I'd love to think that there'd be less wiring, but I find that a bit tough to believe. I think that williamlweaver has it right: the whole system has to be redesigned from the bottom up if we're going to make such a radical shift as moving to a drive-by-wire Ethernet.
Perish the thought that a critical Real-Time system would be using a general purpose office OS such as Windows. Modern Real-Time data acquisition and control systems are extremely reliable. For example National Instruments LabVIEW RT (real time) is a very mature product and has embedded solutions and even a real-time hypervisor for running multiple RT instances in parallel. I don't think the BSOD is the fault of the Ethernet communication standard.
Makes perfect sense. Thanks for clarifying, Chuck. Luckily for me, I use a Mac so I've never encountered the dreaded "blue screen" <grin>, but I can imagine that the possibility of dealing with any kind of unknown or security breach is too risky especially for power train applications that have such a close correlation to driver safety. Question though: Were there specific advances around Ethernet that drove up its level of determinism?
Beth: If you've ever had a computer lock up on you or, worse, gotten the so-called BSOD (blue screen of death), then you can understand why automotive engineers were worried about Ethernet. They need determinism. If a message doesn't get to the brakes or steering or engine on time, they could have a serious safety problem, so they simply stayed away from Ethernet. They used CAN for powertrain and MOST for infotainment because of its bandwidth. But over the last few years, Ethernet has been recognized for greater determinism and MOST has lost some support because it's proprietary and because it's more expensive than Ethernet. So Ethernet has nudged its way into the picture, especially for applications such as video, which isn't safety-critical.
@RadioGuy - I was intentionally exaggerating to illustrate my point. While car companies continue to highlight performance of components under the hood, inside the transmission, or bolted onto the suspension it has not been until recently that median sedans included technology over and above power windows and door locks. With so many commuters spending multiple hours in their automobiles per day, it is difficult to appreciate the technology of Variable Valve Timing while sitting in traffic with only a speedometer and an FM radio to keep you company. If manufacturers are going to incorporate radical changes such as drive-by-wire Ethernet, the entire system will require a redesign or more appropriately, a new design from scratch.
The idea is that you would have LESS wiring with a standardized data network bus instead of "home run" wiring to each device data point. You just have to get to the nearest network switch, which might be a short distance.
>> -- all while the production vehicle is outfitted with a cassette tape deck
>> and a bicycle brake cable actuator for the fuel door.
I don't think I have seen cassette tape decks since around the 2000 model year, when they were replaced by CD players. Actually, my 2000 Mustang had both. The cassette decks were not nearly as reliable as CD players, due to the damage that moisture condensation tended to do to the tapes. In my experience, the CD players wear out after about 8-10 years. Replacing a 10-year old car stereo with an up-to-date aftermarket unit tended to be a nice, affordable upgrade to the vehicle. One that will not be possible with the modern OEM integrated system control/entertainment/navigation systems.
I think the "brake cable" activated fuel door is a perfectly serviceable solution. The alternatives would be a key lock on the door (clunky, especially in an era when the doors are keyless, proximity key actuated) or a solenoid operated mechanism, which I think would be quite a bit more expensive and less reliable.
I hate to be a wet blanket, but this sounds to me like yet more wiring that can go wrong. Automotive wiring is notoriously difficult to fix when some short or simple disconnection happens. This may just add to the "fun" for mechanics. Are there are statistics for failure rates?
Bill makes a good point about regulation. Another, somewhat related issue, is that the user life of a car is now well in excess of what it used to be. I believe people now keep their cars upwards of 9 years. So this means, as new technologies displace older ones, will be have issues about maintainability of legacy vehicles. OTOH, I was always able to get parts for my 1988 Camry. I finally had to get rid of it two years ago when the rust became a danger (the rear quarter panel was about to fall off, and maybe the trunk after that).
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.