I really like the look of the car, but from the title it sounds like a Mustang has been converted into an FSD Tesla ("teslafied" Mustang) - but Tesla suspension, Tesla interior... this smells like a Mustang body fitted onto a Tesla chassis ("mustangified" Tesla).
I suspect that this might be more of a "Mustang body kit" on a Tesla chassis and not retrofitting the Tesla tech into a Mustang chassis + body. Still cool, but maybe misleading.
Yep, that's exactly what it is. Still a cool project. For a split-second after reading the headline my brain thought they had gotten the Tesla software (with a lot of hackery) to control an ICE vehicle drive-train.
It does work in driving the car where it is able to. Where it fails is in the 'full' part. After 10 billion miles driven on 'auto-pilot' [1] it is hard to claim it 'does not work'. Tesla would have been better off removing the 'F' from 'FSD' but that's water under the bridge.
If we judge it as "Self Driving", it does the Driving part pretty well, and is quite bad at the Self part. There exist no roads and no weather conditions where I'm allowed to take my eyes off the road and stop being the safety nanny.
I liked this article's definition of Full Self Driving (Level 4 autonomy), it is very clear - when Tesla directly takes on the legal liability for unsupervised driving.
You just said "where it fails" and then state hard to claim it not working. If you call it Full Self Driving and it doesn't fully self drive, then it doesn't work. Not really sure where the confusion is. It's not water under the bridge. It is what it is. They claimed it would be fully self driving and not some lane/speed maintenance that pretty much all car makers can do now. It was straight up explained to drive the car. Any deviation from that means it is not working and people like you willingly accepting what Musk has lied about for years trying to make the rest of us out to be the weird ones for not falling for it. I'm tired of the gaslighting.
Literally every non-budget brand (and even some of the budget brands) offers automatic lane keeping with traffic-aware cruise control somewhere in their fleet. It might only be on their flagship vehicles, and possibly only on the top trims, but you're living in a Tesla-decorated cave if you think those are still Tesla-only features.
On a Tesla, it's not even an FSD-specific feature. Autopilot does it.
So many cars come with lane assist and adaptive cruise control. You can google those terms for yourself. I don't bother with lmgtfy requests. You're a big boy/girl, and teaching you to fish it much better effort. They also don't cost an additional $10k on top of the price of the car. They are just part of the price of the car.
Maybe the lmgtfy would be a good exercise for you. Ford's BlueCruise, GM's SuperCruise, Rivian's Autonomy+, and Mercedes-Benz’s Drive Pilot all cost money.
Indeed and none of those work outside of (select) highways. Lucid also has DreamDrive, but it's fairly poor from what I've seen. BYD's God's Eye is in the news, and it isn't looking good either.
I'd love to see good competition in this space, but it seems Tesla has a healthy moat.
Full is a relative term. Full compared to what? Compared to a professional rally car driver? Compared to my grandmother? Compared to a properly licensed tourist in a foreign country?
From videos I see on YouTube, I’m struggling to think what is not Full compared to—at a bare minimum—the bottom 10 percent of drivers on the road.
Tesla set their own benchmark, their own goal posts, and their own timelines.
In 2016 Tesla said, "as of today, all Tesla vehicles produced in our factory – including Model 3 – will have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver.":
"Full" in full self-driving is a superfluous modifier. But it does is further emphasize that, what a person would consider "self-driving", it can do. Except it can't, of course.
> The team grafted three sections of the 2024 Tesla Model 3’s floor and seats into the Mustang’s body, shortening the battery case to fit without altering the car’s original dimensions. The result is a classic Mustang shell sitting on top of a Model 3 dual-motor setup
I used to work at a company that did self driving. The sensor setup was more complicated than Tesla (cameras, lidar, etc), but the fact that FSD can still work on this car despite the cameras being in a different place is really impressive to me. Our sensors were pretty sensitive to accurate calibration, and iirc any time we tried to move our sensor array to a new car it took a ton of work to reconfigure it to make the sensor fusion output work.
Most sensors can be implemented in a way that enables self-calibration.
I'm oversimplifying it here, but the macro process is taking some known attributes and mapping them to what you are observing. For example, if you can detect people, and you know the average height of a person, you can compute where your horizon is, and where you should (or shouldn't) expect to see people in the FOV. You can do this with cameras, lidar, etc. When you have multiple sensors you can do a lot more to have them all sample an object in their own ways and converge on agreement of where they are relative to each other and the object.
I’m not sure this has much to do with vision as opposed to fancy self-calibration software. At least a few years ago, Tesla cars would be in self-calibration mode for a while after delivery while they calibrated their cameras. I think the idea is that it’s cheaper to figure out in software where everything is than to calibrate the camera mounts and lenses at the factory.
I see no reason that LiDAR couldn’t participate in a similar algorithm.
A bigger issue would be knowing the shape of the car to avoid clipping an obstacle.
It probably could, but I imagine a LIDAR system would need a similar (large) amount of training data to enable effective self-calibration across a wide variety of situations.
At some point, with enough sensor suites, we might be able to generalize better and have effective lower(?)-shot training for self-calibration of sensor suites.
Isn’t the model needed rather similar to what’s needed for sensor fusion in general? If you can extract features from each sensor that you expect to match to features from a different sensor, then you can collect a bunch of samples of this sort of data and then use it to fit the transformation between one sensor’s world space and another sensor’s world space.
There is actually a photo of a Model 3 in factory line doing that precise positioned target stuff, so they do that precise positioned target stuff at the factory. It's also instant compared to a 10-minute drive.
I think the real reason why Tesla is known to require 10-minute calibration drive is, they shipped APHW2 long before the software matured, so they needed means to do it after the cars were shipped "blank". Other manufacturers only ship finalized hardware and software, and so they don't need a scalable tool-free calibration method.
Anyways, my point is that, Tesla cars need calibrations like anything else. This is same for any multi sensor SLAM systems, whether it uses sets of color cameras or laser spinny thingy or laser flash cameras or laser flash color camera thingy or combinations thereof.
From the article it sounds like the inverse, they took a Tesla and stuck a classic car exterior shell on it, not transplant the electric car parts into a mustang frame. It is still kind of neat but is not the same thing to me. You don't normally upgrade a classic car by chopping out the entire frame and sticking the body panels onto a modern car.
I'd love to see more of this as an option... Getting a modern electic car with a late 60's Camero styled shell would be a nice option IMO. Would need to take a few liberties with some dimensions in the designs as the originals won't 1:1 match up, but there's probably enough leeway to make such things work out.
The dimensional differences of modern cars are striking. You'd need to make a 6:5 larger scale model of the original car in many cases. Or worse, if you've ever seen a classic Mini next to a modern Mini.
(yes, I will admit that a lot of that is for crash safety, but not all of it)
I agree fully. Hyundai has a mockup that starts to get there (different era, but same concept) called the N vision 74[1], but I doubt we'll see it in market anytime soon. The unfortunate reality IIUC is that modern cars (electric vehicles) have certain aero restrictions (for mileage) that heavily limit design options.
Aero is driving a lot of EV design choices yes but it’s not as much of a limitation as you are implying - see the Ioniq 5, Cybertruck, F150 lightning or any of Rivian’s vehicles for examples where they’ve traded a bit of range for a boxier, less aero shape (admittedly often with _some_ rounded corners still).
It’s a trade off most manufacturers are not making because the US market is _so_ range conscious but I think it is fairly small margins we’re talking.
Here's one of a fully custom Toyota 4x4 truck getting a Tesla Model 3 motor that I enjoyed. I would love to have a small electric pickup like this, but I don't want to invest $100k to get it done
It would be cool if we saw the separation of drivetrain from body in the automobile market. This happens with heavy trucks, you can buy a "glider" which is a completely new, finished rolling chassis and you provide your own engine. Originally done (I think) to skirt emissions laws but it would be cool to be able to buy the body and the EV drivetrain (and maybe battery packs?) from different vendors, and for EV drivetrains to be more easily fittable to older chassis.
This was very common in the early days of the automobile, at least for luxury cars. Bentley or Rolls-Royce would deliver a chassis with the entire drivetrain, and a coachbuilder like Mulliner would add a body to the customer's liking.
Take a look at Kia's 'PBV' concept, and the new PV5 van. It's a skateboard chassis with all the EV gubbins in, with a body and cabin that bolts on. Allows for extremely modular vehicle construction.
One cool thing about commercial vehicles is that the CAN bus is all open-spec and fully documented. This is because fleet operators expect to be able to put their own aftermarket components on the bus and manufacturers have to support that.
The team sat Electric Classic Cars are doing that, you can soon buy a skateboard chassis and drivetrain from them, and bolt whatever body you want to it.
that's what Waymo is doing, in their implementation of this idea the "waymo driver" is software intended to be licensed out to vehicles or vehicle chassis with applicable hardware
so if you thought the waymo car rollout was fast and sudden, wait until companies no longer need their own training data, it'll be like a switch got flipped
Well, when there's a software created accident, possibly horrendous, I'm sure Tesla will be relieved that it's one case where they can point to it being a completely unauthorized system for which they reject responsibility.
This sort of conversion gets coverage every once in a while and it's been neat to see old frames getting chopped onto new electric drivetrains. I spoke with one of the people interviewed in this article[1] a couple years back about converting an old truck I have sitting around into an EV.
The Model 3 approach takes their unified rear axle (motor,axle,wheels) and mounts it into an existing frame. Then you just need to find a place to stuff the batteries, retrofit some high-voltage electronics, and you're off to the races. One of the drawbacks of that approach is that it changes the stance of the vehicle, but for this Mustang that doesn't seem to matter much - it still looks classic.
Other converters either go for the high end with a model S and fit the motor into a traditional drivetrain for a sleeper build, or they go for the low end and take an old forklift motor and batteries and build what is effectively a street-legal golf cart. Prices range from $5-100k depending on your level of DIY and how dangerous of a classic car you want on the other side of the process.
> It’s likely the first non-Tesla vehicle to run FSD, and it achieves 258 Wh/mi — roughly matching the efficiency of an actual Model 3.
This claim is implausible, right? The Mustang is unambiguously less aerodynamic than the Model 3; there's no way it is achieving similar efficiency, especially at highway speeds.
The Mustang is from before modern safety laws (and feature expectations) and therefore weighed a lot less than your average modern car.
A stock '66 Mustang hardtop had a curb weight below 3000lb, in the lightest configuration close to 2500lb.
Less mass to move will do a lot for efficiency just like aerodynamics will.
Of course, you will also die or be horrifically maimed in an accident in a 1966 Mustang that you might walk away without any serious injuries from in a modern vehicle.
But this conversion is basically just a Tesla model 3 with the shell taken off and a mustang shell installed over. It's mostly a model 3, including the heavy battery and drivetrain. And airbags.
That's really cool, though I confess I would have preferred the interior to have been more Mustang and less Model 3. Just a quibble, though, the effort is fantastic.
A lot of people use Leaf batteries/powertrains for this type of conversion. Probably would be cheaper than using Tesla parts if you don't need as much range.
Neat. I would have preferred the original interior over Tesla's, but I guess it would then just be an electric conversion and not a "Tesla" conversion with "FSD".
That Vagabond Builds video... They edited it, but left in a claim it had a frunk. The commentary felt like engagement optimized stream of consciousness blather. They cut out shots of the car being moved and never showed it being driven.
Whatever happened to the electric delorean reboot?
EDIT: at one point whoever owned the name also owned a warehouse of spare parts and was going to produce an electric retrofit kit for the old vehicle, and hinting at manufacturing new ones a la retromod. Whoever owns the name now just has concept rendering on their site and a Solana token, so, little more than a meme coin now :(
I believe Stephen Wynne, the man you're referring to, still owns the new DMC company, and he did even produced a concept car of his Alpha5 model. But since 2022, there hasn't been much activity. His classic DMC parts sales side of the business is still active though.
Retro-electric stuff makes so little sense since it's the worst of all worlds. Part of why Teslas get decent range is the slippery body. I wince every time I see people clamoring for the VW Scout reboot. Rivian too with their 140kwh batteries just to give people that nostalgic body-on-frame SUV look with usable range.
You don't take on projects like this because you're on a mission to make the most efficient vehicle. Lots of people are paid good money to do that and produce the slip-slugs we have today.
A project like this is to have a fun experience in a vehicle that was never designed to drive with electric qualities. I don't need the most efficient vehicle for my use so I could afford to trade some of that for fun. I'd probably try a Subaru Impreza STi because it would just be a blast to have a car of that size and stature with an electric powerplant under the hood (or trunk, or wherever it fits)
Mainly size and driving position. The charm of the AWD and live adjustments would need to be kept around but even without the ol turbo you'll still have a small, nimble, darty car to toss around. Something would be lost in the conversion but balancing turbo lag isn't the entire car. The Model 3 is long, wide, with a big wheelbase compared to, especially the older, STi's.
I wouldn't call STIs darty. They're nose heavy and need heavy coaxing with the left foot to rotate. The big wheelbase in the Model 3 is a fair point though. Apparently there are TC/DSC defeat devices and mechanical LSDs available for them now, so I'd expect Model 3s to be rallied more and more as they age out.
It makes sense just like 22 people kicking a ball around a field makes sense. I repair electronics as hobby. This past week I repaired my daughters $15 toy tablet. It took me a few hours to do that. Economically completely unjustifiable. This is probably the only way to look at this project.
Correct. Not sensible at all. Except so much more sensible than retro-petroleum stuff.
Watching my brother-in-law buy a 1971 Chevelle for his 16-year old daughter because she thought it looked cool only to have him sell it at a fat loss 3 months later because she couldn't choke down the gasoline fumes driving out of the school parking lot every day was instructive.
Or tune the engine correctly. Probably has an off-the-shelf "performance" carb that's set much richer than it should and a "full race" cam that only makes sense for a track car, giving horrible fuel economy and actually less low-end power.
My daily driver is roughly as old, has a 400 V8 with a 4-barrel, idles so quietly I've had passengers surprised that the engine was running, and gets around 20-25mpg if I resist the urge to open it up all the way.
> Retro-electric stuff makes so little sense since it's the worst of all worlds.
old cars are bastards to drive. I have a softspot for a mark 2 VW golf. But its not fast, the steering is heavy and the brakes are utterly shite.
However, if I had the time and money, I would totally electrify a golf. it would be zippy quiet and hilarious to drive, especially without any kind of traction control.
However it would be fun.
Basically its like vinyl. It is a demonstrably worse format than anything digital(and other analogue formats), however it looks great. Sure you get lots of audiophiles waffle on about "warmth" and shit, but its all lies. they either like it because its how they think things should sound, or it looks cool. It is not a purer warmer sound.
same with backyard steam engines. useless but fucking cool
Wouldn't it be easier to get a Mk2 GTI and put some decent tires, shocks and pads on it? Steering being heavy just means you have to avoid steering at a standstill which can be fun in itself since you're actively working to be more fluid.
> Wouldn't it be easier to get a Mk2 GTI and put some decent tires, shocks and pads on it?
oh very much so, it would be much easier to do that way, cheaper too.
But I don't think you do this for the ease of it, you do it either for the challenge, or to overcome some blocker (like parts shortage, or the engine is knackered.
> means you have to avoid steering at a standstill
ha! yeah, I still do the creepy and turn, even with the modern cars that I drive. I also still have a strong clutch reflex when driving automatic/electric
Vinyl, for whatever continuing reason, often does sound better than digital formats but only because it’s mastered with more care.
It could be that it’s physically impossible to master vinyl for extreme loudness, but whatever the reason is you can absolutely pick up a vinyl copy of an album and find it sounds much better than the streamed or CD version.
EVs are still far more efficient than any combustion engine per unite of energy. It doesn't have the density of gasoline, but you're paying a lot less per mile driven no matter how boxy it is.
I've been waiting for someone to do something like this as long as I've known electric cars to be a thing. I hope they just start making them like this.
That would be a pretty shortsighted move from a PR standpoint. Nobody is going to blame a restomod FSD wreck on Tesla. But turning off FSD remotely just reminds everyone that Tesla can and will take control of your car.
I took my 1972 Dodge small block engine out and converted it to 400 hp. Had to upgrade the transmission, driveline, brakes, radiator, and suspension to match. I self-drive it.
You are entitled to your opinion, obviously, as is anyone else. Restomods have been around since forever, and this is hardly the first EV conversion (almost certainly not the first classic Mustang EV restomod either).
Personally I think it's pretty damn cool. But I have always been a Mustang fan, and I know that this era of Mustang is not especially collectable. They made quite a large number of them and plenty are still running.
To be fair though, this is beyond typical restomod territory. This isn’t an LS swap with modern wheels and suspension. This most certainly changes the spirit of the car.
There's no question the first generation of Mustangs are the most collectible.
A 66 mustang ... without the sound? A lion without a roar.
And at these prices, you would reduce far more carbon by investing that money in solar panels. 50k buys at least a 20kw system that would more than make up for your summer weekend drives in a classic mustang.
I know this is not the way the industry or regulations work, but I wish electric car platforms let you pick body styles without waiting for a whole model to come out. I'd love an electric Suzuki Jimny body, and could care less about the driving platform.
I have a dream that electric motors + battery get so compact that you can make kits that fit engine blocks and large numbers of cars can be custom swapped relatively cheaply.
There was a shop in Dallas back about 20 years ago that did an electric conversion of a H1 Humvee. Since then there’s been lots more conversions like that and to me that is a valid recycling business.
> It demonstrates that Tesla’s hardware and software stack is more portable than the company’s licensing struggles would suggest.
Unless I missed something, this is a completely unsupported claim by the article. Passion projects and retrofits are nothing at all like manufacturing.
The article claims that the whole project only cost $40,000, and then compares that to electric conversion offerings that cost $75,000 (and mentions that the global conversion market in 2024 was $5.9 billion). I think the implication is that there could be a large market for FSD conversions that goes beyond passion projects because it is not only possible but affordable.
I would be surprised however if this project only cost $40,000, when you factor in the cost of labor and maintaining a facility to do this work.
It's specifically talking about the "FSD" model under the hood being able to run on this retrofit even though the cameras don't align 100% like they originally would.
It is basically a Mustang body on a tesla chasis... which misses the point of having a classic car.
While there is nothing wrong with converting your classic car to electric, if the powertrain is shot (they are harder to maintain as they age), but IMO, it looses the charm of the point of having a classical car.
Few years ago, there was a trend to do these conversions, but that stopped as people realised the car loses its charm and the feel of having old classic car, and most of them are not being used as dailies anyways.
I really like the look of the car, but from the title it sounds like a Mustang has been converted into an FSD Tesla ("teslafied" Mustang) - but Tesla suspension, Tesla interior... this smells like a Mustang body fitted onto a Tesla chassis ("mustangified" Tesla).
I suspect that this might be more of a "Mustang body kit" on a Tesla chassis and not retrofitting the Tesla tech into a Mustang chassis + body. Still cool, but maybe misleading.
Yep, that's exactly what it is. Still a cool project. For a split-second after reading the headline my brain thought they had gotten the Tesla software (with a lot of hackery) to control an ICE vehicle drive-train.
For a split-second after reading the headline, I thought they were claiming FSD works.
It does work in driving the car where it is able to. Where it fails is in the 'full' part. After 10 billion miles driven on 'auto-pilot' [1] it is hard to claim it 'does not work'. Tesla would have been better off removing the 'F' from 'FSD' but that's water under the bridge.
[1] https://electrek.co/2026/05/03/tesla-fsd-10-billion-miles-no...
If they renamed it Featureless Sometimes Driving we might be onto something.
If we judge it as "Self Driving", it does the Driving part pretty well, and is quite bad at the Self part. There exist no roads and no weather conditions where I'm allowed to take my eyes off the road and stop being the safety nanny.
I liked this article's definition of Full Self Driving (Level 4 autonomy), it is very clear - when Tesla directly takes on the legal liability for unsupervised driving.
You just said "where it fails" and then state hard to claim it not working. If you call it Full Self Driving and it doesn't fully self drive, then it doesn't work. Not really sure where the confusion is. It's not water under the bridge. It is what it is. They claimed it would be fully self driving and not some lane/speed maintenance that pretty much all car makers can do now. It was straight up explained to drive the car. Any deviation from that means it is not working and people like you willingly accepting what Musk has lied about for years trying to make the rest of us out to be the weird ones for not falling for it. I'm tired of the gaslighting.
> lane/speed maintenance that pretty much all car makers can do now
What car can I buy in the US today that's as good as the latest fsd?
Literally every non-budget brand (and even some of the budget brands) offers automatic lane keeping with traffic-aware cruise control somewhere in their fleet. It might only be on their flagship vehicles, and possibly only on the top trims, but you're living in a Tesla-decorated cave if you think those are still Tesla-only features.
On a Tesla, it's not even an FSD-specific feature. Autopilot does it.
FSD isn't basic ADAS. You could just say no.
So many cars come with lane assist and adaptive cruise control. You can google those terms for yourself. I don't bother with lmgtfy requests. You're a big boy/girl, and teaching you to fish it much better effort. They also don't cost an additional $10k on top of the price of the car. They are just part of the price of the car.
Maybe the lmgtfy would be a good exercise for you. Ford's BlueCruise, GM's SuperCruise, Rivian's Autonomy+, and Mercedes-Benz’s Drive Pilot all cost money.
Indeed and none of those work outside of (select) highways. Lucid also has DreamDrive, but it's fairly poor from what I've seen. BYD's God's Eye is in the news, and it isn't looking good either.
I'd love to see good competition in this space, but it seems Tesla has a healthy moat.
Maybe re-reading what I wrote would help you lmgtfy.
The did, kind of. Instead of removing "Full", they added a disclaimer to the end.
https://www.caranddriver.com/news/a70420085/tesla-drops-auto...
even my 95 miata drives itself on a straight flat road...
but unlike the Tesla, the Miata is a car designed for the delight of the driver, rather than as a futuristic driving appliance / infotainment centre.
if they replaced F with S (Somewhat) all would be swell (hard to make any sales when not lying though…)
Full is a relative term. Full compared to what? Compared to a professional rally car driver? Compared to my grandmother? Compared to a properly licensed tourist in a foreign country?
From videos I see on YouTube, I’m struggling to think what is not Full compared to—at a bare minimum—the bottom 10 percent of drivers on the road.
> From videos I see on YouTube, I’m struggling to think what is not Full compared to—at a bare minimum—the bottom 10 percent of drivers on the road.
Try sitting in the back seat or even just acting like a passenger and you'll see the difference very quickly.
For me it has a very specific meaning: "Full" means "Unsupervised and without a geofence". Anything less is not Full Self Driving.
Nobody admits they're in the bottom 10%. Nobody even admits they're in the bottom 75%.
> Full compared to what?
Tesla set their own benchmark, their own goal posts, and their own timelines.
In 2016 Tesla said, "as of today, all Tesla vehicles produced in our factory – including Model 3 – will have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver.":
https://electrek.co/2024/08/24/tesla-deletes-its-blog-post-s...
https://web.archive.org/web/20240730071548/https://tesla.com...
That was, of course, a lie. Tesla has spent the last 10 years lying about the state of FSD. Tesla keeps claiming FSD will be achieved "next year".
What about 1 million robotaxis on the road by 2020: https://www.thedrive.com/news/38129/elon-musk-promised-1-mil...
More lies: https://en.wikipedia.org/wiki/List_of_predictions_for_autono...
But you knew all that already. Defending a decade's worth of lies is intellectually dishonest.
"Full" in full self-driving is a superfluous modifier. But it does is further emphasize that, what a person would consider "self-driving", it can do. Except it can't, of course.
I have FSD in a 2026 Model Y and it does a solid job for me.
It'll work next year, don't worry.
It works, it just doesn’t fully drive by itself.
I was 100% FSD today. Went to the office, ran some errands, went home. Never touched the steering wheel once. 2026 Model S Plaid.
> Never touched the steering wheel once.
Don’t you have to?
> The team grafted three sections of the 2024 Tesla Model 3’s floor and seats into the Mustang’s body, shortening the battery case to fit without altering the car’s original dimensions. The result is a classic Mustang shell sitting on top of a Model 3 dual-motor setup
how some people gobble up Tesla FSD - when it's clearly inferior to options out there from Waymo, GM Supercruise or hell even Ford.
I used to work at a company that did self driving. The sensor setup was more complicated than Tesla (cameras, lidar, etc), but the fact that FSD can still work on this car despite the cameras being in a different place is really impressive to me. Our sensors were pretty sensitive to accurate calibration, and iirc any time we tried to move our sensor array to a new car it took a ton of work to reconfigure it to make the sensor fusion output work.
This is one of the real advantages to the (often insulted and/or chastised) vision only approach to FSD.
People can easily adapt to different vehicles in a similar manner.
Most sensors can be implemented in a way that enables self-calibration.
I'm oversimplifying it here, but the macro process is taking some known attributes and mapping them to what you are observing. For example, if you can detect people, and you know the average height of a person, you can compute where your horizon is, and where you should (or shouldn't) expect to see people in the FOV. You can do this with cameras, lidar, etc. When you have multiple sensors you can do a lot more to have them all sample an object in their own ways and converge on agreement of where they are relative to each other and the object.
I’m not sure this has much to do with vision as opposed to fancy self-calibration software. At least a few years ago, Tesla cars would be in self-calibration mode for a while after delivery while they calibrated their cameras. I think the idea is that it’s cheaper to figure out in software where everything is than to calibrate the camera mounts and lenses at the factory.
I see no reason that LiDAR couldn’t participate in a similar algorithm.
A bigger issue would be knowing the shape of the car to avoid clipping an obstacle.
It probably could, but I imagine a LIDAR system would need a similar (large) amount of training data to enable effective self-calibration across a wide variety of situations.
At some point, with enough sensor suites, we might be able to generalize better and have effective lower(?)-shot training for self-calibration of sensor suites.
Isn’t the model needed rather similar to what’s needed for sensor fusion in general? If you can extract features from each sensor that you expect to match to features from a different sensor, then you can collect a bunch of samples of this sort of data and then use it to fit the transformation between one sensor’s world space and another sensor’s world space.
Teslas require a camera calibration after windshield replacements, same as any cars
The calibration is a 10 minute drive on any marked road though, not the precise positioned target stuff that others use.
There is actually a photo of a Model 3 in factory line doing that precise positioned target stuff, so they do that precise positioned target stuff at the factory. It's also instant compared to a 10-minute drive.
I think the real reason why Tesla is known to require 10-minute calibration drive is, they shipped APHW2 long before the software matured, so they needed means to do it after the cars were shipped "blank". Other manufacturers only ship finalized hardware and software, and so they don't need a scalable tool-free calibration method.
Anyways, my point is that, Tesla cars need calibrations like anything else. This is same for any multi sensor SLAM systems, whether it uses sets of color cameras or laser spinny thingy or laser flash cameras or laser flash color camera thingy or combinations thereof.
From the article it sounds like the inverse, they took a Tesla and stuck a classic car exterior shell on it, not transplant the electric car parts into a mustang frame. It is still kind of neat but is not the same thing to me. You don't normally upgrade a classic car by chopping out the entire frame and sticking the body panels onto a modern car.
I'd love to see more of this as an option... Getting a modern electic car with a late 60's Camero styled shell would be a nice option IMO. Would need to take a few liberties with some dimensions in the designs as the originals won't 1:1 match up, but there's probably enough leeway to make such things work out.
The dimensional differences of modern cars are striking. You'd need to make a 6:5 larger scale model of the original car in many cases. Or worse, if you've ever seen a classic Mini next to a modern Mini.
(yes, I will admit that a lot of that is for crash safety, but not all of it)
I agree fully. Hyundai has a mockup that starts to get there (different era, but same concept) called the N vision 74[1], but I doubt we'll see it in market anytime soon. The unfortunate reality IIUC is that modern cars (electric vehicles) have certain aero restrictions (for mileage) that heavily limit design options.
[1] https://www.hyundai-n.com/en/models/rolling-lab/n-vision-74
Aero is driving a lot of EV design choices yes but it’s not as much of a limitation as you are implying - see the Ioniq 5, Cybertruck, F150 lightning or any of Rivian’s vehicles for examples where they’ve traded a bit of range for a boxier, less aero shape (admittedly often with _some_ rounded corners still).
It’s a trade off most manufacturers are not making because the US market is _so_ range conscious but I think it is fairly small margins we’re talking.
Here's one of a fully custom Toyota 4x4 truck getting a Tesla Model 3 motor that I enjoyed. I would love to have a small electric pickup like this, but I don't want to invest $100k to get it done
https://www.youtube.com/watch?v=siEhd4Z-6Ts
I check multiple times a year for a small electric truck. Plenty announced but none you can actually buy year after year
Hoping the Slate makes it... for that matter, would be nice if the Scout does as well.
Another candidate that I hope isn't vaporware: https://www.telotrucks.com/
It would be cool if we saw the separation of drivetrain from body in the automobile market. This happens with heavy trucks, you can buy a "glider" which is a completely new, finished rolling chassis and you provide your own engine. Originally done (I think) to skirt emissions laws but it would be cool to be able to buy the body and the EV drivetrain (and maybe battery packs?) from different vendors, and for EV drivetrains to be more easily fittable to older chassis.
This was very common in the early days of the automobile, at least for luxury cars. Bentley or Rolls-Royce would deliver a chassis with the entire drivetrain, and a coachbuilder like Mulliner would add a body to the customer's liking.
Take a look at Kia's 'PBV' concept, and the new PV5 van. It's a skateboard chassis with all the EV gubbins in, with a body and cabin that bolts on. Allows for extremely modular vehicle construction.
One cool thing about commercial vehicles is that the CAN bus is all open-spec and fully documented. This is because fleet operators expect to be able to put their own aftermarket components on the bus and manufacturers have to support that.
The team sat Electric Classic Cars are doing that, you can soon buy a skateboard chassis and drivetrain from them, and bolt whatever body you want to it.
Their recent videos showcase what they're doing in that area https://www.youtube.com/@ElectricClassicCars/videos
that's what Waymo is doing, in their implementation of this idea the "waymo driver" is software intended to be licensed out to vehicles or vehicle chassis with applicable hardware
so if you thought the waymo car rollout was fast and sudden, wait until companies no longer need their own training data, it'll be like a switch got flipped
Tesla has also offered their FSD to other companies, with no takers: https://www.businessinsider.com/elon-musk-tesla-fsd-licensin...
However, Tesla hasn't achieved anywhere near the autonomy of Waymo, so that may be the main sticking point.
nobody trusts non-lidar except tesla buyers
Superfastmatt did something similar with a very very old Jag... (Sans the FSD)
https://youtube.com/playlist?list=PLoTU9_iCGa6i_C38pwQyg0pBG...
His channel comes highly recommended for anyone even remotely interested in cars by the way.
Second this. I love his dry humour.
Well, when there's a software created accident, possibly horrendous, I'm sure Tesla will be relieved that it's one case where they can point to it being a completely unauthorized system for which they reject responsibility.
This sort of conversion gets coverage every once in a while and it's been neat to see old frames getting chopped onto new electric drivetrains. I spoke with one of the people interviewed in this article[1] a couple years back about converting an old truck I have sitting around into an EV.
The Model 3 approach takes their unified rear axle (motor,axle,wheels) and mounts it into an existing frame. Then you just need to find a place to stuff the batteries, retrofit some high-voltage electronics, and you're off to the races. One of the drawbacks of that approach is that it changes the stance of the vehicle, but for this Mustang that doesn't seem to matter much - it still looks classic.
Other converters either go for the high end with a model S and fit the motor into a traditional drivetrain for a sleeper build, or they go for the low end and take an old forklift motor and batteries and build what is effectively a street-legal golf cart. Prices range from $5-100k depending on your level of DIY and how dangerous of a classic car you want on the other side of the process.
[1] https://coloradosun.com/2023/06/25/classic-cars-electric-veh...
> It’s likely the first non-Tesla vehicle to run FSD, and it achieves 258 Wh/mi — roughly matching the efficiency of an actual Model 3.
This claim is implausible, right? The Mustang is unambiguously less aerodynamic than the Model 3; there's no way it is achieving similar efficiency, especially at highway speeds.
The Mustang is from before modern safety laws (and feature expectations) and therefore weighed a lot less than your average modern car.
A stock '66 Mustang hardtop had a curb weight below 3000lb, in the lightest configuration close to 2500lb.
Less mass to move will do a lot for efficiency just like aerodynamics will.
Of course, you will also die or be horrifically maimed in an accident in a 1966 Mustang that you might walk away without any serious injuries from in a modern vehicle.
But this conversion is basically just a Tesla model 3 with the shell taken off and a mustang shell installed over. It's mostly a model 3, including the heavy battery and drivetrain. And airbags.
Yes it is implausible. But I imagine that they are probably not running it at 70 mph everywhere, just pootling about in suburbia
That's really cool, though I confess I would have preferred the interior to have been more Mustang and less Model 3. Just a quibble, though, the effort is fantastic.
I’d love to do something similar to an El Camino. I don’t even need triple digit range; I’d use it as a local runabout, mostly to my art studio.
A lot of people use Leaf batteries/powertrains for this type of conversion. Probably would be cheaper than using Tesla parts if you don't need as much range.
If I ever find myself with the money to do this, that was exactly what I was thinking.
Neat. I would have preferred the original interior over Tesla's, but I guess it would then just be an electric conversion and not a "Tesla" conversion with "FSD".
That Vagabond Builds video... They edited it, but left in a claim it had a frunk. The commentary felt like engagement optimized stream of consciousness blather. They cut out shots of the car being moved and never showed it being driven.
Whatever happened to the electric delorean reboot?
EDIT: at one point whoever owned the name also owned a warehouse of spare parts and was going to produce an electric retrofit kit for the old vehicle, and hinting at manufacturing new ones a la retromod. Whoever owns the name now just has concept rendering on their site and a Solana token, so, little more than a meme coin now :(
Carson did it: https://www.youtube.com/watch?v=-eCvUGx6OkE&list=PLYIgpKT34L...
I believe Stephen Wynne, the man you're referring to, still owns the new DMC company, and he did even produced a concept car of his Alpha5 model. But since 2022, there hasn't been much activity. His classic DMC parts sales side of the business is still active though.
Retro-electric stuff makes so little sense since it's the worst of all worlds. Part of why Teslas get decent range is the slippery body. I wince every time I see people clamoring for the VW Scout reboot. Rivian too with their 140kwh batteries just to give people that nostalgic body-on-frame SUV look with usable range.
You don't take on projects like this because you're on a mission to make the most efficient vehicle. Lots of people are paid good money to do that and produce the slip-slugs we have today.
A project like this is to have a fun experience in a vehicle that was never designed to drive with electric qualities. I don't need the most efficient vehicle for my use so I could afford to trade some of that for fun. I'd probably try a Subaru Impreza STi because it would just be a blast to have a car of that size and stature with an electric powerplant under the hood (or trunk, or wherever it fits)
90% of the STi's charm is the drivetrain. How would a hacked up STi differ from a model 3?
Mainly size and driving position. The charm of the AWD and live adjustments would need to be kept around but even without the ol turbo you'll still have a small, nimble, darty car to toss around. Something would be lost in the conversion but balancing turbo lag isn't the entire car. The Model 3 is long, wide, with a big wheelbase compared to, especially the older, STi's.
I wouldn't call STIs darty. They're nose heavy and need heavy coaxing with the left foot to rotate. The big wheelbase in the Model 3 is a fair point though. Apparently there are TC/DSC defeat devices and mechanical LSDs available for them now, so I'd expect Model 3s to be rallied more and more as they age out.
> slip-slugs
That's a good one. I'm partial to David Frieberger's "mobility blobs".
It makes sense just like 22 people kicking a ball around a field makes sense. I repair electronics as hobby. This past week I repaired my daughters $15 toy tablet. It took me a few hours to do that. Economically completely unjustifiable. This is probably the only way to look at this project.
Correct. Not sensible at all. Except so much more sensible than retro-petroleum stuff.
Watching my brother-in-law buy a 1971 Chevelle for his 16-year old daughter because she thought it looked cool only to have him sell it at a fat loss 3 months later because she couldn't choke down the gasoline fumes driving out of the school parking lot every day was instructive.
Sounds like the dude needs to learn how to fix exhaust leaks.
Or tune the engine correctly. Probably has an off-the-shelf "performance" carb that's set much richer than it should and a "full race" cam that only makes sense for a track car, giving horrible fuel economy and actually less low-end power.
My daily driver is roughly as old, has a 400 V8 with a 4-barrel, idles so quietly I've had passengers surprised that the engine was running, and gets around 20-25mpg if I resist the urge to open it up all the way.
From the article:
> it achieves 258 Wh/mi — roughly matching the efficiency of an actual Model 3.
> Retro-electric stuff makes so little sense since it's the worst of all worlds.
old cars are bastards to drive. I have a softspot for a mark 2 VW golf. But its not fast, the steering is heavy and the brakes are utterly shite.
However, if I had the time and money, I would totally electrify a golf. it would be zippy quiet and hilarious to drive, especially without any kind of traction control.
However it would be fun.
Basically its like vinyl. It is a demonstrably worse format than anything digital(and other analogue formats), however it looks great. Sure you get lots of audiophiles waffle on about "warmth" and shit, but its all lies. they either like it because its how they think things should sound, or it looks cool. It is not a purer warmer sound.
same with backyard steam engines. useless but fucking cool
Wouldn't it be easier to get a Mk2 GTI and put some decent tires, shocks and pads on it? Steering being heavy just means you have to avoid steering at a standstill which can be fun in itself since you're actively working to be more fluid.
> Wouldn't it be easier to get a Mk2 GTI and put some decent tires, shocks and pads on it?
oh very much so, it would be much easier to do that way, cheaper too.
But I don't think you do this for the ease of it, you do it either for the challenge, or to overcome some blocker (like parts shortage, or the engine is knackered.
> means you have to avoid steering at a standstill
ha! yeah, I still do the creepy and turn, even with the modern cars that I drive. I also still have a strong clutch reflex when driving automatic/electric
Yeah, it would also be easier to just buy an e-golf.
Vinyl, for whatever continuing reason, often does sound better than digital formats but only because it’s mastered with more care.
It could be that it’s physically impossible to master vinyl for extreme loudness, but whatever the reason is you can absolutely pick up a vinyl copy of an album and find it sounds much better than the streamed or CD version.
EVs are still far more efficient than any combustion engine per unite of energy. It doesn't have the density of gasoline, but you're paying a lot less per mile driven no matter how boxy it is.
I drive by an old Scout that just sits there and wish I had time to update it. It could be a fin project.
I want to like the VW Scout, but it's so big compared to IH Scouts.
Okay so its not perfect as far as efficiency, but isn't it still significantly more efficient than whatever gas engine was in it before?
This is so cool. I would love to revitalize a generation of great, but perhaps boring older cars with FSD. Just so much work...
OMG, that steering wheel is just hideous!
I've been waiting for someone to do something like this as long as I've known electric cars to be a thing. I hope they just start making them like this.
People have been doing Tesla powertrain swapped vehicles for a long time. Here's a Jaguar from 5 years ago: https://www.youtube.com/watch?v=DxxaHYlCuN8
So where is the startup to convert any car to electric?
That is the safest 1966 Mustang on the road.
Yeah, kinda boring in that respect.
This is such a waste of time and money and wow it’s absolutely gorgeous and I want it
I hope they implement this in other car models as well, and make autonomous driving possible in Japan too.
When will I be able to get an affordable restomod for my early-2000s Jetta?
I wonder if tesla will see this and try to invalidate the VIN from using FSD
That would be a pretty shortsighted move from a PR standpoint. Nobody is going to blame a restomod FSD wreck on Tesla. But turning off FSD remotely just reminds everyone that Tesla can and will take control of your car.
Don't they already block using their supercharger network for rebuilt titles?
Interesting, love the concept. Don't love the modern interior.
It should've been a black Pontiac Trans-Am.
A Mercedes Benz 770K would have been even more appropriate.
I took my 1972 Dodge small block engine out and converted it to 400 hp. Had to upgrade the transmission, driveline, brakes, radiator, and suspension to match. I self-drive it.
Nice work, but is it just me or does this take away from the car’s original spirit?
You are entitled to your opinion, obviously, as is anyone else. Restomods have been around since forever, and this is hardly the first EV conversion (almost certainly not the first classic Mustang EV restomod either).
Personally I think it's pretty damn cool. But I have always been a Mustang fan, and I know that this era of Mustang is not especially collectable. They made quite a large number of them and plenty are still running.
This is more of a body kit for a Tesla than a drivetrain swap in a Mustang.
To be fair though, this is beyond typical restomod territory. This isn’t an LS swap with modern wheels and suspension. This most certainly changes the spirit of the car.
There's no question the first generation of Mustangs are the most collectible.
It is like putting a Samsung shell on iPhone. It is still an iPhone.
A 66 mustang ... without the sound? A lion without a roar.
And at these prices, you would reduce far more carbon by investing that money in solar panels. 50k buys at least a 20kw system that would more than make up for your summer weekend drives in a classic mustang.
I know this is not the way the industry or regulations work, but I wish electric car platforms let you pick body styles without waiting for a whole model to come out. I'd love an electric Suzuki Jimny body, and could care less about the driving platform.
Why destroy this beauty
To keep it alive
Still bothers me that “full self driving” is not fully self driving. They shouldn’t be allowed to call it that.
I have a dream that electric motors + battery get so compact that you can make kits that fit engine blocks and large numbers of cars can be custom swapped relatively cheaply.
People have been doing these conversions forever with teslas.
There was a shop in Dallas back about 20 years ago that did an electric conversion of a H1 Humvee. Since then there’s been lots more conversions like that and to me that is a valid recycling business.
> It demonstrates that Tesla’s hardware and software stack is more portable than the company’s licensing struggles would suggest.
Unless I missed something, this is a completely unsupported claim by the article. Passion projects and retrofits are nothing at all like manufacturing.
The article claims that the whole project only cost $40,000, and then compares that to electric conversion offerings that cost $75,000 (and mentions that the global conversion market in 2024 was $5.9 billion). I think the implication is that there could be a large market for FSD conversions that goes beyond passion projects because it is not only possible but affordable.
I would be surprised however if this project only cost $40,000, when you factor in the cost of labor and maintaining a facility to do this work.
It's specifically talking about the "FSD" model under the hood being able to run on this retrofit even though the cameras don't align 100% like they originally would.
good luck getting it repaired though honestly this is really cool
It is basically a Mustang body on a tesla chasis... which misses the point of having a classic car.
While there is nothing wrong with converting your classic car to electric, if the powertrain is shot (they are harder to maintain as they age), but IMO, it looses the charm of the point of having a classical car.
Few years ago, there was a trend to do these conversions, but that stopped as people realised the car loses its charm and the feel of having old classic car, and most of them are not being used as dailies anyways.
As a classic car owner/hobbyist, this disgusts me. But thankfully there are at least 200k other restored/restorable 60s Mustangs out there.
Good news is that you can get a brand new 66 shell stamped in Taiwan for about 20k.
I wonder if they're as "good" as the Chinese AE86 shells.
Oh, a good looking Tesla.