Model 3: steering by wire?

[jbcarioca]

Is that how modern airplanes actually work? My impression is that planes are actually similar in concept to what you see in newer cars in that there is "by-wire" braking and assisted steering but ultimately there are still mechanical linkages in case the assistive systems fail. In a plane, the hydraulic controls can fail and the pilot can still pull back on the yoke and control the flaps but it takes a lot of physical effort on a large jet.

The B787 is all electric, no hydraulics even. Since the original A320, then the B777 , neither has built an air plane with mechanical control surface linkages. A fair number of military aircraft would have been impossible to control with mechanical controls. A fair number of business aircraft also are FBW (fly by wire). There will probably be almost new designs with mechanical controls.
Five reasons:
1. They are lighter,
2. They are cheaper,
3. They are more reliable,
4. They are more adaptable and updatable,
5. Maintenance is much cheaper.

Cars should have steering by wire just as modern ICE do for engine controls. The same advantages as for aircraft would apply, with pretty serious cost savings.

 

[jbcarioca]

Reliability aside, there's one thing steer by wire doesn't do well that is intrinsic to mechanical steering: Feedback.

I can see this becoming less important as people with actual driving skill become less commonplace, but for now it's going to be an unwelcome change for many.

Steering feel is a Big Deal. If aircraft are an indication it is no problem. A Falcon (big business jet) still has all the feel of the old style. Steering feedback is not a difficult technological task, builder that care about drivers will provide it. Builders who value driver serenity will have zero feedback. Therefore, nothing much will change except for cheaper, more durable, lighter, simpler car systems.

 

[JER]

Aircraft control might not be a good analogy here.

Steering torque directly tells the driver how hard the contact patch is working and how far forward the stick/slide transition has moved as you approach breakaway. There isn't a directly equivalent feedback in aircraft control.

 

[jbcarioca]

Aircraft control might not be a good analogy here.

Steering torque directly tells the driver how hard the contact patch is working and how far forward the stick/slide transition has moved as you approach breakaway. There isn't a directly equivalent feedback in aircraft control.

There is an exact analogy. aileron and rudder controls need precise and direct feedback in order to assess, among other things, gust loading and airframe stress. A few aircraft types eliminate taht process, but most today provide the equivalent of manual system feedback.
In both aircraft and motor vehicles automated systems manage traction control and braking control.
Please don't kid yourself; in modern transportation devices we, as operators, manage systems delivering acoustic and physical feedback when these are sometimes designed to make us feel as though we are all doing it directly. The only question is: how much do we want to deceive ourselves that we are in direct control?

That does not remove the desirability for highly skilled operators, nor the pleasure we all might find from a well designed machine that lets us imagine we're in charge completely.

1. For the record I see a future market for full autonomy in aircraft and ground vehicles. In that mode all those nice feedback devices are irrelevant. 2. I also see a future market for aircraft and ground vehicles that are engineered to provide high visceral pleasure to operators.
Also for the record I want my car to be capable of doing 1 while doing 2 when I am operating it myself.
Ah, yes, I want a Tesla with fully functional level 5 autonomy, and with Maximum Plaid when I desire it, still cornering and responding pretty much as my P85D does right now.

 

[bxr140]

Steering feedback is not a difficult technological task...

That's right--and technological progress will continually improved DBW feel.

There's no reason to think tesla will underwhelm with DBW feel.

 

[JER]

Loading from dynamic pressure is not an exact analogy. Not even close.

 

[jbcarioca]

Loading from dynamic pressure is not an exact analogy. Not even close.

To be fair you're right. the aircraft situation is a couple orders of magnitude more difficult. Sorry I suggested otherwise. It's true that the car situation is nearly trivial in comparison. Thanks for pointing that out.

 

[JER]

You really are missing the point entirely.

 

[Mark C]

There's a day coming for DBW just as the day came for FBW. When I test drove a Model S, I did not ask how the steering wheel sent information to the steering rack assembly. I drove the car and liked what it did. It doesn't matter much to me if it's a mechanical linkage (that will eventually wear out) or an electronic device (that will eventually fail) as long as there is some way to let me know failure it is coming.

As a retired USAF fighter aircraft crew chief, I know there are parts on the aircraft that are required to be replaced at a certain interval, because testing shows where the safe end of operation is. Failure to replace it could mean a smoking hole or a pilot who goes swimming.

In the case of the Model 3, I would hope that you'd have some indication of the end of useful life, and when a critical component reaches it without the operator having replaced the component, the car would require replacement before it would be drivable again. Failure to replace it would leave you (perhaps / hopefully) an expensive V2G (Vehicle 2 Grid) backup battery.

That said, when I drive the Model 3 I hope to own, I have no intention of verifying DBW or not. That's just me.

 

[igotzzoom]

Steer by wire is a very bad idea. It is not needed and removes a crucial safety link. Do not put too much faith in electronics. There is ZERO benefit to it.

I don't know if I'd go so far as to say there's zero benefit, but as it stands today, it's more trouble than it's worth. As noted previously, federal law currently requires a redundant mechanical backup in the case of a system failure, effectively obviating any advantage from simplified manufacturing. Until the FMVSS is updated to allow for SBW without a mechanical backup, it just makes more sense to employ a conventional mechanical steering system.

 

[Topher]

With autopilot, your car is able to do the steering by itself. Why bother having a steering axis installed when everything could be done electronically?

At some point you need to transform electrical signals to physical movement. What is your design for that system, and why is it simpler than the current system?

Thank you kindly.

 

[jbcarioca]

I don't know if I'd go so far as to say there's zero benefit, but as it stands today, it's more trouble than it's worth. As noted previously, federal law currently requires a redundant mechanical backup in the case of a system failure, effectively obviating any advantage from simplified manufacturing. Until the FMVSS is updated to allow for SBW without a mechanical backup, it just makes more sense to employ a conventional mechanical steering system.

The NTSB will be strong for it due to the numerous safety benefits, which they've already documented with aircraft. Further the autonomous driving legislation they are advocating will specifically allow for such advances, as it must if technological advance is to continue. The mechanical systems are too heavy, too mechanically complex, so too expensive. BTW, the aircraft pilot furore over FBY still goes on, but the opponents are dying out (literally). In cars it already has become a fait-accompli, but cost has been the big impediment.

Given my posts. As an enthusiast my favorite airplane to fly was the Learjet 25, a Luddite of a business jet adapted from a fighter design. Glorious, but shitty accident record.
DBW will soon be happening and we'll need to adapt, just as we airplane drivers were forced to do.

 

[timk225]

The cause for things like autopilot, this idea of no mechanical steering link, and the whole NVH crusade of the last 20 years is sheer, all-out LAZINESS on the part of the consumer, who has become a wussified crybaby, driven by automotive advertising. God forbid we hear any engine noise, or feel a bump in the road, and now even steering the car is getting to be too much trouble to bother with!

Newer technology is not always better technology. Some things, the 1970s cars got right.

 

[alseTrick]

I think right now there is bimodal distribution of people interested in the model 3.

A) Those who wish they could buy a google car with no steering wheel, and are accepting the 3 as a poor substitute because it is (will be) too engaging.

And B) People who care about their immediate environment and want to maintain some connection to and dexterity with it. Visceral people. They are alive.

I disagree with this supposition. And to intimate people who want autonomous cars are dead? What's up with that?

I think most people don't fall into either of those groupings. I know I don't care about the performance of the 3 at all, but I also don't want a "google mobile".

 

[jbcarioca]

The cause for things like autopilot, this idea of no mechanical steering link, and the whole NVH crusade of the last 20 years is sheer, all-out LAZINESS on the part of the consumer, who has become a wussified crybaby, driven by automotive advertising. God forbid we hear any engine noise, or feel a bump in the road, and now even steering the car is getting to be too much trouble to bother with!

Newer technology is not always better technology. Some things, the 1970s cars got right.

Other than lousy fuel economy, poor handling and lower reliability what did they do then that was better?
For perspective I am a fan of old cars, from the 1903 Baker that was my first experience with BEV, to the Chrysler 300F owned by the Baker owner, to the 1978 Ferrari GT4 that I owned. None of the 1970's cars I experienced come anywhere near the middling level of 2016 cars. Just compare tests of the stars of the 1970's to anything average in 2016. Virtually everything is better now and inflation-adjusted costs have not risen.

FWIW, if it is the Sturm und Drang that you miss, my personal all time favorite is from 1966, a Shelby 427SC that I drove extensively back then because it's young owner lost his driving license so I drove him around. It was tremendously exciting but in 2016 terms it was a truly dangerous, lousy car. Nostalgia does cover all the warts and blemishes and adolescence made them invisible then too.

 

[dsvick]

.

I think right now there is bimodal distribution of people interested in the model 3.

Bimodal customer distributions are scary as they can destroy the essence of the product. Large car companies confuse the distribution network (dealers) with end customers. That bimodal focus makes most modern cars have no real essence. Maybe some styling or perfume. They are choosing to be commodities.

I'm sure Tesla has nothing to worry about as I'd be willing to bet that neither of those reasons are the main reason that very many people are getting the Model 3. I'd put more money on:
They want an electric car for the environment
They want one for the cost savings
They (US customers) want it because it is mostly American made
They want it for all the tech that is in it (including autopilot and autonomous driving)
They want it for the performance
They want it so they can show off to all their friends.

Any one, or combination, of the above is more likely than "they wanted a Google car".

 

[dsvick]

Newer technology is not always better technology.

Except when it is. Autopilot and self driving cars are not being developed because people are lazy, they are being developed because they are safer than cars driven by people, and because it a great way to advance a whole slew are various technologies.

NVH isn't about laziness either, it's about comfort. Wanting to be comfortable isn't being a wuss, it's a personal choice that most intelligent people decide is worth pursuing. If you'd rather not then you can go right ahead and drive around in a go cart with wooden wheels.

The same with steer by wire, just because some people don't want it so they can "feel the road" doesn't mean that everyone that does want it only wants so they don't have to feel the road.

 

[S3XY]

One thing that steer by wire could provide is adjustable road feel. If you think the default handling is too numb, just turn the dial to increase road feel.

 

[22522]

I wonder how much road feel is integrated with other sensory inputs. I am confident some DBW will be good. Some will not be.

 

[jbcarioca]

One thing that steer by wire could provide is adjustable road feel. If you think the default handling is too numb, just turn the dial to increase road feel.

Yes.
Another thing in aircraft is that cruise flight in FBW can be more efficient by adopting an unstable configuration that a human could not manage. One of the biggest fuel efficiency advantages in FBW is the fuel saving achieved by eliminating the need for positive stability ( it takes a long discourse to explain this, but basically a very aft center of gravity saves fuel, but severely reduces aircraft stability. FBW handles that. Some military designs are so radical that a direct human hand could not fly them at all.
Back to cars: DBW can permit traction control vastly better than what exists today by managing each wheel independently, dynamically adjusting suspension setting power, braking, etc. The surface has barely been scratched. None of the advances would be [possible with ICE because of the inherent lags and transmission delays inherent in those antiquated processes. No matter how good they get ICE cannot generate individual wheel behavior in nano-seconds or less to cope with changing conditions.

Ten years from now cars and trucks are likely to behave like nothing we've yet seen. Tesla will be leading if there is any luck at all.