Bill Schweber's original question was why hasn't the steering wheel gone away? Well it's pretty darn clear, in this blog alone, why is hasn't, people are not mentally ready to turn loose of that control, just that simple. Most comments center around something going wrong and causing an accident. Any engineer worth their salt should know there are ways around ALL of the design related issues listed here.
Cars applying brakes if they sense an imminent collision were unheard of a few years ago but they are on their way to becoming standard. There are brake by wire (BBW) cars and in fact some of the naysayers posting here may have driven one without even knowing it. DBW is just the next step.
Will there be an occasional failure resulting in a crash and possible death? I can guarantee it. It has been and will continue to be the risk of being transported by something weighting thousands of pounds and going much faster than you can walk or run.
Posters talk about single point failures that could occur without warning causing loss of control, think tires! Try a having a blow out while negotiating a decreasing radius turn (like an off ramp) and see how much control you have. Funny thing about control, you think you have it until you don't.
I drive a '71 VW bus not because I'm afraid of technology but because it's fun, cheap and I have been known to fix it in a parking lot with a swiss army knife. So I'm very familiar with old vs new technologies. Bottom line, with cars generally getting safer each year I would happily drive a properly engineered DBW/BBW vehicle and have more real and less illusionary safety and control.
There is an intrinsic fault with steer-by-wire, which is that it uses a computer in the loop, at least in almost all of the versions that I have seen recently, which means that there will be software that will not be able to handle unusual conditions. The software will be set up to protect a 16 year old beginning driver on city streets. That alone is a show-stopper. The FBW systems on aircraft are multiply redundant and usually set up to work with a very well trained pilot. Besides that, they are far more carefully maintained than even most race cars. One final thing is that we all see a wide spread of failures in our current vehicles. Who would wish to add steering to that list? As for the assertion that vehicle electronics is becoming more reliable, it is clear that the individual is in sales of vehicle electronics.
True indeed. And to add to your point about failure mode, this failure was sudden, but far from catastrophic in the overall system impact. I just felt the frame drop a bit and suddenly get squishy and bouncy feeling.
With two handed driving, the driver's input is much like a balanced line so that anything that affects both inputs (hands) like a sneeze has minimal to no effect on the direction of the vehicle. Would a single joystick react the same or would two joysticks be required? I'm certain DBW could mimic the wheel functionality but it may raise the complexity of the joystick system.
As for drawing parallels to the joystick in aircraft, how much flying is done manually with a joystick and not on autopilot? Would a better comparison be operating construction equipment or lift trucks with joystick controls?
I recall that a project was published in magazine (Popular Science?) in the 1960's or maybe early 1970's that changed the steering wheel system. The dash mounted wheel was removed and wheels were installed beside the driver's seat on both sides. The added wheels were parallel to the tires. The driver turned the car by turning one wheel forward and the other wheel back. I don't recall any more about the project other than is was not suitable for bench seats. Maybe a reader recalls the article or can find it.
You don't find carbon fiber in critical systems (i.e. steering, suspension in particular) because there isn't a ductile mode failure, just sudden collapse. Sounds a lot like the silicon failure mode to me.
Fe all the way. I'm fine with Cu/Si as supplements.
By the way, as I understand it the most strenuous environments for products are: Aerospace due to extreme environment, but mitigated by expert operators and money available. Medical due to biocompatibility and usage requirements but mitigated by financial availability. And last, automotive due to environment and usage but exacerbated by inexpert operation and commodity pricing.
The trouble with this is that a yearly inspection will not catch a silicon junction that is about to fail. Unlike mechanical wear (such as ball joints) silicon is packaged where it cannot be seen, and a transistor will fail instantaneously and without warning.
Analogy: Predict when a light bulb will fail.
I think the risk of a bad driver is far higher than that of faulty electronics. And Toyota is a good example for that.
Frankly speaking any electronics that controls the average sloppy uninterested driver will increase safety and not decrease it.
And it doesnt't matter if it is a joystick or an electric steering wheel. Both can be easily pushed in the wrong direction if drivers grab a cup of coffee or turn to their kids while traveling at 70 mph. The trick is that electronics would be able to detect this and prevent the accident.
In a car the driver may move. Vibrations, inertia... are transmitted to the driver. And the driver "secures" himself/herself by "catching" the steering wheel. That is, the vibrations... into the steering wheel do not affect the steering functionality (at least in a "normal" case).
It is not this case with a joystick. You will see, whenever a joystick is used, that either the driver suffers no vibrations or inertia, or the driver is securely "fixed" to the seat. If we accept a more "secure fixing" to our seats, then a joystick could start to be a possibility.
Would it be correct to presume that your BCM is still under warranty and cannot be replaced unless the shop can return a proven bad unit to the manufacturer? In this case a little bit of assistance could help, such as current-limited reverse polarity applied to the 12 volt input (with the unit out of the vehicle of course). Do not apply 120VAC since the damage would be too obvious.
Been there where a shop that will quickly "shotgun" out-of-warranty at the customer's expense will not repair intermittent problems under warranty.
If not under warranty try resoldering all components with a small tipped soldering iron designed for surface mount work. Do not bridge solder across fine pitch IC pins. I had to do this to a Dodge Neon dashboard pcb, several solder joints had cracked. (That's one of many reasons I never buy Dodge/Chrysler any more).
Now imagine this scenario as applied to an intermittent steering joystick: "Sorry Sir, we cannot replace your intermittent Steering Control Module until it fails permanently and you actually run head-on into a truck."
NASA's Orion Flight Software Production Systems Manager Darrel G. Raines joins Planet Analog Editor Steve Taranovich and Embedded.com Editor Max Maxfield to talk about embedded flight software used on the Mars on EE Times Radio. Live radio show and live chat. Get your questions ready.