191 comments

[ 15.8 ms ] story [ 276 ms ] thread
(comment deleted)
So we could have a slightly bigger inside to the usual EV SUV. woo
A smaller battery would imply fewer resources required per vehicle, which will be important as demand for batteries increases.
Optimistically, we could have smaller cars without compromising on cabin space or storage.

…except when large SUVs are bought as a status symbol, or the race to drive the biggest tank on the road in defense from the other tanks or the road.

I actually suspect part of the reason people keep buying larger cars is because larger cars are safer and sometimes get better safety scores because in a two vehicle collision a heavier vehicle will experience less acceleration in the crash. Also more space to absorb the impact, increasing the time over which the acceleration happens. This genuinely makes heavier and larger cars safer if they are designed properly and maintain the same braking performance, which is totally possible. However the issue is that a larger car is only a safety advantage if your car is large compared to other cars on the road, so if everyone else also buys heavier cars no one wins and pedestrians, the environment and so on all lose.
Also they’re worse for everything else: the poor handling, especially with overpowering, means you’re more likely to get in a crash, etc. Manufacturers have spent billions making them less likely to lose control, roll over, etc. but the physics are unforgiving.

Drivers discount that because they like to think that those are things they control and just won’t do, but every year some fraction of people will hit ice, be distracted or angry, etc.

Bigger inside than modern SUVs? What does people transport? How many people ride regularly? What size are people?
Two adults, two kids in car seats, and a dog. With that, even a mid-size SUV has no space for cargo. If you drive a sedan, you have to leave the dog at home.
Maybe if they don’t put a foot-wide console in the middle there would be more space for people and cargo.
That's why station wagons and vans exist. SUVs do not provide more space, they just occupy more (on top of being dangerous to others)
Large station wagons no longer exist in the US market. The biggest one remaining is probably the Volvo V90, which is a nice car but can barely fit 5 people, and if you put a dog in the back then there's no room for cargo. It's also too expensive for most families; several large SUVs are actually cheaper.

Minivans are a good option for some families. But they suck for towing trailers, and the ground clearance is terrible if you ever have to drive on a dirt road.

Yeah really, like I’m gonna drive a minivan or station wagon out in the woods, or drive my four wheeler into the back when I need to take it somewhere.
Well, this is why A6 Allroad and the newly bastardized E-Class wagon exist.
Right.

I am all in on retrofitting a 1973 Ford Country Squire faux-wood-panel-in-green station wagon with a set of Kia uniwheels.

>> the Volvo V90, which is a nice car but can barely fit 5 people

It's a lot bigger than my Prius, and that happily takes 4 (5 with a squeeze) climbers with all their gear!

This sounds like a very US related problem that is unfortunately expanding to Europe due to car manufacturers marketing strategy.

I still find it crazy that no station wagon exists other than Volvo.. like Audi does not export their A4/A6/A8 estate version? Skoda, Mitsubishi, Toyota neither?

I can understand if you live in an area where roads are not well maintained or mostly dirt, SUV is really alluring. But for city driving is a massive (literally) problem for everyone.

There are a few other station wagons still available on the US market. Audi in still sells the A6, but it's very expensive and has limited interior space. For most middle-class families a large SUV like a Chevrolet Tahoe is a much more affordable and practical choice. Most Americans don't spend much time driving in dense cities, so that's not a real problem worth worrying about.
The average number of occupants in a road vehicle is under two, yet they drive a massive tank. The typical pickup truck on the road also has an empty bed.

Meanwhile, I can fit four people and some light cargo comfortably in my hatchback, or put all the seats down and carry more cargo than the typical SUV.

I drive a small petrol powered MPV. It has a luggage capacity of 710 liters in the trunk with 5 passengers, 160 liters with 7 passengers, or 1900 liters with 2 passengers. Moreover the trunk is square, so I can go to Ikea and buy a whole new kitchen and fit it in the trunk. The cargo loading capacity with 5 adult passengers is 450 kg. It has a towing capacity of 1650 kg. It also has a large roof box. New price was around $30k depending on trim level.

Now show me a modern EV that has similar specs, and you are talking about a large SUV that's up around the $100k price range.

It's a simple fact that while modern EVs are often styled as SUVs, the inside is comparatively tiny since the battery pack takes up a lot of space.

I think more importantly is that you can add more battery capacity, making EVs much more competitive with gas vehicles.
(comment deleted)
As a drivetrain nerd, this is awesome!
More mass inside the wheel increases unsprung weight and worsens some ride/suspension characteristics of a vehicle-

that being said the tradeoff might not be too bad in this case.

It’s just reduction gears, not the motor, so hopefully it can be made light enough. Shame that their promo didn’t say a word about the mass.

Quad-motor all-wheel torque vectoring sounds awesome tho.

My semi-educated guess is the heavier weight of the vehicle means the wheels can bear a bit more weight.
Is there some way that this or similar can be copied by all automakers, or is this all patented by Kia/Hyundai in such a way that most improvements are completely locked in?
Assuming the technology has actual value, Kia/Hyundai would likely be more incentivised to license it to other vehicle-makers, than to keep an exclusive license for themselves.
Especially considering how their name and brand are in the dumpster, and likely won't recover. Their cars will be targeted for break ins for a long time to come due to how viral the vulnerability is, regardless of whether or not the model is actually susceptible to it.
Maybe I can't tell from the video, or maybe they're not working on this yet... But what about the steering wheels? Surely the driveshaft will have to have a joint on it on the front wheels in order for the front wheels to turn? Also this doesn't look like it would deal with camber changes which are needed to keep your tires flat on the road when cornering...
I also notice that unsprung weight was not mentioned. The "Uniwheel" clearly has advantages over CV joints, but it also looks like a significant increase in unsprung weight.

They also gloss over "splitting the motor". It looks like 4 individual wheel motors, which would indeed be optimal for the torque vectoring that they mention.

Still, looks like a very cool development. I hope it works out to be a real advance

It is a compromise - maybe slightly more unsprung weight vs a CV joint, but less than a hub motor.
Right, it seems like this will only work on non-steering solid axles, not with independent suspension.
The front wheels still have a CV joint to facilitate steering. See 4:17 in the video.
This will work and you will still be able to steer.
FWIW, there was recently a now-defunct car company (Lordstown motors) that utilized an in-wheel motor from the company Elaphe (https://in-wheel.com/en/technology/#in-wheel-motor-technolog...)

I think the tradeoff is between more flexibility & fewer parts (no cv joints) but possibly more rolling resistance.

But watch the video, this uni-wheel thing is different. It isnt an in-wheel motor, more of a cv-joint alternative?

unsprung weight I believe is also a big issue with hub motors esp as you get up to car size at speed
It is the issue for in-hub motors. Solve that one and the world will beat a path to your door and shower you with money.
The other issue is that potholes my get quite expensive.
Yes, the impact from a pothole can be substantial. Having an upset alignment is one thing, damaging the guts of an in-wheel motor is a totally different picture. But: when properly packaged and lightweight enough I think that is a secondary concern, you can design with that in mind given that it is more or less predictable from which direction and with which magnitude such an impact would happen. But every pound counts and the shocks+spring will have to be matched to the increased weight.
Have you got a source. Everyone mentions this, but I've never come across anyone trying and failing.

Motor scooters have unsprung motors, the motor is a significant proportion of the mass.

Motor scooter motors weigh a fraction of what the equivalent motor for a car would weigh, and it's exactly the problem:

The issue is that it scales poorly, unsprung weight is bad enough that on some cars the disc brakes live 'inboard' on the front wheels (older hydraulic suspension cars).

Essentially the larger the unsprung mass the bigger the hit every time the road surface irregularities cause the wheels to move relative to the vehicle body. The larger the mass the more energy gets imparted, and the more energy that gets imparted the bigger the impact and all that energy then needs to be absorbed so you'll need a stronger (and heavier) suspension system, roughly half of which is unsprung mass!, and a larger battery to compensate for the losses in the suspension which also adds weight (harder suspension -> bigger range but less comfortable ride). So it feeds back on itself and that's why it scales so poorly.

I can sort of buy this at the upper end. But a small car with 4 relatively small motors with normal tyres v a range rover with run flat tyres and massive disc brakes?
In a small car it could work, but you'd be very much limited in power and weight.

See upthread links and subsequent updates

https://news.ycombinator.com/item?id=38485097.

Note that more unsprung weight has other effects as well, such as poor handling because the wheel bounces higher and so will be out of contact with the road for a longer time when it encounters an obstacle. If you think of the weight of the wheel as the weight on a pendulum it is easy to see why that would affect the period of the swing.

You can buy a car powered with in-wheel motors here:

We have developed the first production car to use in-wheel hub motors. https://mwmotors.cz/luka-ev/

This car maybe too, as far as I understand it. A Russian UAZ on steroids: https://www.youtube.com/watch?v=RCOf_33C1qc

815 Kg curb weight, aluminum chassis. It makes sense for a vehicle with that form factor (2 seater little sports car) but you can't really compare it with a full sized sedan, and note too that it has a tiny battery pack (22 KWh), range is 'unofficial' 300 Km and likely will be adjusted down based on weight/pack size. The four motors together are less than half the rated power of what you'd find in a more conventional (is there such a thing already?) EV. But with that light total weight it will be a zippy little car anyway (most EVs are over-powered for their weight).

Still, very impressive achievement and a nice testbed for this kind of tech, it may also make sense for motorcycles and for vehicles that are not meant for highway use.

What is the advantage of in-wheel motors?
That you save on a whole bunch of linkage and drive train components, four wheel drive without the associated complexity, that you can integrate the wheels, motors and brakes, cleaner airflow under the car and that the system as a whole can be lighter weight than the same package with motors near the centerline of the car, so you get somewhat better range. It also gives more space for batteries and/or more storage space. That's why manufacturers will keep trying to get this done, if they can get it to work cost effectively and without impacting that unsprung weight too much it will give them an advantage.
"cleaner airflow under the car"?

Why should the airflow be cleaner?

Because that's automatically better range, cleaner airflow is less drag. That's why cars have all of those underbody skirts, it improves their mileage (and helps with not fouling up the more sensitive bits).
I think they mean smoother or less turbulent airflow, not cleaner in the sense of less pollution.
That Luka has some pretty interesting tech and is so sharp-looking. I hope it's a massive success.
Gorgeous little car! Reminds me of the sports cars of the 50's and 60's. But they apparently dropped the hub motors.
The height of that little sports car reminded me of the ultra-low-height 1960s Lotus Europa (first version) - one didn't so much exit it as roll outwards from it.
I've had a kei-car sized sports car that was a hoot to drive but it felt like putting on gloves just to get in.
Bummer. Like Mazda’s experiment with the Wankel engine it would have been very cool to see the results
Well, they did make them. I suspect that the safety concerns were the main issue (less road contact compared to inboard motors).
There's another startup, Aptera, which is ostensibly still working towards producing a vehicle using Elaphe hub motors. Interestingly they are efficiency focused so I wonder how the motors are working out for them in that respect.
Very well in fact, you have to look at the whole drivetrain efficiency, not just the motor. Elaphe has a webinar on this topic on their youtube channel, look it up.
Yeah, it’s not a hub motor, so the unsprung weight should be a lot less.
I'll be interested to see if the Elaphe in-wheel motors are designed into a high ground clearance 4WD. The site says added weight 25-65kg but maybe that is two motors for 2WD? A grunty 4WD tyre like a 33 inch (33/12.50R20) might weigh 33kg per tyre (not sure if rim weight), so the additional unsprung weight for the motor is comparatively reasonable per wheel? In-wheel would allow the suspension to be designed for much better clearance (since no shafts). Tyre might have a speed rating of Q (99mph).
Pretty sure I measured a 33" BF Goodrich MT (mud traction) tire on an aluminum 15x8 wheel at 93lbs. However, it was about 25 years ago so memory's probably a bit fuzzy.
Is Lordstown defunct?

As far as I know, this Czech company uses in-wheel motors:

https://spartan.mwmotors.cz/ev-traditional-technologies/

Looks like they already abandoned that idea: https://mwmotors.cz/luka-ev-traditional/
Interesting, I wonder what the technical reasons were for the change. It has gone this way with pretty much every hub motor concept that I've seen so far, either they don't make it of the drawing board, or they end up abandoned at the prototyping stage. The GP had me wondering if someone had actually managed to conquer it but alas that doesn't seem to be the case. It's a neat little car though, reminds of the Daihatsu Copen that I had at some point.
unsprung weight is what every engineer who looks at a hub motor will think. The less of that the better. Engineers would love to move the brakes off the wheel (there are safety reasons not to, but it would make the car handle notably better if you did)
Citroen DS/ID had that, SM as well. CX and XM did not.
Don't even need to be an engineer just go ride some nice bikes, it is an amazing and very tangible lesson in the cost of unsprung weight.
Cc can you still do a burnout with these? It looks kind of fragile.
Disappointed that they didn't explain how steering works. Seems like a major oversight.
A very simple and incomplete explanation. You have an electrical or hydraulic steering rack that spans your vehicle just behind the wheels behind your engine and close to the bottom. It connects to each wheel with a tie rod that has a ball joint on the end (not to be confused with your actual ball joints that are part of your control arms). When you turn your steering wheel it moves the tie rods to steer the car. Steering doesn't really have anything to do with the CV Joints or power to your wheels, it is a separate system.
>Steering doesn't really have anything to do with the CV Joints or power to your wheels

When you turn the wheels the axles need to flex somehow around that rotation. This is one of the things CV joints do well: they can handle not just moving up and down, but also the wheel turning left/right. This thing is a bunch of gears and seems to handle only wheels moving up and down and not turning.

My understanding is they can't shorten the CV Shaft/Joint because it causes problems.If they can move it into the wheel hub, it allows them to make the motor smaller which is where they save the room. By taking one big motor to drive both front or rear wheels, they now have smaller ones much closure to each individual wheel. This reconfiguration leaves more room for interior space or extra batteries.
Seems like it's only in-plane motion, so the wheel can't tilt/turn?

I assume that means they still need a CV joint for the front wheels regardless, and that it gains no camber when cornering.

Most likely they will have a ball joint (like the current inner CV Joint shaft has) on the end of the shaft that connects to the device in the wheel hub.
One of their animations indeed showed a CV joint in between the motor and the device. But you only need one and it can be absolutely straight when the car is not turning so the efficiency should be good.
Correct me if I’m wrong here but this is just sort of an alternative to a cv joint, sort of because it can’t handle rotation. It also has nothing to do with EVs specifically, a ICE could use this.
An ice could use this but there is no real purpose. The whole goal was to make the electric motor smaller and closure to each wheel so there is more room for other things.
Yeah fair, though ICE cars also benefit from more space. Maybe I’m just salty because I had to watch 10 minutes of their video just for a clip that was about 1s long that actually shows what it is instantly, they should have lead with that…
You probably don't want to use a small ICE at each wheel though.
An ICE still needs all the shafts connecting the wheel to the engine.
I honestly thought that Tesla made a version of their sedan which had 4 motors -- one for each wheel. When I looked into it, all I can find is that the CyberTruck should have 4 motors. Wasn't there a quad and duel version of the sedans -- what are they referring to?
They have dual and triple, the triple (Plaid) has two motors on the rear if I remember correctly.
The thing I see not mentioned in the article/video, is how the wheels can tilt/turn/angle - without needing a CV joint again.

The designs seems to cover all the motion of the wheel in the suspension - but I don't see how this could accommodate steering.

Is this for back wheel drive only?

Curiously the animations shown in the video, show the uni wheel used in the front of a front wheel drive car - but without ever showing how the wheels steer.

How is that supposed to work?

(comment deleted)
Maybe the motor is supposed to turn with the wheel, that doesn't sound like a good idea.
That'd make the motor the pivot point. No idea if that's good or bad though. :)
Why not?

You'd have an added complication of making the shock absorber the pivot point, and attaching the motor to the top point. Or having a separate pivot point where the motor or wheel can slide up and down.

This does raise the question of why not just attach the motor straight to the wheel.

The standard response to that is a hand wavey "unsprung mass". But has it actually been tried and shown to be an issue?

Well, if the motor is unsprung mass, you no longer need this contraption at all, and you can just directly connect the motor to the wheel...

So, that's probably not it.

I'm no mechanical expert, but I wonder how you prevent damage to the motor if the wheel encounters an obstacle when it tries to rotate (against a curb for example) or if it's forced to rotate by an exterior force.
Cars already have traction control, and there's probably a thermal cutout already in the motor.
Unsprung mass sucks. You don’t notice on smooth surfaces, but if you are cornering on a rough surface the wheel ends up losing contact with the ground after each bump. It’s extra-dangerous when its the rear wheels because the car can spin out.
It's not unsprung. The wheel and gear bit are the only unsprung parts; think of it like the motor attaching directly to the wheel bearing, and the wheel is able to move vertically on the bearing.

I think this only works with RWD, because the turning wheels can't be sprung without also having to rotate the springs.

It would be a better suspension (closer to racing suspensions) than standard, because wheels wouldnt toe in/out when they moved down/up. Looks expensive as hell though

tlb was replying to this part:

"This does raise the question of why not just attach the motor straight to the wheel. The standard response is ...unsprung mass"

> The standard response to that is a hand wavey "unsprung mass". But has it actually been tried and shown to be an issue?

Afaik with high performance motors, if you want optimal performance and space efficiency you take the unsprung mass for granted and move everything in wheel and try to reduce the mass as much as possible. As a bonus you do not need joints, just a planetary gear system.

Why only high performance though.

A motor per wheel simplifies everything.

No diff, no CV joints, you don't need this, just a simple planetary gearbox.

You gain space, lower manufacturing costs (theoretically, assuming multiple motors doesn't offset the other cost reductions, although as you could base a 2wd and 4wd high performance version on the same motor, that's an even harder case to make).

Everyone says "unsprung mass" but you also have to provide cooling without ingesting all the dirt and water off the road. It's not that simple.
Brake lines are a solved problem. Water lines wouldn't be any more difficult. If air cooled it could be a sealed unit.
The wheel can move up and down relative to the motor, so you can still attach a shock absorber to the wheel, without the motor being involved. The video shows this at 2:36.
Yes my point is, why have the complication of this set up at all.
To me it doesn't seem any more complicated than a pair of CV joints.

Benefits according to the article: "reduced packaging size, improved ride quality, greater durability and, importantly, increased efficiency."

I've been under the impression that putting the motor on the wheels increases gyroscopic forces which can cause handling problems.
It isn't rotational mass though, just unsprung.
Probably a combination of having some way to rotate wheel + motor around some axis and the ability to simply vary the motor speeds on each wheel independently.
But then you need empty space around the motor, so that it can rotate there.

That somehow seems a weird choice for a device that's all about improving space efficiency.

(comment deleted)
You can see the motors turning in some empty space in the video yes. But it’s mostly in the space that would be taken up by the CV joint.

Just look closer at the animations in the video if you want to see how the space is being used

In the video there was a brief animation at 5:06 showing the entire motor rotating with the wheel. So the CV joint (if you can call it that) wasn't really removed, just moved to the part between the motor and the chassis.
The next logical step would be integrate the motor directly into the wheel. With 4 wheel drive motors it wouldn't require huge motors, they could each be small enough to be packed right up against the inside of the wheel.

I'm actually surprised they would even need the gearbox. Why not just put four Outrunner[1] motors, one on each wheel, and let the wiring be the only thing that has to flex when the wheel is steered.

1. https://en.wikipedia.org/wiki/Outrunner

Hub motors have existed for a long time. The issues you'll have to overcome are the same.
Hub motors make wheels quite heavy, which means that there's more unsprung mass [1] in the vehicle, making it less responsive under acceleration/braking and more noisy and sensitive to vibrations.

They're common in electric scooters, bikes and bicycles though.

[1] https://en.wikipedia.org/wiki/Unsprung_mass

newer designs are silent, weigh about 2lbs or a little under a kilo, can produce upwards of 100nm torque at 1000w.
1kW isn't going to suffice. Car motors are >100kW.
Well, that’s overkill for Hyundais, 20-50kW max is all you need per wheel. You can produce >100kW with some high end racing hub motors [1]. You can also scale it up to generate more torque. Gear it down, etc.

[1] https://www.koenigsegg.com/quark-emotor

I thought the same thing and was disappointed to see a planetary gear setup to turn an outer ring gear. Shame, we have come so far in EV motor design and yet Hyundai seems stuck on drivetrain power design.
My experience with a planetary gear in a bicycle is that it produces additional noise. I wonder if this would be noticeable for an automobile.
I'm sure it would. Any mechanical gears will produce noise at higher velocities so curious what this sounds like myself. Electric hub motors produce a wine from the EMF that is audible but that's a different sound (the hum of a tesla going by) or a Surron that uses a radial flux motor with chain link gears (this is my favorite hobby design, chain gears are super useful).
> Electric hub motors produce a wine from the EMF

the sound is not coming from EMF, but rather from speakers due to safety regulations for pedestrians

> The designs seems to cover all the motion of the wheel in the suspension - but I don't see how this could accommodate steering.

It actually doesn't. A wheel's camber changes as it moves up and down, which is why race cars typically have negative camber when going straight, so the tire's entire contact patch is used when going around a turn, which compresses the suspension of the outside wheel.

This design doesn't appear to handle this case at all. I'm not sure how a traditional suspension setup could be modified only move a wheel up and down, without any change in the camber, considering the fixed attach points of the various control arms.

How do they keep this clean? A CV boot is compact, flexible, and fixed to the shaft. What kind of dust shield can cover the big vertical gap and handle a spinning shaft?
Yea I don't see how this will work. Just a little bit of sand or water in there and it is a goner.
Unsprung weight is a clear disadvantage to the solution. It will significantly affect ride quality, no way around that. Those planetary wheels look more fragile than the traditional solution. You also need to keep that entire thing closed and lubed as well and even so it looks like added friction, wear and a more expensive repair job when it goes bad.

All that being said it could be a good niche solution. A small city car would benefit from the space savings and does not need to go high speeds over less than perfect roads. Nor does it rack up that many miles in wear and tear. In general I think electric will bring more diversity in car designs depending on purpose.

> Unsprung weight is a clear disadvantage to the solution. It will significantly affect ride quality, no way around that. Those planetary wheels look more fragile than the traditional solution. You also need to keep that entire thing closed and lubed as well and even so it looks like added friction, wear and a more expensive repair job when it goes bad.

But uni-wheel changes paradigm!

At least they are nine identical gears? So just keep a barrel of them handy!
I expect the bearings wear out before the gears. So just replace the whole thing. Then rebuild it with new gears and bearings later.
Worth noting, placement at the wheel is somewhat arbitrary. There's nothing to prevent this contraption from working the other way around and being mounted at the chassis instead, aside from squeezing out every bit of space optimization. Axle support would be a concern either way. Controlling deflection of the drive axle adds a structural requirement, certainly a well supported bearing close to this device.

The gears will be turning at a large multiple of wheel speed, so will require roller bearings except for low power and speed use: a radial and thrust bearing for each end of each gear, or 22 of each per wheel. Increasing torque also increases the level of thrust support required for the moving gear carriers handling all of those helical gears.

It's a neat idea for highly space constrained applications that have to package a drivetrain very close to a wheel, which does end up being a rather niche use case. Mars rovers, perhaps. Trying to compete against CV joints in passenger vehicles with this would be quite difficult.

It's trying to compete against a CV joint and gear box.

That seems like an easier sell.

I'm not sure if we're talking past each other but you wouldn't need an axle to deflect. The motor drive could attach directly to the 'uni wheel' which is in turn connected directly to the wheel. If you're talking about the motor drive shaft, then surely you'd just be moving thrust bearings from the gearbox to the motor (if motors don't already have them)

If the drive source is attached directly to the uniwheel, then handling radial drive inputs isn't required and the uniwheel isn't necessary. It's effectively a hub motor with reduction gears at that point. Once the motor is mounted remotely from the uniwheel and the uniwheel is handling radial offsets, then we need to prevent the axle from getting pushed around radially. Rotating shafts that deflect introduce off-balance vibration and destructive mechanical loads. If the motor is located very near the uniwheel, keeping the shaft's free length very short, the motor's output bearing does effectively provide that support. Then we're only constrained by relocating the suspension elements.

Thrust bearings are for the gears. The depictions show helical gears, which would surely be necessary when operating at transportation speeds for noise reasons. Helical gears experience axial forces, which would need to be handled for each gear. Mainly on one side for ICE drivetrains, since coasting loads are small, so one side could get away with a ground and hardened rub surface. EV's using regenerative braking would need robust thrust handling on both sides.

Herringbone gears would eliminate the axial thrust issue, at substantially more expense. The easiest method is to separate the gear areas with a central runout groove so you'd have a hope of grinding in a decent surface on the faces, then it's only a little more than double the work; effectively cutting two gear shapes per gear plus the runout.

https://en.wikipedia.org/wiki/Herringbone_gear

> you wouldn't need an axle to deflect

This is really cool for rear wheels. I don't know why their demo shows it on all 4, since it has to be able to steer (note that the demo does show a CV joint in the axle on the front, but that seems like it would limit steering angle severely, since it is just 1 joint. They're probably planning to apply torque to steer).

> Worth noting, placement at the wheel is somewhat arbitrary

You're suggesting placing the ring-gear in the chassis and running the axle to the wheel? That would involve two reduction gearsets in the chassis instead of the current "one plus a CV gear," which kind of defeats the purpose of reducing chassis space usage.

> The gears will be turning at a large multiple of wheel speed, so will require roller bearings except for low power and speed use: a radial and thrust bearing for each end of each gear, or 22 of each per wheel. Increasing torque also increases the level of thrust support required for the moving gear carriers handling all of those helical gears.

I imagine the final product would use herringbone gears, since the large number of thrust bearings would all add to unsprung weight.

An axial flux wheel motor is a much simpler and less fragile design for ev’s. I’d like to see more experiments with that.
Regardless of the merit or practicality of the idea, I really loved the video and the animations. I'd happily watch a series of cars and other complex machines explained in the same style.
I wonder how long will it take this to break if use on real (not-so-perfect) roads.
My first thought, too. All those unsealed tiny gears. One fragment of metal filing in there? Perhaps a cascade failure of teeth chipping and flinging fragments until total breakdown.

Reminds me of the head of a waterpic toothbrush - a packet of tiny gears, the weakest point in the device, constantly wearing out.

This wouldn't have worked for ICE cars, because in order to have a reasonable range of up-and-down movement the gear ratio needs to be quite high (ie, the pinion gear has to be much smaller than the ring gear). It's 6.2:1 in the article, about 2x what's appropriate for the top gear on an ICE car.
I wasn't clear why this couldn't be used on all vehicles but that makes sense, thanks.
In and ICE you have the transmission - just put different gear ratios in the transmission if you need to get a different output speed.

I think this could have worked with an ICE, but there is no real point: the CV joint is simple and the advantages to this don't apply when you have a single large power source.

This mostly makes sense if you have four individual motors right next to the wheels, that turns with the wheel. That’s obviously also not viable with ICE cars.
I love innovation.

There still needs to be a CV joint to allow steering, but short of hub motors[1] There isn't much you can do about that.

What I would be interested to see is how lubrication works. Do you think they just have a bunch of oil in the bottom and rely on centrifugal force to fling it to the ring gear? Then let surface tension transmit it to the sun gear?

[1] They are heavy and heavy wheels affect ride quality, as suspension works essentially by allowing a light wheel to move up and down dampening against the mass of the heavy body. If the body is light, then you need to have less dampening, which means more shock/impulse is transferred into the main body of the car/

The hard thing with lubrication is sealing it so it doesn’t leak. Considering the pinion gear and shaft moves in 2 axes relative to the housing, that could be tough.
> There still needs to be a CV joint to allow steering

Naw, because if you look closely in the video you can see that the motors turns with the wheel. With small EV motors that’s not unreasonable.

Presumably the motor is free to tilt up and down a bit as well to deal with changing camber.

> but short of hub motors[1] There isn't much you can do about that.

I mean, it is pretty much a hub motor that has been decouple in the vertical motion so it doesn’t contribute to unsprung mass, and with a good reduction gear as a bonus.

Hmm steering and ride quality questions aside - i'm no mechanical engineer...

How much would damaging a wheel cost?

I was wondering if it would increase the incentive to steal car wheels.
s/wheels/engines in this case :)