Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Mitigating Speed Hump Discomfort With Air Suspension

This site may earn commission on affiliate links.
Would it not be a relatively easy add-on to an air suspension equipped S to detect and rapidly move the front and then back wheels to all but eliminate the effect of speed humps? A small sensor (or sensors) mounted under the front bumper could detect the bump and at a calculated time period later (according to the car's speed) quickly raise and lower the front - and then back - wheels to effectively eliminate the bump. Perhaps they can't move quickly enough? Anyone got any observations? MW
 
It takes two or three seconds (estimate: I haven't timed it) for the air suspension to raise or lower one end of the car a couple of inches. That's far too slow to be of any assistance with speed bumps.

But there is one thing you can do whether or not your car has air suspension: if you arrange it so only the left wheels or the right wheels actually go over the speed bump, you can comfortably take it at a much higher speed than if all the wheels pass over it. Many speed bumps end short of the curb to facilitate drainage; just be careful you don't rash your rims...
 
But there is one thing you can do whether or not your car has air suspension: if you arrange it so only the left wheels or the right wheels actually go over the speed bump, you can comfortably take it at a much higher speed than if all the wheels pass over it. Many speed bumps end short of the curb to facilitate drainage; just be careful you don't rash your rims...

Also nails and trash tend to collect by the curb, so be a bit careful when you use this tactic. Going slower of the speed bumps helps too.
 
Would it not be a relatively easy add-on to an air suspension equipped S to detect and rapidly move the front and then back wheels to all but eliminate the effect of speed humps? A small sensor (or sensors) mounted under the front bumper could detect the bump and at a calculated time period later (according to the car's speed) quickly raise and lower the front - and then back - wheels to effectively eliminate the bump. Perhaps they can't move quickly enough? Anyone got any observations? MW


Actually, I've pondered something similar for a few years now.

Most suspensions react to the road. They read speed, and movement of the axles/suspension.

What you are describing, and what I've always wanted to build, would be a proactive suspension. While technically, it's still reactive, it would use cameras, radar, magic, whatever. the point is, it would react to road issues BEFORE the wheel actually struck them.

You could 'see' a pot hole and soften the suspension quickly.

you could 'see' a raised surface and soften it

you could see a speed bump, and absorb it.

My system design also calls for a learning system.

Just like how I assume the AWD tesla works... it senses what it's doing, what the result is, and learns from it.

You don't pre-program "hey this is a pothole" into the camera and say "soften the suspension"

nope. not one line of code.

What you do is program, ok, this is occurring (pot hole), try (something) it will (stiffen suspension). Then record the results. Then try something else next time.

record results.

when results move in a favorable direction, keep doing that thing, until they do not. Then use that value.

ta-da, self learning suspension.

I assume the AWD S will get more efficient over time, as it records the driving habits of 1,000s of cars. Driver input: 45% gas pedal, record speed, record acceleration, record energy usage per motor.

adjust energy levels per motor, slightly, over time, over every acceleration, forever adjusting and millisecond learning.

eventually, you'll produce the most efficient acceleration pattern.

All without knowing or caring what the actual efficiencies are. Because the car just learned the most efficient way of getting up to speed.

Of course that all goes to hell when you mash the go pedal :)

This allows accounting for subtle differences in all machinery... when you make 100,000 of something... some of them will be different... so why should they all have the same programming?

Observe, react, record, improve.

Lucky for you guys I'm more John Connor than SkyNet :)
 
Actually, I've pondered something similar for a few years now.

Most suspensions react to the road. They read speed, and movement of the axles/suspension.

What you are describing, and what I've always wanted to build, would be a proactive suspension. While technically, it's still reactive, it would use cameras, radar, magic, whatever. the point is, it would react to road issues BEFORE the wheel actually struck them.

You could 'see' a pot hole and soften the suspension quickly.

you could 'see' a raised surface and soften it

you could see a speed bump, and absorb it.

My system design also calls for a learning system.

Just like how I assume the AWD tesla works... it senses what it's doing, what the result is, and learns from it.

You don't pre-program "hey this is a pothole" into the camera and say "soften the suspension"

nope. not one line of code.

What you do is program, ok, this is occurring (pot hole), try (something) it will (stiffen suspension). Then record the results. Then try something else next time.

record results.

when results move in a favorable direction, keep doing that thing, until they do not. Then use that value.

ta-da, self learning suspension.

I assume the AWD S will get more efficient over time, as it records the driving habits of 1,000s of cars. Driver input: 45% gas pedal, record speed, record acceleration, record energy usage per motor.

adjust energy levels per motor, slightly, over time, over every acceleration, forever adjusting and millisecond learning.

eventually, you'll produce the most efficient acceleration pattern.

All without knowing or caring what the actual efficiencies are. Because the car just learned the most efficient way of getting up to speed.

Of course that all goes to hell when you mash the go pedal :)

This allows accounting for subtle differences in all machinery... when you make 100,000 of something... some of them will be different... so why should they all have the same programming?

Observe, react, record, improve.

Lucky for you guys I'm more John Connor than SkyNet :)


Didn't BOSE experiment with such a suspension?

Bose Suspension System – Bose Automotive Systems Division

Never heard about it being built into cars you can buy.
 
Anyone heard of a Citroën DS? It could pick up one wheel off the ground for a change of tire (assuming proper balance) and was reportedly awesome at soaking up bumps. But I don't think it handled as we prefer these days and was pretty soft and mushy. Citroën DS - Wikipedia, the free encyclopedia

I had one. Not only did the suspension allow you to change the tires without a jack, if a tire went flat it picked the tire off the road to prevent further damage. You could actually drive on any three wheels. As far as the handling goes, they had a prototype that really gave it sport-type handling (more than it already was, which was very good for the day) but when Andre Citroën went broke, Michelin took over because they were the largest creditor. Michelin never did anything with Citroën for fear of alienating their other car manufacturer customers, so many items that were on the drawing board never made it to production.
 
I had a 1968 Austin America, my first new car, that had hydro-lastic suspension with 15/16 steel lines for and aft on each side connecting the rubber suspension bags. Actually that was the wrench size for the single flare nuts. If you hit a bump at speed the induced pressure increase from the front travelled to the rear piston which raised the rear of the car just in time to effectively smooth out the bump. No motors, pumps or valves. I believe it was unique and I never had issues with it, except for one slight leak - had to buy a pair of combo wrenches to fix. You could throw that car into corners with aplomb esp with those 4 tiny 12in wheels at the periphery.
--
 
The Active Body Control of the past 10 years in some Mercedes can do amazing things like adjust over 20 times a second, and keep the car completely flat on turns. Search YouTube for "Mercedes Rodeo" to see the "rodeo" diagnostic test, which looks exactly what it sounds like, and you can really see the degree and speed it is able to adjust! The problem is it's a complex hydraulic system pressurized to 3,000psi, expensive, heavy, and failure-prone.....Ask a Mercedes owner who's seen the dreaded ABC malfunction light and there's a good chance they were on the hook for a $5k repair bill! :scared:

I don't see Tesla adding the complexity, weight, and expense any time soon. The 2014 S550 did bring Magic Body Control, which adds the ability to scan the road ahead to preadjust as the road changes, but with drivers having had some real world time with it, word is it needs ideal conditions to work properly like fair weather and daylight, and that it's kinda gimmicky. But it does smooth out speed bumps!
 
...how bout just slowing down?

you know why those things exist, i hope? usually there's a school or playground nearby...

I think the idea was that even at a slow speed they are still uncomfortable.

On my street speed cushions are there because the typical speed was about 50 mph (30 mph zone) and no amount of enforcement could slow it down. Before I put up some stonework and significant vegetation, I often found tire marks in the yard where people had driven over the curb and on the side lawn. A house up the street from me had a car driven into the bedroom, fortunately no one was injured.