Hmmm, the NHTSA-NASA Study of Toyota Unintended Acceleration
I took a quick look at the 177 page NHTSA-NASA Study of Toyota Unintended Acceleration full report this evening. I had hoped that with NASA involved, it would have been a really hard core report. Sadly, the design of the experiments aspect of it appears to so financially and resource constrained, it seems to raises more questions than answers.
For sure, the obvious things tested for in published standards are ruled out. In addition, the human factors analysis appears well done, albeit such is mostly outside the scope of my expertise. Perhaps the biggest disappointment is the lack of detail provided… yes, its likely written for politicians, rather than EMC engineers, but being NASA, I guess my expectations were a lot higher than reality.
Since I have a number of years in the EMC field, I’ll take a look at section 6.8 My guess is there are likely substantial internal NASA discussions on these factors, albeit they didnt make it into the report.
Sample size:
Six vehicles were chosen for testing, all of which were reported to experience UA… Sounds good, but none of them have the same option set, and the years 2005 and 2006, were skipped entirely. Granted, I am not familiar with Toyotas, but I would certainly expect substantial year to year variation.
Frequency range of radiated susceptibility testing
The frequency range of testing lined up with ISO, Toyota, and NESA standards, and on the outset appear a good thing. However, there is a gap between conducted and radiated susceptibility tests (its a royal pita to test there, standards dont require it, and its rare in todays world to have problems in such an area… but I’ve run into it before (science is no respector of standards)).
Dyno testing
Units were tested on a dyno… and a dyno that can withstand 250V/m radiated fields, without causing distortion in said field has to be a really really cool device. On the other hand, in likely trying to avoid killing the dyno, its design probably shoots a number of anomalies into the field surrounding the car (Granted, I’m not sure I’d want to drive a car while subjected to a 250V/m field for an extended time… but it is a test limitation to consider).
High Current Audio Injection Failure precluded some testing
NHTSA-NVS-ETC-SR15 failure report is not included… perhaps the high current failure was an anomaly, perhaps not. Granted, the audio injection test is an unusual one. On the other hand, why did 4 pass, and 2 fail?
Also, being a 3300rpm increase was detected at 2Vpp at 150Khz… perhaps this test should have been run on more than 1 vehicle? Granted, outside of the test lab, such should never occur, but it is a point of interest. I’m a bit surprised testing was not explored at 125Khz, as such has repeatedly bit me in the field over the years. Likely civil litigation will explore this further with differing power levels, and/or frequency sweeps.
DTC Issues
Its assumed that DTC’s which were triggered were reliably recorded. Ie a Toyota Tech Scan was attached and data was pulled before, and after testing. (Obviously having it connected during tests would either kill it, and/or potentially disturb the results).
Orientation
Orientation was limited to 1 orientation for the TEM cell, and 8 orientations for the semi-anechoic. Such seems a reasonable approach, on the other hand… which orientation for TEM cell setup? Were other orientations possible?
Tests not performed (see table 6.8.1-1)
Early failures of vehicles 13C and 14C precluded many tests, and the fact that so many tests were skipped is a cause for concern.
The victim terminology
This is just a point of interest. Years ago, the usage of the terms source and victim were discouraged, as the legal field would tend to run with such. I gotta admit it is nifty seeing such terms once again.
The lack of any discussion or test data concerning ESD
On the outset, short of a minor discussion in section 6.9, this is not mentioned. The ToC suggests that data is available in appendix B, which is pretty nebulous as well. Section 6.9.6 is not a confidence booster, and such does leave the door open for civil litigation. Granted a Camry is not the space shuttle, but I bet the NASA guys were going WTF when they saw the watchdog/heartbeat monitor setup.
The observations in section 7.1
These are interesting…
sample size again is a concern
F-3. At full throttle, brake vacuum boost can be diminished or lost (apparently unlike Nissan of 20 years ago, they dont see the need for an auxiliary vac pump for such situations). It was also interesting, that even with the vacuum boost lost, one can decel at 0.25g with 112lbf on the brake pedal at 30mph. I wonder what the deal is a 75mph?
F-5 Tin whiskers… no wonder Uncle Sam and aerospace is exempt from Rohs. At least Toyota expected this might happen and designed around it. The big question though, is what if the whiskers show up in other places, and what about other manufacturers?
O-3, and O-4 are interesting… yes, the smog stuff is important, but I’m not sure its wise to not hold safety to a similar standard.
O-7 and O-8… yep, this is a problem. If I had a Camry, I’d code up an Iphone app to address this exact situation. If they really want to get to the bottom of it, wide scale testing either via an Iphone, Android, or something related might be a cheap and fast way to explore this aspect.
O-11, I agree 100%… even more so if the driver is tired, and its a different vehicle than what they normally drive. Ie a rental car fleet an airport might well provide for a great study of this human factors issue.
7.3-R-1 Likely they wont approach any of the concerns presented in this report, until either something really high profile occurs, or as the vehicles age, more problems show up.
My conclusion
Most of the obvious issues have been ruled out, or put in the category of being statistically low probability events. This does not mean there isnt some type of anomaly, but that if there is, it will be rare, and it will be a really bugger to get to the bottom of it. The biggest deal to me, is that so many of these issues transcend the whole industry, Toyota just happened to get in a jam mechanically (floormats, and/or brake pedal sticking) which put them in the spotlight.



