Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 10:02 am
http://www.youtube.com/watch?v=lZ4PtafR ... r_embedded" onclick="window.open(this.href);return false;
ww for offensive language
ww for offensive language
Cal Club Autocross Forums
Hate to say it but guys got a point
Page 1 of 2
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 11:25 am
Why is he dressed like a Taliban?
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 11:27 am
Hilarious video that demonstrates how a properly functioning Prius works.
But what happens when the Prius malfunctions?
That shifter knob? Software controlled.
That power switch? Software controlled.
The brakes (via ABS unit)? Software controlled.
Now let's see how funny it is when a software glitch slams the throttle open, disables the brakes via the ABS unit, and ignores the shifter and power buttons.
But what happens when the Prius malfunctions?
That shifter knob? Software controlled.
That power switch? Software controlled.
The brakes (via ABS unit)? Software controlled.
Now let's see how funny it is when a software glitch slams the throttle open, disables the brakes via the ABS unit, and ignores the shifter and power buttons.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 11:52 am
I got a ride with a Chrysler engineer in one of the early SRT-8 prototypes. It had a big RED kill button on the center console. Why? I asked the guy. Oh, the software hasn't been approved for consumer release yet, so the lawyers won't let us drive these things without kill switches. Just in case there's a bug.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 1:05 pm
I've always been leery about drive by wire. Knowing what I know about the various malfunctions caused by malfunctions in the solid state circuitry in the HVAC business is enough to make me very suspicious about driving a car controlled by computers and servos.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 1:36 pm
That guy reminds me of Walter in The Big Lebowski.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 2:19 pm
Fly-by-wire exists in Airplanes and lot of devices and they do just fine.
If this is indeed a software issue, it is a rotten shame at this point in time.
1. Redundant paths for the transmission of the signals?
2. Sensor to detect the presense of sustained acceleration without any driver input?
3. Sensor to detect acceleration when deceleration is demanded?
Those things listed are no more complex than the steering angle sensor in a Covette or the oxygen sensor in any car.
If this is indeed a software issue, it is a rotten shame at this point in time.
1. Redundant paths for the transmission of the signals?
2. Sensor to detect the presense of sustained acceleration without any driver input?
3. Sensor to detect acceleration when deceleration is demanded?
Those things listed are no more complex than the steering angle sensor in a Covette or the oxygen sensor in any car.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 2:53 pm
http://jalopnik.com/5492199/exclusive-e ... -a-scammer" onclick="window.open(this.href);return false;
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 2:58 pm
Yep, all that's fine until a catastrophic failure takes place. I can land a plane if a cable break by neutral control surfaces by opening and closing doors to steer, using trim to control an elevator, etc. But it's all contingent on control surfaces finding a neutral position. I can't see a fail safe way to insure that happening with a servo attached. Have you ever heard the saying "hope for the best, prepare for the worst".Theo Osifeso wrote:Fly-by-wire exists in Airplanes and lot of devices and they do just fine.
If this is indeed a software issue, it is a rotten shame at this point in time.
1. Redundant paths for the transmission of the signals?
2. Sensor to detect the presense of sustained acceleration without any driver input?
3. Sensor to detect acceleration when deceleration is demanded?
Those things listed are no more complex than the steering angle sensor in a Covette or the oxygen sensor in any car.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 3:13 pm
Opening the door on a 777 to steer, YIKES!George Schilling wrote:Yep, all that's fine until a catastrophic failure takes place. I can land a plane if a cable break by neutral control surfaces by opening and closing doors to steer, using trim to control an elevator, etc. But it's all contingent on control surfaces finding a neutral position. I can't see a fail safe way to insure that happening with a servo attached. Have you ever heard the saying "hope for the best, prepare for the worst".Theo Osifeso wrote:Fly-by-wire exists in Airplanes and lot of devices and they do just fine.
If this is indeed a software issue, it is a rotten shame at this point in time.
1. Redundant paths for the transmission of the signals?
2. Sensor to detect the presense of sustained acceleration without any driver input?
3. Sensor to detect acceleration when deceleration is demanded?
Those things listed are no more complex than the steering angle sensor in a Covette or the oxygen sensor in any car.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 3:25 pm
You put a spring in place so that a loss of power (through malfunction or the user hitting a kill switch) will create a closed-throttle condition.George Schilling wrote:But it's all contingent on control surfaces finding a neutral position.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 3:34 pm
It's difficult to have anything 100% fail safe, but I really like the idea of stuff being connected mechanically. Call me silly, I just feel more secure.Will Kalman wrote:You put a spring in place so that a loss of power (through malfunction or the user hitting a kill switch) will create a closed-throttle condition.George Schilling wrote:But it's all contingent on control surfaces finding a neutral position.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 3:36 pm
It's difficult to have anything 100% fail safe, but I really like the idea of stuff being connected mechanically. Call me silly, I just feel more secure. Not to mention I would think one could feel trouble brewing easier when connected mechanically.George Schilling wrote:Will Kalman wrote:You put a spring in place so that a loss of power (through malfunction or the user hitting a kill switch) will create a closed-throttle condition.George Schilling wrote:But it's all contingent on control surfaces finding a neutral position.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 3:38 pm
None of those would help. #1 isn't a software issue anyways.Theo Osifeso wrote:If this is indeed a software issue, it is a rotten shame at this point in time.
1. Redundant paths for the transmission of the signals?
2. Sensor to detect the presense of sustained acceleration without any driver input?
3. Sensor to detect acceleration when deceleration is demanded?
The issue with software would be a crash or a hang. Let's say in normal use, the cpu receives a throttle-pedal position number of 0-255 and outputs a throttle position command to the servo controller of 0-255. Now we're truckin' along and the computer glitches out and hangs but in it's dying breath, it outputs an erroneous 255 to the throttle servo controller and maybe throws a command to the ABS controller that we're in a skid so the ABS effectively disables the brakes. And since the computer is hung, it's not reading the shifter or power buttons. There is no electrical hardware you can throw at that problem that will fix it.
Now, most of these systems will have what's called a "watchdog timer" that expects to get a signal or data message from the computer at regular intervals. Now, when the computer hangs, it will stop sending the messages/signal and the watchdog will do a hard reset on the computer. Then, in almost all cases, it will begin operating normally, at least until the specific circumstance occurs again. It's safe to assume that the circumstances are very rare to cause this problem as all these cars have millions upon millions of miles on them.
I hear Woz has claimed that it's a software problem and that he can replicate it. Last word is that Toyota is ignoring him but I'd sure like to see a public demonstration.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 3:43 pm
Henry Ford distrusted hydraulic brakes. He tried to save us all and we ignored him!George Schilling wrote:It's difficult to have anything 100% fail safe, but I really like the idea of stuff being connected mechanically. Call me silly, I just feel more secure. Not to mention I would think one could feel trouble brewing easier when connected mechanically.
But I don't have to tell you that, you were there }:) ;)
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 4:01 pm
...or on a B2. Thankfully all 4 computers will have to fail simultaneously at which point you better ejectAaron Goldsmith wrote:Opening the door on a 777 to steer, YIKES!George Schilling wrote:Yep, all that's fine until a catastrophic failure takes place. I can land a plane if a cable break by neutral control surfaces by opening and closing doors to steer, using trim to control an elevator, etc. But it's all contingent on control surfaces finding a neutral position. I can't see a fail safe way to insure that happening with a servo attached. Have you ever heard the saying "hope for the best, prepare for the worst".Theo Osifeso wrote:Fly-by-wire exists in Airplanes and lot of devices and they do just fine.
If this is indeed a software issue, it is a rotten shame at this point in time.
1. Redundant paths for the transmission of the signals?
2. Sensor to detect the presense of sustained acceleration without any driver input?
3. Sensor to detect acceleration when deceleration is demanded?
Those things listed are no more complex than the steering angle sensor in a Covette or the oxygen sensor in any car.
"hope for the best, prepare for the worst".
Actually that's a big part of my job. I can make your system as fail safe as you are willing to pay. I have multiple paths between my data centers, in case crap happens. A bank will never tell a customer that they forgot how much the customer owes the bank, bonus depends on it
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 4:08 pm
Purely mechanical systems have failures too. When I was 16 I had a minor fender bender when the bracket holding my double throttle return springs vibrated loose. Probably caused by the solid motor mounts 
I turned the car off pretty quickly, but it was a car (the barracuda) that could accelerate 0-60 in 3.5 seconds. I was approaching the end of a cul-de-sac, and slowing down from probably 70mph when a scared Postman trotted out into the middle of the street then froze like a deer in headlights. I locked up the brakes (damn unmodulatable power brakes....) and the heavy crown of the road took my right front bumper corner into a parked car maybe 20ft before I would have stopped, and maybe 40 feet short of the still frozen postman. Did nearly no damage to the barracuda. Totaled the old beatup mini truck.
The postman is probably still standing there.
I turned the car off pretty quickly, but it was a car (the barracuda) that could accelerate 0-60 in 3.5 seconds. I was approaching the end of a cul-de-sac, and slowing down from probably 70mph when a scared Postman trotted out into the middle of the street then froze like a deer in headlights. I locked up the brakes (damn unmodulatable power brakes....) and the heavy crown of the road took my right front bumper corner into a parked car maybe 20ft before I would have stopped, and maybe 40 feet short of the still frozen postman. Did nearly no damage to the barracuda. Totaled the old beatup mini truck.
The postman is probably still standing there.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 4:24 pm
All of those will help and are part of the software system. A software system is only as good as the inputs it uses. 3 or more computers in different parts of the car on 3 different wiring paths, with a requirment that at least two have to be in agreement or shut down the car gracefully.Will Kalman wrote:None of those would help. #1 isn't a software issue anyways.Theo Osifeso wrote:If this is indeed a software issue, it is a rotten shame at this point in time.
1. Redundant paths for the transmission of the signals?
2. Sensor to detect the presense of sustained acceleration without any driver input?
3. Sensor to detect acceleration when deceleration is demanded?
I agree that this is most likely not a software issue but the complaints can be easily dismissied if redundancy is built in, maybe it is , I don't know.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 5:24 pm
Lot's of good times, Henry and me.Will Kalman wrote:Henry Ford distrusted hydraulic brakes. He tried to save us all and we ignored him!George Schilling wrote:It's difficult to have anything 100% fail safe, but I really like the idea of stuff being connected mechanically. Call me silly, I just feel more secure. Not to mention I would think one could feel trouble brewing easier when connected mechanically.
But I don't have to tell you that, you were there }:) ;)
Crap, you think I'll old now, wait til tomorrow.
Re: Hate to say it but guys got a point
Posted: Fri Mar 12, 2010 5:55 pm
Additional sensors, wiring paths, and more computers are hardware solutions even if they have software behind them. But you are right that a redundant monitoring and control system is a good idea.Theo Osifeso wrote:All of those will help and are part of the software system. A software system is only as good as the inputs it uses. 3 or more computers in different parts of the car on 3 different wiring paths, with a requirment that at least two have to be in agreement or shut down the car gracefully.Will Kalman wrote:None of those would help. #1 isn't a software issue anyways.Theo Osifeso wrote:If this is indeed a software issue, it is a rotten shame at this point in time.
1. Redundant paths for the transmission of the signals?
2. Sensor to detect the presense of sustained acceleration without any driver input?
3. Sensor to detect acceleration when deceleration is demanded?
I agree that this is most likely not a software issue but the complaints can be easily dismissied if redundancy is built in, maybe it is , I don't know.
Re: Hate to say it but guys got a point
Posted: Sat Mar 13, 2010 5:44 pm
http://gizmodo.com/5462205/steve-woznia ... s-problems" onclick="window.open(this.href);return false;Will Kalman wrote:...
I hear Woz has claimed that it's a software problem and that he can replicate it. Last word is that Toyota is ignoring him but I'd sure like to see a public demonstration.
What he is complaining about is issues with the cruise control, and to me it sounds a bit questionable as detailed in this user comment:
Judging by Woz's description he posted on Slashdot (quoted below) it
sounds like he merely doesn't know how to use the cruise control
properly.
Basically he's on a 85mph highway in a Prius, so he nudges up the
speed on the cruise control.
Thing is, the Prius isn't the most powerful car around, so it doesn't
build up speed very quickly when you are traveling 80+ mph.
But Woz is nudging that lever "over and over" again, and then he's
pressing it "10 times or more" because the Prius isn't building up
speed fast enough relative to his expectations. So eventually he hits
97 mph because you've press the cruise control lever enough times the
Prius is going to respond with WOT (wide open throttle) because it
thinks you wants to go really fast.
Its like those idiots that keep clicking the mouse over and over
because their computer or internet is slow, and suddenly find out they
purchased 15 copies of Transformers 2. That's Woz.
Quote Woz:
"Let's say that I'm in some place where the speed 85 mph is legal. I
can nudge my cruise control speed lever and my speed barely goes up,
say from 80 to 81.I nudge at again and again, up to 83. Then I nudge
it again and the car takes off, no speed limit. Nudging the cruise
speed control lever down has no effect until I've done it about 10
times or more. By then my Prius is doing 97. It's scary because it's
so wrong and so out of your normal control. I tested this over and
over the night I observed it."
Re: Hate to say it but guys got a point
Posted: Sat Mar 13, 2010 11:04 pm
Saw this posted on Facebook....good article. outlines the media frenzy pretty well IMO.
http://www.businessweek.com/lifestyle/c ... 403524.htm" onclick="window.open(this.href);return false;
http://www.businessweek.com/lifestyle/c ... 403524.htm" onclick="window.open(this.href);return false;
Re: Hate to say it but guys got a point
Posted: Sun Mar 14, 2010 12:10 am
You get to see the mighty full force of the Prius giving everything it's got, as it slowly comes to a stop from its dragging rear tires.Will Kalman wrote:Hilarious video that demonstrates how a properly functioning Prius works.
But what happens when the Prius malfunctions?
That shifter knob? Software controlled.
That power switch? Software controlled.
The brakes (via ABS unit)? Software controlled.
Now let's see how funny it is when a software glitch slams the throttle open, disables the brakes via the ABS unit, and ignores the shifter and power buttons.
What's the E stand for in E-brake again?
Re: Hate to say it but guys got a point
Posted: Sun Mar 14, 2010 4:59 am
They should also bring suicide doors back. They could remake that one James Dean movie w/ a Prius }:)Jason Rhoades wrote:You get to see the mighty full force of the Prius giving everything it's got, as it slowly comes to a stop from its dragging rear tires.
What's the E stand for in E-brake again?
Re: Hate to say it but guys got a point
Posted: Sun Mar 14, 2010 11:20 am
Good and enlightening read. Thanks for posting.Dan Shaw wrote:Saw this posted on Facebook....good article. outlines the media frenzy pretty well IMO.
http://www.businessweek.com/lifestyle/c ... 403524.htm" onclick="window.open(this.href);return false;
Too bad Toyota doesn't make a car I want. Looks like now would be a great time to buy one.