March 31, 2016
Earlier this month, Rich Grandy of Crystal Beach, Florida was easing his 2005 Toyota Tacoma into a parking space in front of his local 7-Eleven, when it took off, hit the front doors of the convenience store and shattered the adjacent picture window. The only thing that kept the Tacoma from advancing further into the convenience store was the low wall that framed the window and Grandy’s foot clamped on the brake. As his vehicle attempted to labor forward, Grandy shut off the ignition, and the event stopped.
Grandy loved his Tacoma – he bought it used in May 2013 after much research, and with full knowledge of Toyota’s unintended acceleration problems.
“I’ve read the newspaper my whole life. I probably knew more than the average person about Toyota, and I was familiar with the [fatal Saylor crash] in California,” says Grandy, a retired general contractor. “I thought there was probably a problem, but I also thought they were a big corporation dodging a bullet. I figured that it was such an isolated problem, it probably would not happen to me.”
But it did. Twice. In late March 2014, Grandy was making a slow right into a head-to-head parking space, and was braking to a stop, when the Tacoma accelerated forward and over a Honda sedan, which became wedged under this truck. In both cases, the UA occurred after the same sequence of events: Grandy had taken his foot off of the accelerator, and was coasting with his foot resting on the brake. When he actually engaged the brake to bring the vehicle to a full stop, it surged forward.
“I left a really big tire patch on the asphalt, and you could hear the tires squealing like crazy while I was trying to hold it back on the brake,” he recalled.
With the GM ignition switch crisis, closely followed by the exploding Takata airbag inflator crisis, it may be hard to recall that just five years ago, our dearly departed Secretary of Transportation Ray LaHood declared: “The jury is back. The verdict is in. There is no electronic-based cause for unintended high-speed acceleration in Toyotas. Period.” But, far from ancient history, Toyota unintended acceleration incidents continue to happen with regularity to older vehicles, newer vehicles, hybrid vehicles, vehicles that received the sticky pedal and floor mat recalls and those that did not. Vehicles that….well, you get the picture. Neither public relations, nor the intervention of NASA and the National Academy of Sciences, nor million dollar fines, nor billion-dollar deferred settlement agreements between the government and Toyota have done anything to solve this technical issue.
To date, consumer complaints to the National Highway Traffic Safety Administration and Safety Research & Strategies (vetted to eliminate as many duplicates as possible) have reached more than 9,400. Speed control complaints for two Toyota vehicles, the 2006 Camry and the 2010 Corolla, continue to occupy spots on The Safety Institute’s Vehicle Watch List, a quarterly report monitoring potential vehicle defect trends and NHTSA’s recall and enforcement activities, using death and injury claims and early warning reports filed to the agency. Here are a couple of typical complaints added to the VOQ database this month involving late model Toyotas:
From Louisville, KY: “The contact owns a 2015 Toyota Camry. While attempting to shift the vehicle in reverse, the vehicle independently accelerated and the check engine indicator illuminated on the instrument panel. As a result of the independent acceleration, the contact crashed into her neighbor's car port. The air bags failed to deploy. There were no injuries and a police report was filed. The vehicle was towed to the dealer, but was not repaired. The manufacturer was not made aware of the failure. The failure mileage was 5,400.” (ODI 10839884)
In Plano, Texas, the owner of a 2015 Lexus ES 350 complained: “I was pulling into my garage slowly, braking to stop, when the car accelerated. I could hear the engine rev and I pressed hard on the brake and put the car in park. The car traveled about 4 feet before it came to a stop. I reported to Lexus who turned over to 'legal' and I have been waiting for a week for an investigator to contact me to look at the car, after which they say it will be another 30 days for a finding. The paperwork I filled out termed this as 'unintended acceleration". This is the second time this has happened with this car, which I purchased new and have had for a little over a year.” (ODI 10840059)
Amazing. No one can get a Toyota in or out of a parking space without making a big whoopsie! NHTSA is so concerned about this new generation of inept drivers not seen since the advent of electronic throttle controls, that in May it actually issued a consumer advisory entitled Reducing Crashes Caused by Pedal Errors. It makes the dubious claim that each year “approximately 16,000 preventable crashes occur due to pedal error when drivers mistake the accelerator for the brake.” And it contains helpful hints such as: Adjust your seat, mirrors, steering wheel and pedals; aim for the middle of the brake and wear light-weight, flat soled shoes when driving.
The agency, however, is not concerned enough to actually investigate the causes of these preventable crashes; it’s too busy gas-lighting the public. To date, NHTSA has fielded nine defect petitions for Toyota Unintended Acceleration. Nine! This is an astounding number of petitions for a single defect that has no parallel in agency history. But, it’s given the Office of Defects Investigation plenty of time to refine its driver error arguments and smooth out its boilerplate petition denial language.
In the last two years, the agency turned back three requests for defect investigations from Robert Ruginis (Read NHTSA Denies Unintended Acceleration Defect Petition), James Stobie, and Gopal Raghavan, Toyota owners who experienced UA events in parking scenarios. The trio used the contradictions in their Event Data Recorder (EDR) readouts to buttress their arguments that NHTSA ought to pursue a vehicle-related cause. But, its pedal error today, pedal error tomorrow and pedal error forever over at 1200 New Jersey Avenue, SE.
In the past, NHTSA argued that pedal misapplication is the result of a driver accidentally depressing the gas instead of the brake, and, when the vehicle moves forward or backward, the driver compounds this error by pressing the accelerator pedal harder until the inevitable crash. Never mind that that this theory was derived from a “thought experiment,” (Read The Pedal Error Error.) is not supported by any empirical research, and makes no sense in high-speed unintended acceleration incidents when one’s foot is already on the accelerator or in parking scenarios where practiced drivers ease on and off the pedals, because NHTSA made up its mind in 1989.
Access to EDR data made this theory a little trickier to apply. Luckily, NHTSA found a trapdoor in the asynchronous nature of individual datum points to justify their denials. EDR records the pedal voltage at vehicle idle, while the rpms zoom upwards? The driver’s just hitting the pedal in the milliseconds in between data samples. In this way, investigators can read the data any way they want.
The agency took particular pains to sneer at Gopal Raghavan, an electrical engineer with a PhD from Stanford University and more than 20 years’ experience in high-speed circuit design and device modeling. Dr. Raghavan worked as a senior engineer with Intel Corporation, and a principal engineer at Conexant designing integrated circuits. He also holds 10 patents and has published more than 30 technical publications.
He submitted the EDR readouts of two other crashes that shared similarities to his, but NHTSA dismissed the idea that the pattern indicated anything other than pedal misapplication and informed Dr. Fancy-Pants-Ten-Patents that “the common pattern is that the ‘glitches’ occur at the moments in the events when the driver should be initiating braking, but no braking has occurred,” and called this “a signature of pedal misapplication by the driver.”
Electrical engineer Antony Anderson, who frequently writes about unintended acceleration in automotive electronics, published a critique of NHTSA’s denial in IEEE Access. Case Study: NHTSA’s Denial of Dr Raghavan’s Petition to Investigate Sudden Acceleration in Toyota Vehicles Fitted with Electronic Throttles, notes:
In the hypothetical case of such a panic-induced sudden acceleration the accelerator rate signal would go to its maximum value very shortly after the pedal was fully depressed and would stay there. Such a constant accelerator rate signal is not found in any of the examples cited by Dr Raghavan. In an attempt to explain this failure of the EDR results to fit the pedal error hypothesis, NHTSA has developed a new hypothesis in this DP which appears to require that drivers when coasting in to park, engage in some kind of multiple foot stomping action on the accelerator pedal which isn’t detected by the EDR.
Consider such hypothetical pedal stomping activity, assuming for the moment that is a realistic possibility. Since the EDR data sampling rate is once per second (1 Hz), any stomping would also have to take place at the same frequency and be precisely syncopated with the EDR data sampling over a period of 4 to 5 seconds. To carry out a successful sequence of stomps, the driver would have to synchronise his foot actions with the data sampling of the EDR. NHTSA has so far failed to produce any experimental evidence, peer-reviewed articles, or research reports that demonstrate that panicked drivers either could, or would, ever go into such a precisely timed, synchronized and syncopated pedal stomping routine.
Anderson says he was moved to dismantle NHTSA’s “formulaic approach” out of “a real annoyance” with the agency’s “disdain” of a competent engineer “who came forward with a good case to do some investigation. It was so nasty. It’s not right.”
Toyota Keeps on Settling
The automaker has taken a different, but complementary tack to NHTSA’s denial of electronic defects – it settles death and injury cases. To date, the Intensive Settlement Process (ISP) has resolved 422 cases. The ISP is a two-step process that begins with an initial settlement conference, and if the matter is not resolved, proceeds to a formal mediation. More than half of the cases – 233 – were settled, 196 were dismissed or awaiting dismissal, and the rest are in some stage on the process.
It only took one stinging loss in civil litigation to persuade Toyota that it was much better to negotiate confidential settlements than risk a wave of headlines about spaghetti code. Bookout v. Toyota turned the tide in October 2014, after an Oklahoma jury determined that the automaker acted with “reckless disregard,” and delivered a $3 million verdict to the plaintiffs. The trial emanated from a September 2007 crash. Jean Bookout and her friend and passenger Barbara Schwarz were exiting Interstate Highway 69 in Oklahoma in a 2005 Camry when it experienced a sudden acceleration. Bookout tried to stop her car by pulling the parking brake, leaving lengthy skid marks. Her Camry continued to rocket down the ramp, stopping only after its nose was embedded in an embankment. Schwarz died of her injuries; Bookout spent two months recovering from head and back injuries.
Embedded software expert Michael Barr, who spent nearly 20 months reviewing Toyota’s source code, testified that the software code was poorly written and the safety architecture was “a house of cards. (See Toyota Unintended Acceleration and the Big Bowl of “Spaghetti” Code) Barr explained that many of the vehicle behavior malfunctions could be caused by the death of tasks within the CPU; in particular, the death of a proprietary-name task, called Task X, at trial. Barr dubbed it “the kitchen-sink” task, because it controls a lot of the vehicle’s functions, including throttle control; the cruise control – turning it on, maintaining the speed and turning it off ; and many of the failsafes on the main CPU. Barr testified that Toyota’s watchdog supervisor design – software to detect the death of a task — “is incapable of ever detecting the death of a major task. That's its whole job. It doesn't do it. It's not designed to do it.” Instead, Toyota designed it to monitor CPU overload, and, Barr testified: “it doesn't even do that right. CPU overload is when there's too much work in a burst, a period of time to do all the tasks. If that happens for too long, the car can become dangerous because tasks not getting to use the CPU is like temporarily tasks dying.” Barr also testified that the operating system contained codes that would throw away error information, ignoring codes identifying a problem with a task.
Toyota hastily settled Bookout before the jury could determine punitive damages, and it’s been cutting deals ever since.
None of this helps owners like Grandy, who took his complaints to Toyota, to no avail.
“As soon as it happened, I knew I was screwed.”