Author Topic: NASA will help probe Toyota accelerators  (Read 9362 times)

Offline TheFallen

  • Full Member
  • ****
  • Posts: 800
  • Liked: 77
  • Likes Given: 159
NASA will help probe Toyota accelerators
« on: 03/30/2010 05:11 am »
I don't know whether it's awesome or disconcerting that an automaker needs the help of a space agency to figure out what's wrong with its cars.  Oh well

http://www.msnbc.msn.com/id/36092407/ns/business-autos

Offline Danny Dot

  • Rocket Scientist, NOT Retired
  • Senior Member
  • *****
  • Posts: 2792
  • Houston, Texas
  • Liked: 17
  • Likes Given: 1
Re: NASA will help probe Toyota accelerators
« Reply #1 on: 03/31/2010 11:04 pm »
I think it is awesome.

Danny Deger
Danny Deger

Offline Chris Bergin

Re: NASA will help probe Toyota accelerators
« Reply #2 on: 03/31/2010 11:08 pm »
Yep, awesome. Get the best engineers on the job.
Support NSF via L2 -- Help improve NSF -- Site Rules/Feedback/Updates
**Not a L2 member? Whitelist this forum in your adblocker to support the site and ensure full functionality.**

Offline Namechange User

  • Elite Veteran
  • Senior Member
  • *****
  • Posts: 7301
  • Liked: 0
  • Likes Given: 0
Re: NASA will help probe Toyota accelerators
« Reply #3 on: 03/31/2010 11:18 pm »
Well, NASA doesn't have a roll in space anymore so why not fix cars.

GM is building robots for space

And in further stunning news about the shake-up of various responsibilities NOAA will be assuming the roles of the DoD
« Last Edit: 03/31/2010 11:40 pm by OV-106 »
Enjoying viewing the forum a little better now by filtering certain users.

Offline robertross

  • Canadian Member
  • Senior Member
  • *****
  • Posts: 17939
  • Westphal, Nova Scotia
  • Liked: 659
  • Likes Given: 7692
Re: NASA will help probe Toyota accelerators
« Reply #4 on: 03/31/2010 11:55 pm »
Well, NASA doesn't have a roll in space anymore so why not fix cars.

GM is building robots for space

And in further stunning news about the shake-up of various responsibilities NOAA will be assuming the roles of the DoD

LOL...nice

Offline Nascent Ascent

  • Full Member
  • ****
  • Posts: 739
  • Liked: 124
  • Likes Given: 106
Re: NASA will help probe Toyota accelerators
« Reply #5 on: 04/01/2010 12:14 am »
Since the problems are with Toyota cars shouldn't JAXA be tasked with this effort?

« Last Edit: 04/01/2010 12:15 am by Nascent Ascent »

Offline TheFallen

  • Full Member
  • ****
  • Posts: 800
  • Liked: 77
  • Likes Given: 159
Re: NASA will help probe Toyota accelerators
« Reply #6 on: 04/03/2010 03:28 am »
Since the problems are with Toyota cars shouldn't JAXA be tasked with this effort?



Good point!  But I do believe it is us Yankees who are the only ones raising heck about Japan's automobiles.

That...and maybe NASA might learn a thing or two about preventing sudden acceleration by Curiosity and other Mars rovers from this investigation.  I kid. ;D
« Last Edit: 04/03/2010 03:31 am by TheFallen »

Offline Patchouli

  • Senior Member
  • *****
  • Posts: 4490
  • Liked: 253
  • Likes Given: 457
Re: NASA will help probe Toyota accelerators
« Reply #7 on: 04/03/2010 04:45 am »
Jokes aside it's because the sudden acceleration issues are believed to be partly caused by cosmic rays.

Toyota has a history of trying to use the most cutting edge electronics in cars and they where using smaller process ICs then the other manufactures which are more susceptible to radiation issues.

One interesting thing the electronic throttle's benefit is mostly on making hybrid drives more seemless.
It is pretty much pointless complexity on a non hybrid vehicle.

Some stuff was just stupidity on their part the single fault layer on the throttle control ,the 3 second delay in shut down , removing the mechanical links in the shifter for selecting neutral, and not having the neutral position of the shifter disable power to the hybrid drive are just bad engineering.
The last one could literally be implemented with just a switch ,a contactor and some protection diodes.

The person who made those decisions may have not been experienced in vehicles but got the position due to credentials in other sectors or due to having a degree from a prestigious university.

I'm not familiar with Toyota's ECMs but GM and Ford ECMs generally have a second cpu that has it's own firmware etc that will take over if the primary controller goes into a fault mode.
They also still have cables and linkages on the transmissions and a mechanically operated slider valve that cuts hydraulic pressure to the rest of the valve body even though they have been computer controlled since 1990.
Same goes for all the unaffected Toyota models.

This is likely why there have not been many run away incidents with them.
Hopefully this will result in new rules for robustness of vehicle electronics and systems for all manufactures.
« Last Edit: 04/03/2010 05:21 am by Patchouli »

Offline Jorge

  • Senior Member
  • *****
  • Posts: 6404
  • Liked: 529
  • Likes Given: 67
Re: NASA will help probe Toyota accelerators
« Reply #8 on: 04/03/2010 05:04 am »
Jokes aside it's because the sudden acceleration issues are believed to be partly caused by cosmic rays.

Cosmic rays which mysteriously seem to strike disproportionately at Toyotas driven by persons over age 60. Nope. It's driver error.
JRF

Offline spinkao

  • Member
  • Posts: 21
  • Prague, Czech Republic, EU
    • The RAMA UAV system
  • Liked: 0
  • Likes Given: 0
Re: NASA will help probe Toyota accelerators
« Reply #9 on: 04/03/2010 06:48 am »
Since the problems are with Toyota cars shouldn't JAXA be tasked with this effort?

Since the "unintended acceleration" seems to occur in the US only, it is logical that a US agency steps in. JAXA has better things to do - like flying in space, you know ;).

Well, I enjoy living on the edge with my dangerous Prius, "The Widowmaker". I enjoy the heck outta pulsing-and-gliding in the slow lane with death as my co-pilot 8)! It makes you *really* feel alive and savor every moment, you know ;).

There are three things in the world I don't believe in, lest somebody convincingly proves me wrong: The Loch Ness monster, Ogopogo and runaway Priuses.
Any sufficiently advanced technology is indistinguishable from magic. - Arthur Charles Clarke

Offline spinkao

  • Member
  • Posts: 21
  • Prague, Czech Republic, EU
    • The RAMA UAV system
  • Liked: 0
  • Likes Given: 0
Re: NASA will help probe Toyota accelerators
« Reply #10 on: 04/03/2010 07:46 am »
Patchouli, please, don't say the crap of the things you know nothing about. I am a happy Prius owner and I know the car well, I have a Ph.D. in embedded systems, I design control systems for a living and I worked as a car mechanic for many years during my college studies. I read the Gilbert report and the related documents that were released, I also have detailed service manuals for my Prius and I did some experiments with the car myself, I disassembled the throttle pedal and the ECU in question and I am as familiar with the topic as anyone outside of Toyota could be.

Jokes aside it's because the sudden acceleration issues are believed to be partly caused by cosmic rays.

Do you have any relevant sources to support that claim?

Toyota has a history of trying to use the most cutting edge electronics in cars and they where using smaller process ICs then the other manufactures which are more susceptible to radiation issues.

Again, I'd love to see where is this information coming from.

One interesting thing the electronic throttle's benefit is mostly on making hybrid drives more seemless.
It is pretty much pointless complexity on a non hybrid vehicle.

Wrong. Please see how the Toyota HSD system works. It *cannot* work without the electronic throttle, it is an essential building block of the system. And it is not that complex for that matter.

Some stuff was just stupidity on their part the single fault layer on the throttle control

There are *two* sensors in the go-pedal (for redundancy), the signals are routed to the ECU using separate wirings. These are Hall sensors, so contactless measurement is used, therefore the sensors are not subject to any wear. The only "single point of failure" is that both signals are routed to a single MCU in the ECU, which is pretty normal. But even if that MCU fails, the car is not going anywhere. It would require a very special kind of failure to cause unintended acceleration in hardware level, causing *both* signal paths to be grounded (as is clearly shown in the Gilbert report).

This is not an airliner; the car would cost a fortune if everything was made completely redundant. Please show me a car with better pedal position sensing and throttle control design. I've seen *many* cars from inside and I can tell you that I've seen a lot worse systems than this. I don't know any one with complete redundancy in throttle control.

the 3 second delay in shut down

This is just to prevent unintentional shutdown of the car if you just happen to tap that button (not that I can imagine how this might happen, due to it's location. You can still select neutral or stomp on the brakes to stop the car in case of "unintended acceleration". Contrary to popular belief, the car *has* the brake override system (so if you stomp on the brakes with the gas pedal floored, the engine would shut down). If you depress the brakes only slightly, they are fighting the engine and the car still goes, but if you depress the pedal to more than 50% of its travel, the engine disengages instantly.

I know it, I tried that myself in my own car. It works exactly as advertised. And even if it doesn't, the brakes are more than capable to defeat the engine pretty quickly if you stomp on them and bring the car to a rapid stillstand. And yes, there *is* a direct mechanical (hydraulic) link between the brake pedal and the brakes. The only way to destroy the brakes is if you brake only slightly with the car running, overheating them (which is what Sikes did for that matter).

removing the mechanical links in the shifter for selecting neutral

Again, you are only showing your ignorance. There is no true neutral in a Prius, since there is not a gearbox at all. Please see how the planetary gear power split device works (for example here: http://prius.ecrostech.com/original/Understanding/WhatsGoingOnAsIDrive.htm) - the gears are meshed all the time. The neutral is just an electrical state, when both MG1 and MG2 are not powered nor generating power. So there simply could not be any "mechanical link".

In other Toyotas with classic drivetrain, there is a mechanical link to select a neutral.

and not having the neutral position of the shifter disable power to the hybrid drive are just bad engineering.

This is complete BS. You *can* select neutal all the time, even with the gas pedal floored. You can do it two ways: either you select "N" with the shifter and hold it there for 1s, or you tap the "P" (Park) button and the neutral is engaged immediately (if you press the "P" button while driving, it selects neutral, not park - this is a safety feature). If you do that while stomping on the gas, the engine would shut off automatically. Again, I *tried* that with my *own* car and I can confirm that it works as I say.
Any sufficiently advanced technology is indistinguishable from magic. - Arthur Charles Clarke

Offline HIPAR

  • Member
  • Full Member
  • ****
  • Posts: 585
  • NE Pa (USA)
  • Liked: 1
  • Likes Given: 0
Re: NASA will help probe Toyota accelerators
« Reply #11 on: 04/03/2010 12:45 pm »
For those knowledgeable or interested about computerized control systems, there's an explanation of the Toyota electronic throttle here:

http://www.insideline.com/toyota/toyota-throttle-control-and-electromagnetic-interference-testing-presentation.html

What is wrong with the system concept?

Off course, the presentation doesn't get into the nitty-gritty details of how the system is implemented.  Ground loops,  marginally stable signal conditioning amplifiers, differences in CPU's from alternate sources that cause logic faults not detected by test vector sets (i.e. the Intel numerical processor problem that occurred after a specific sequence of logical events).  Then, what about quality control issues with other components or manufacturing processes.  Or what about interconnection issues as connectors are a constant source of irritation with electronics.

Why not let NASA/DoD look into this?  They have a wealth of experience with ultra-reliable computing systems that must function within harsh environments.  The automotive operational environment is about the harshest I can think off for a commercial/consumer product.

---  CHAS

Offline cd-slam

  • Full Member
  • ****
  • Posts: 610
  • Singapore
  • Liked: 18
  • Likes Given: 315
Re: NASA will help probe Toyota accelerators
« Reply #12 on: 04/03/2010 03:57 pm »
I thought this sort of investigation is exactly what NTSB is supposed to be doing?

Offline vt_hokie

  • Senior Member
  • *****
  • Posts: 3054
  • Hazlet, NJ
  • Liked: 118
  • Likes Given: 436
Re: NASA will help probe Toyota accelerators
« Reply #13 on: 04/03/2010 05:49 pm »


Jokes aside it's because the sudden acceleration issues are believed to be partly caused by cosmic rays.

Do you have any relevant sources to support that claim?


Evidence to support the claim that it's being considered at least:

http://cosmiclog.msnbc.msn.com/archive/2010/04/02/2255979.aspx

Offline kevin-rf

  • Elite Veteran
  • Senior Member
  • *****
  • Posts: 8823
  • Overlooking the path Mary's little Lamb took..
  • Liked: 1318
  • Likes Given: 306
Re: NASA will help probe Toyota accelerators
« Reply #14 on: 04/05/2010 02:03 am »
I honestly wonder if it is a shielding/grounding issue, I had a GEN 1  '02 prius a while back and a first generation XM Delphi radio. On long drives the XM would randomly freak out and stop working. Same radio works fine in my wife's honda (been in there the last to years), never an issue. To get it to work fine in my ole prius I had to slap some ferrite's on the thing. When I got my newer Gen 2 '05 prius I no longer needed the ferrite's for the skyfi, but still leaves me wondering.  I since upgraded to an XM inno and no issues to date.

I suspect they have some grounding/shielding issues myself... Maybe cosmic, maybe mechanical (They did identify two separate weaknesses in the design), possible sw?  Only a good rigorous systems engineering based review of the design will find if any additional ghosts lurk in the machine. I think that is what NASA brings to the table.

NASA, the only place I know that can have endless meetings and at the end of the day actually do some amazing things despite all the meetings. It may be ITAR to explain how they pull that one off, but I really really would love to know how they do it ;)
If you're happy and you know it,
It's your med's!

Offline Patchouli

  • Senior Member
  • *****
  • Posts: 4490
  • Liked: 253
  • Likes Given: 457
Re: NASA will help probe Toyota accelerators
« Reply #15 on: 04/05/2010 03:10 am »
I honestly wonder if it is a shielding/grounding issue, I had a GEN 1  '02 prius a while back and a first generation XM Delphi radio. On long drives the XM would randomly freak out and stop working. Same radio works fine in my wife's honda (been in there the last to years), never an issue. To get it to work fine in my ole prius I had to slap some ferrite's on the thing. When I got my newer Gen 2 '05 prius I no longer needed the ferrite's for the skyfi, but still leaves me wondering.  I since upgraded to an XM inno and no issues to date.

I suspect they have some grounding/shielding issues myself... Maybe cosmic, maybe mechanical (They did identify two separate weaknesses in the design), possible sw?  Only a good rigorous systems engineering based review of the design will find if any additional ghosts lurk in the machine. I think that is what NASA brings to the table.

NASA, the only place I know that can have endless meetings and at the end of the day actually do some amazing things despite all the meetings. It may be ITAR to explain how they pull that one off, but I really really would love to know how they do it ;)

Some of their troubles could simply be skimping on the electrical connector plating.
Use anything but gold or at the very least nickel you're asking for trouble in this application.


This is complete incorrect. You *can* select neutal all the time, even with the gas pedal floored. You can do it two ways: either you select "N" with the shifter and hold it there for 1s, or you tap the "P" (Park) button and the neutral is engaged immediately (if you press the "P" button while driving, it selects neutral, not park - this is a safety feature). If you do that while stomping on the gas, the engine would shut off automatically. Again, I *tried* that with my *own* car and I can confirm that it works as I say.

First I am familiar with the synergy drive train and the fact it's a planetary gear train with two motor generators and by a safety I mean a cut off control separate of the primary drive train controller one that will work even if the ECU enters a race condition no pun intended where it'll ignore farther inputs.
If Steve Wozniak says the Prius has a flaw I think I'd believe him over you.

http://www.huffingtonpost.com/2010/02/04/steve-wozniaks-prius-prob_n_448778.html
I don't think it's comic rays ,tin whisker or anything exotic but just a simple software flaw where the ECM enters a race condition.
http://en.wikipedia.org/wiki/Race_condition
The fact it happens with cruise control seems to point to a race condition.
The cars don't seem old enough for tin whisker to have developed unless some truly poor quality solder was used.
You may have even seen this on you own PC where it'll seem to ignore inputs for several seconds if you have several applications open or one decides to not play nice.
Some newer models lack dedicated cruise a servo and the cruise control just sends data on the CAN bus to the ECM which tells the throttle servo to open.

I suspect the ECM may be entering a race condition if it sees input from the accelerator and a throttle request from the cruise control.

Audi had a similar problem back in the 80s.
http://www.edn.com/blog/1690000169/post/630052463.html

BTW why aircraft don't suffer the same problems more testing and everything is triple redundant.
But then an aircraft with flyby wire is $10,000,000 while a car could be $18,000

There's some aerospace people in here who know more on the subject then I do that can explain in more detail.
« Last Edit: 04/05/2010 03:28 am by Patchouli »

Offline vt_hokie

  • Senior Member
  • *****
  • Posts: 3054
  • Hazlet, NJ
  • Liked: 118
  • Likes Given: 436
Re: NASA will help probe Toyota accelerators
« Reply #16 on: 04/05/2010 04:03 am »
I don't necessarily mind some of the drive by wire technology, but it's scary if the claims of some Toyota drivers that the automatic transmission wouldn't respond in runaway incidents are true.  And e-brakes are becoming common.  Subaru, for example, replaced the good 'ol cable operated parking brake with an electronic brake controlled by a button on the dash on the redesigned Legacy.  There has already been a software update in response to uncommanded parking brake releases on that vehicle.  Call me old fashioned, but when we start relying on electronics and software for even the parking brake, maybe we're taking it a bit too far!

Offline Patchouli

  • Senior Member
  • *****
  • Posts: 4490
  • Liked: 253
  • Likes Given: 457
Re: NASA will help probe Toyota accelerators
« Reply #17 on: 04/05/2010 04:28 am »
I don't necessarily mind some of the drive by wire technology, but it's scary if the claims of some Toyota drivers that the automatic transmission wouldn't respond in runaway incidents are true.  And e-brakes are becoming common.  Subaru, for example, replaced the good 'ol cable operated parking brake with an electronic brake controlled by a button on the dash on the redesigned Legacy.  There has already been a software update in response to uncommanded parking brake releases on that vehicle.  Call me old fashioned, but when we start relying on electronics and software for even the parking brake, maybe we're taking it a bit too far!

I could not agree more I don't mind electronic throttle if there is redundancy as found on GMs and Fords but the parking brake should be a pure mechanical affair that will work even if the battery is dead.
If I was head of the company I'd fire the baka who came up with the button operated brake on the spot.
I'd do it personally vs have HR do it.

Electronic e-brakes and primary brakes that cannot operate in pure hydraulic mode as a backup really need to be outlawed by the NHSTA

I actually had a battery cable short on car once because the positive cable was near the exhaust.
The positive cable went strait to the solenoid which was mounted to the starter.
GM and several Japanese makes do this.

The car had ABS,computer controls etc.
I did not panic as I still had steering and brakes even though the engine went dead.
« Last Edit: 04/05/2010 04:42 am by Patchouli »

Offline A_M_Swallow

  • Elite Veteran
  • Senior Member
  • *****
  • Posts: 8906
  • South coast of England
  • Liked: 500
  • Likes Given: 223
Re: NASA will help probe Toyota accelerators
« Reply #18 on: 04/05/2010 10:04 am »
I don't necessarily mind some of the drive by wire technology, but it's scary if the claims of some Toyota drivers that the automatic transmission wouldn't respond in runaway incidents are true.  And e-brakes are becoming common.  Subaru, for example, replaced the good 'ol cable operated parking brake with an electronic brake controlled by a button on the dash on the redesigned Legacy.  There has already been a software update in response to uncommanded parking brake releases on that vehicle.  Call me old fashioned, but when we start relying on electronics and software for even the parking brake, maybe we're taking it a bit too far!

I could not agree more I don't mind electronic throttle if there is redundancy as found on GMs and Fords but the parking brake should be a pure mechanical affair that will work even if the battery is dead.
If I was head of the company I'd fire the baka who came up with the button operated brake on the spot.
I'd do it personally vs have HR do it.

Electronic e-brakes and primary brakes that cannot operate in pure hydraulic mode as a backup really need to be outlawed by the NHSTA

The parking brake is the hand brake.  It is safe for that one to be electronic, when it fails you simply cannot move the car.  It is the speed controlling foot brake that needs the backup.

Offline madscientist197

  • Regular
  • Full Member
  • ****
  • Posts: 1014
  • Liked: 6
  • Likes Given: 0
Re: NASA will help probe Toyota accelerators
« Reply #19 on: 04/05/2010 10:56 am »
Did you read the bit about "uncommanded parking brake releases?" IMHO there's something to be said for simplicity.

Actually, I was just over in the US recently -- one of the things that surprised me was how few people over there actually use the parking brake when they park. They seem to assume that putting the gearbox in park is enough (not to say that it isn't, but...).
« Last Edit: 04/05/2010 10:57 am by madscientist197 »
John

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
0