for blinkers and drive/reverse.
I like the car in general, but it has the dumbest fails, things everyone else seems to have figured out.
Other cars also have dumb mistakes, like electric cars with no frunk. Literally bolted down hoods. Looking at you, German auto industry…
They never had lidar, as far as I know.
The cameras that do what is needed are 10’s of thousands of dollars.
