57 comments

[ 4.7 ms ] story [ 101 ms ] thread
(comment deleted)
I thought there are already plenty of active suspension systems on the market or is this different?
The Mercedes ABC (Active Body Control) system is good. But if you watch the video - this one is amazing.

The Mercedes system used hydraulics. If the Bose system works the same way their truck seat does (the "Bose Ride") then it's run by a combination of pneumatic & electromagnets, and I'd be curious how they get enough current to run it.

ABC: https://www.youtube.com/watch?v=clyUmA5lRcg

Bose suspension: https://www.youtube.com/watch?v=q8sVDenpPOE

Bose Ride: https://www.youtube.com/watch?v=T_G9GCpWAcM

I'm guessing that's what makes it viable.

If it's going into future vehicles, well, a number of those are going to have plenty of current.

That suspension would make a pretty sweet market differentiator for electric cars. Would go a long way toward making up for range anxiety.

There's been talk of switching to 48v batteries in cars (for the starter and accessory circuits, not for the EV circuits which already run at much higher voltages). AFAIK, only some Class B RV makers are using them for their house batteries.
I think the updated version is called Magic Body Control [1] and the videos show an impressive behavior. But when this tech came out a couple of years ago, Autobild have tested it thoroughly and said that it works well on big bumps, but on everyday small street bumps and cracks that you mostly see on the streets, there was no effect. And it costs additional ~10k EUR...

[1]: https://youtu.be/ScpgI1w5F6A

They claim that it responds faster. Following the article's link to the source AutoCar article, there's this quote:

> “Our technology is the fastest proactive ride system that exists today,” Clearmotion CEO Shak Avadhany told Autocar in an exclusive interview.

There may be other differences. I'm not very knowledgeable about all the active suspension systems available, but it seems that the most common system today may be MagneRide:

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

This is basically a shock absorber but with some magnetic fluid inside it whose viscosity can be adjusted. So it does not move the wheels up and down, it just adjusts the opposing force. Though it can do it quickly.

There is one system on the new Audi R8 that does more:

https://www.audi-mediacenter.com/en/press-releases/looking-a...

This seems very similar to what the Clearmotion system does, as the above AutoCar article says the Clearmotion system "works by turning a car’s dampers into actuators".

So it seems the Audi R8 system and the Clearmotion system are doing similar things, but the Audi system does it "18 times a second" whereas Clearmotion technology page ( http://www.clearmotion.com/technology ) says "it only takes five milliseconds for the Activalve’s motor to counteract", so apparently 10 times as fast.

The problem with magnetorheological fluid dampers is that they wear out very quickly compared to conventional dampers. The magnetic particles are basically very fine grit, which abrades away at the valving inside the dampers.
The active suspension systems I have seen let you switch suspension profiles on the fly. Some are advertised with "millisecond" response time implying automatic tuning based on force sensors. This demo shows aggressive predictive tuning.
They need to make it actually work first. The well-known demo video is all staged, the suspension was specifically tuned before each demo, it wasn't universally capable at all. There was probably a lot of manual control involved as well.

Citroën's hydropneumatic system was always better, especially on the C6 and 3rd gen C5. It's a crying shame they're moving away from it.

Somewhat surprisingly, according to their technology page (http://www.clearmotion.com/technology), they are doing some kind of cloud upload and learning thing:

> Connected by a central gateway, it collects road data and stores it in the cloud to be analyzed and instantaneously retrieved.

...

> Using a precise, millimeter level HD map of the road surface ahead, advanced control theory algorithms continually adjust suspension parameters to enable the best ride possible. An additional layer of reinforcement learning algorithms continually learns the road’s profile, and your ride improves over time.

So, it's just a guess, but perhaps they've overcome the need to specifically tune it by having it learn and share data.

This description suggests a plausible synergy with all the sensors designed for self-driving cars.

With them, the suspension could "see" where exactly a bump or a pothole is ahead, and adjust accordingly.

I vaguely recall seeing some self-driving car presentations that have mentioned exactly this possibility.
Ha. If the car has a detailed map of every bump and pothole it can "stage" it's driving everywhere!
The detail level on that Mario would have to be insane. And probably rather interesting for municipalities as well.
Great now they can make it a subscription service.
That demo is clearly staged. However this other one https://www.youtube.com/watch?v=kJ0ljaqeZ1g (0:40) seems more legit. It is short but the car is driving on a real bumpy road rather than on a flat surface with carefully placed obstacles.
Well with today's lidar type tech and computing power available because of smaller more efficient chips they can probably auto tune the car in real time for the surface.
It looks like an unrealistic driving game.
This may be one of those happy situations where changing requirements make a bug fix much easier than before. Now maybe that driving game is realistic and they can close the bug by just saying, oh yeah, these cars have active suspensions.
Montréal needs this.
You beat me to it. But would they just allow the nid de poules to get bigger!
Is this basically maglev at the point where the body meets the chassis?
No, it still has springs as I understand it. However the dampers are replaced with some sort of linear actuator so the degree of damping and even the position of the suspension can be controlled.

If there were no springs it'd bottom out when the car was switched off, you'd still need a normal ride height when getting in to the car.

I was thinking super strong opposing magnets, with some sort of limiter device.
This is on my list of technologies I've been waiting to make commercial availability. Just waiting for tech to catch up. First is the wasp x-jet (flying pulpit) from the 1980s.
What if you need to swerve while going over a bumpy road? Steering won't do much good if your wheels are in the air.
A vehicle's suspension serves 2 purposes:

1. to make the ride smoother and more comfortable

2. to keep the wheels in contact with the ground to ensure consistent, safe handling

This suspension is capable of lifting the wheels off the ground, but the practical purpose of active suspension is to use a proper PID loop and sensor-enabled anticipation in order to keep the wheels pinned on the ground while minimizing body movement.

Would this tech render speed bumps useless as a safety feature, by removing all the unpleasantness of driving over one?
Yes, although I suppose it would be possible to build in some kind of exception to require it not to respond to them? Maybe you could even have “smart bumps” that would just be some location data pushed to the cloud that tells the suspension to “rumble” without the need for a real bump. Of course that would only work with very widespread adoption of the suspension.
If you're going to require cars simulate speed bumps why wouldn't you just enforce a digital speed limiter?
Lobbying. I can’t think of another practical reason that isn’t primarily based on fantasy. A digital speed limit could allow for a minute say, of exceeding the limit for the sake of evasive maneuvers, passing, etc. lobbying and public perception though, won’t allow it.
Yes. Controversially, they then become a risk for cyclists, which I found out the hard way.
And firetrucks, ambulances, and mobility scooters for the elderly. The solution? drive a large enough 4x4 truck so they don't matter.
Similarly, I wonder if having this would cause people to damage their cars by encouraging them to drive too fast on road surfaces their car can't handle. I've always assumed the bumpy-road tactile feedback was a useful "slow down!" indicator.
You'd do that only if you could afford fixing broken suspension. And that wouldn't be that cheap, I suppose. It increases comfort, yet does not prevent damaging suspension when used in such manner.
Actually.. The BMC Hydrolastic suspension did the same job very effectively.. And no electronics and complex control systems to service and go wrong.

I had a BMC Morris Nomad 1300 and its handling was superb

https://en.m.wikipedia.org/wiki/Hydrolastic

They were terrible to work on, you had to have the special magic tool to pump it all back up again
Well, try working on a new car with no special tools!
One of the demos showed the car "staying level" as it swerved aggressively. I wonder -- might it be reasonable to have it overcompensate? If a car could "lean into" a curve like a motorbike, it could help driver comfort by turning some horizontal acceleration into (passenger-pose relative) vertical acceleration.

Not sure about safety/grip implications.

I was thinking the same thing. It would be sort of like banking in an airplane, that effectively removes the lateral force of a turn.

The effect in a car wouldn't be as dramatic, but I could still imagine it being useful.

There was an attempt by British rail to develop a rail car that would lean into corners. How the car leans in a turn has a big effect on ride comfort.

How a car leans and how the suspension responds has a big effect on how the tires contact the road which has a large influence on how sure footed the car feels.

Historically there are trade offs between ride vs handling vs cost and 'branding'.

What interesting to me about electro-mechanical systems is the ability use highly non linear control algorithms to improve the ride and handling. There are some potential benefits beyond just ride comfort, reduction of road wear[1] and anti-washboard control on unimproved roads.

[1] Think of cars and trucks hammering a weak spot like a crack in a road surface.

Most roads already have lean built into their turns, no tilting suspension necessary. And since it's built into the road, it positively impacts traction
That's true for a lot of rural roads. Not so much for parking lots, four-way intersections, roundabouts etc. And in many cases the road camber is suboptimal (or goes the wrong way...)
Mercedes CL and S-class have had active suspension ('ABC') for about 20 years now with that feature, I believe Rolls Royce had it too. It's crazy expensive to repair if something breaks. Citroen pioneered it but is moving away from it. I've had a couple of DS's in the late 80's and after switching to another car (Austin Maxi) I once took a speedbump without slowing down and hit the roof with my head. The Citroen dealt with obstacles so well it made you forget they were there.
> It's crazy expensive to repair if something breaks.

I've heard that a lot but I don't believe that to be true. It's actually just a fairly simple hydraulics system with a pump, oil lines, and some valves, all conected to the shock absorbers. The pneumatic part consists of these gas filled, self contained orbs, that you easily and affordably can get serviced.

So I think this is probably a rumor maybe stemming from mechanics in generic car shops unfamiliar with these systems and unwilling to try to understand them working on them. Or using this rumor to charge people as much as they can get away with.

Anyhow, it's unfortunate because mechanical engineering wise these are sound systems (not mentioning the software controller issues in some models, that all modern cars seem to be plagued with) and they offer a ride comfort that's unparalleled. And all this using simple, cheap and widely applied and proven techniques (pneumatics/hydraulics).

Is this the same Bose that (primarily) works on audio technology?
If so, I wonder if their experience cancelling sounds waves is just similarly applied to what is ultimately also a problem in counteracting input movement with the inverse.
If it is the same Bose, it will only ever work on special staged roads (viz. Bose's insistence on only showing their products completely separately from everything else, lest you realize that you can get significantly better sound for 1/10th of the price).
The first thing that occurred to me after seeing the vids: this will hurt the taxpayer - the government morons will want to rebuild all speed bumps if this thing comes out.
This kind of tech is going to be a critical pick and shovel for driverless cars.
Suspension is a solved problem since the 60's. The issue is one of economics not of technology.