Drive By Wire

The technology was first developed for aircraft (which had used servo-assisted hydraulic controls since the

1940s). Total "fly-by-wire" has been standard in airliners for more than a decade. Automobiles first used it in ABS braking and antiskid systems. Fail-safe systems seem simpler for automobiles than aircraft, because it does no (direct) harm simply to switch off all servo- controlled functions. Perhaps the OP could repost explaining what disasters he foresees.

phaeton wrote in news:7273a5a8-5b07-47b5-88e9- snipped-for-privacy@m36g2000hse.googlegroups.com:

I was under the impression it was being done as part of continued emissions-reduction strategy.

"Drive by wire" means any system where there is a completely electronic connection between the control and the physical mechanism. What you describe is an "Electronic throttle" which is one type of "Drive by Wire." "Electronic steering" would also be "Drive by Wire" but isn't being made.

It allows the ECU to actually control the throttle position directly at all times, not just at idle. With an automatic transmission, this could be integrated into the transmission control algorithm to allow the computer to override the driver for better mileage, for instance.

I think this is a horrible idea, personally, because I want to be in control of the vehicle completely. But then, I hate automatic transmissions as well, for the same reason.

As far as being disasterous if it malfunctions, it is. But mechanical throttle linkages fail too, sometimes in disasterous ways.

Basically, the fact that cars really haven't changed all THAT much since the Model A. They still get you from one place to another. The problem is that if people didn't keep buying new cars all the time, car manufacturers would go out of business. So they need to find something exciting and new to make you want a new car.

--scott

As I understand it, "Drive By Wire" means that there is a throttle position sensor which is read by an ECU or other electronic device which then changes the position(s) of throttle plate(s) in accordance without any mechanical link whatsoever between the gas pedal and the metering device.

This is showing up on a lot of cars these days, and auto rags (and some manufacturers) are hailing it like the second coming of Jesus. I can't see where it would or even could improve anything, but it sure could be disastrous if it malfunctioned.

Is there a real benefit, or is it all hype?

Thanks.

Throttle control MAY be OK this way, but the electronics wonkies are talking steering and brakes controlled the same way. I'm not having anything that has no direct mechanical connection between me and the road. "Power" steering and brakes will still function at some level if the more-sophisticated machinery fails. The geeks doing this are on it for the same reason a dog licks his balls: because he CAN. There's no practical advantage for the driver and passengers in the vehicle.

it can have advantages in building the vehicle. If EVERYTHING is controlled electronically then there is no longer a need to run cables and rods thru the firewall. And no longer do you need to arrange things so the brake booster/master cylinder is where the persons feet are. Same with steering, if you don't need to make the steering wheel connect to the steering box mechanically you can put stuff in more advantageous locations as well as making bolting things up a lot easier. Electric power steering should be more efficient since you don't have a pump running all the time with the drag that goes along with it. For brakes, the coming thing is Hybrid and that means electrical stuff at the wheels so you may as well get rid of all the hydraulics and just make the whole thing electrical.

I have the same reaction as most people, NO WAY do I want to be betting my life that a fuse won't blow and I have no steering or brakes!!! But I think they can make it safe since computers are so powerful now and it's so easy to provide redundancy that a failsafe strategy should be easily doable that is at least as safe statistically as current systems.

I just realized that my new work car, Crown Vic, has electronic throttle, no more cable. I can't tell the difference and didn't even realize if for two months. The only reason I noticed was that I had the cruise control on and then braked and it disengaged. When I pushed the resume button it did but my foot never felt the pedal pull away like a mechanical linkage would have. This car is a 2007 and the one before it was a 2005. The 2007 actually has better feeling throttle response with the electronic throttle.

It's hard to believe you can't sense the difference. With the cable gone, theres no longer any finesse to applying throttle. It may work fine in most situations, but the human reflex is far quicker than a car's computer in this respect. There's a definite adjustment curve you experience in changing to a car with "DBW".

Yes, but they are much simpler to repair on the side of the road if need be.

A friend of mine had this happen years ago on a trip -- broken throttle cable while driving on the interstate. He coasted onto the shoulder and fixed it with a shoelace. He was able to continue his trip and for that matter drove the car for several more years with the shoelace doing its job.

Sure there is. Traditional mechanical throttle linkages require a lot of space. Cable operated throttles require maintenance and still need more of a "straight shot" from the pedal to the linkage. Throttle by wire does not, which allows for less holes in the firewall.

MB had a brake by wire system:

my opinion of them: "meh"

Ray

Vettes have been DBW since 1997. Haven't heard much complaining about it.

Ray

Sheesh, I've been complaining about 'Vettes since about 1970.

--scott

Depends on the type of servo used. If it can be overridden by a spring fine. But if it is a servo that freezes in position when power is removed, that isn't nice.

Also, in aircraft, SOME could be overridden with redundant cables, others had the fly-by-wire system only. But the ones that were electronic only were at least triple redundant, others quad redundant, and with MBTFs in the millions of hours. I cannot see Detroit putting electronics with that kind of reliability into a normal passenger car. The electronics alone would cost far more than present cars. Yeah, you can hit the ignition off, but you lose power steering and brakes, although at any speed at all most cars aren't that hard to steer, and in a panic most people can get their leg muscles to push pretty hard. Biggest thing to me is the response delay to realize the system has failed and take action. Look at how many people crash after driving for a block or more after a mechanical throttle linkage sticks.

That's "throttle by wire." I wouldn't use the term "drive by wire" unless steering and braking were also purely electrical.

Throttle by wire has some serious advantages to the car designer. One of the biggest advantages is that it lets the computer close the throttle when the automatic transmission is shifting, which makes the shifts much smoother and drastically reduces the amount of heat energy that has to be dissipated in the transmission clutch packs. As automatics got smaller and smaller in the 90s, they also got more and more fragile. A Torqueflite A-727 from the 60s could survive a lifetime of full-throttle upshifts from a 440 putting out 500 ft-lb of torque, but a 90s 41TE had trouble surviving behind a 140-horse 3.0 wheeze-6 in a minivan.... but it was a far more efficient transmission because the internal parts were smaller and lighter. With throttle-by-wire, an unnecessary load is simply never applied to the automatic, so it can last without being over-built.

There are other advantages too- the computer can subtly (or not so subtly- see the downside paragraph below) modify the driver's throttle inputs to achieve the desired performance with lower emissions and higher fuel efficiency than the driver would otherwise achieve. For example, instead of snapping the throttle wide open from completely closed, it can be programmed to open smoothly as the engine begins to respond. This can be done well or badly, its all in the software....

Which leads to: the downside, for me, is that a lot of car companies have really crappy throttle response programmed into their throttle-by-wire cars, with a very noticeable delay between me pressing on the loud pedal and the car, well, getting loud! That's not always true- One of the best I ever drove was a 4.0L Chrysler Pacifica (of all things), where you never noticed that it was TBW except for the fact that wide-open upshifts sounded like a manual transmission being shifted very quickly and expertly. At the other extreme, I've had a Kia rental that I could practically count to 2 before the car would start moving after hitting the gas hard from a stop.

As for reliability and safety, the odds of a servomotor failing to close the throttle really aren't any greater than the probability of a throttle cable binding or a throttle spring breaking on an old style car.

As I said in another post, whether its "horrible" or not really depends on how well the control software is written. I've had TBW vehciles that respond superbly, and others that feel like I'm driving the Queen Mary moving the engine telegraph and signalling down to the black gang in the boiler room who then start turning some steam valves to add some power a few minutes later....

The last time brakes had a direct mechanical connnection with the pedal was way back in the early 1930's. When hydraulic brakes first came out I am certain that the same arguments were heard.

Ever had a mechanical steering or suspension component go out? I see them alongside the road. I see no greater risk with a welll-designed electrical system than with a hydro/mechanical system..

An incompressible fluid in rigid or braid-reinforced tubes IS a direct mechanical connection, which operates without any external power source whatsoever.

Many silly arguments were offered, but not the same ones as offered againstbrake-by wire. Face it- brake by wire *does* have more likely failure points than hydrualic brakes, which actually had *fewer* failure points than old link-and-bellcrank mechanical systems. I'm not saying that ultimately brake-by-wire won't be the way to go, but its just not quite soup yet.

Nothing is failure-PROOF. You also see cars along side the road with electrical failures, and there are a lot of others you don't see beside the road because an electrical problem kept them from making it out of the garage! Again, I'm not being a Luddite- I'm an electrical engineer. I'm just reiterating that there is a very, very high standard of reliability in existing brake systems, and its going to take a lot to match it.

Then I don't think you're looking closely enough. There really is *very* little that can fail in a hydraulic brake system, and the things that do fail usually give ample warning. But most importantly, its a self-contained driver-powered system that doesn't need an external power source from an alternator, battery, or engine to function and I consider that a pretty hard standard to replicate with servo systems.

Yes, brakes need to be fail-safe - obviously. However they should not just fail into the "applied" mode: the thought of the brakes coming on hard when I'm bombing down the motorway does not bear thinking about.

If there's a loss of electrical power (blown fuse, corroded wire) you need the brakes to come on (for safety) but to do so in a controlled manner so that when you get a warning of power failure you have chance to steer to the side of the road and come to rest.

Consider also the disruption that could be caused if a car's brakes come on, even in a gentle and controlled manner, when the car is in the middle of the road, and the brakes can't be temporarily released to drive/push the car out of the way.

So there need to be overrides in the system.

But for years, he walked funny.

Pray that the airbag system doesn't fail at the same time as the brakes. I would hope that electric brakes would be more reliable than the abs systems we have now.