Combustion

With Misleading Messages And Customer NDAs, Tesla Performs Stealth Recall

By E.W. Niedermeyer


Combustion is The New Cartographer's guide to the
present and future of the auto industry.

We provide insight via our email newsletter to executives, investors, and decision-makers on where technology, policy, and the market are headed next.



Enter your email to read a free preview, or learn more about a paid subscription


In early 2014, a Tesla Motors Model S that was part of the Edmunds.com test fleet suddenly lost power while merging onto a freeway. The car flashed several warning messages — "12V Battery Power Low - Car May Shut Down Unexpectedly" and "Car Needs Service - Car May Shut Down Unexpectedly" — before coming to a stop on the freeway onramp. Eventually a tow truck came to haul the stricken vehicle to a Tesla service shop, where the company replaced its drive unit for the third time.

This was a big deal. Not only is it embarrassing when a media outlet’s car breaks down in such a dramatic fashion, but any defect that causes a sudden loss of vehicle power is typically considered safety-related. There was also a grim point of comparison: the GM ignition switch scandal that was just starting to explode into the mainstream media had involved a defect that produced a similar sudden loss of vehicle power. Several months later, on Tesla’s second quarter earnings call, CEO Elon Musk addressed the incident by telling analysts:

“Well, there's definitely some genuine issues we have with the car, but they had one of our early production units, and in fact most of the problems that they have encountered there are not present in current cars. We also -- I think this may be ending up being counterproductive, but the service team was ultra proactive with the Edmunds car. So they would -- they were doing their best to make Edmunds happy, and I think unfortunately that resulted in them changing things up, just on the off-chance something might go wrong.

So that drive unit issue that I mentioned earlier were, the drive units [were] replaced even though it wasn't a drive unit problem, that happened with them twice. So, unfortunate sort of case, but I don't think it's broadly correct.”

Earlier in the call Musk had addressed the issue of Model S drive unit failures more broadly, explaining that the issues were often unrelated to the drive unit itself and that the necessary fixes --a $3 cable tie and a 50 cent shim for the differential-- were not expensive. With Musk’s assurance that these mysterious issues were not costing the company significant amounts of money, investors and the public eventually lost interest. NHTSA, which had its hands full with the GM scandal, never investigated. Though apparent drive unit failures continued to be reported, including several instances of cars suddenly losing power as the Edmunds car had, the whole issue steadily receded from public attention.

Then, in January of 2015, a Model S owned by a French taxi driver lost power while accelerating on a highway. The car threw the same error messages as the Edmunds car had before coming to a stop on the side of the road. The Model S taxi was also towed to a Tesla repair shop, but there its story diverges from the Edmunds story. Tesla didn’t replace the car’s drive unit, as it had with Edmunds, and its service invoice made no reference to the terrifying loss of power the driver had just experienced. In the field designated “Description du probleme: customer states,” Tesla wrote (in English):

“As part of providing peace of mind and a great ownership experience, Tesla vehicles are equipped with telematics systems to provide remote diagnostics support. We have been notified this vehicle has been remotely diagnosed that the Power Switch and Power Supply would benefit from the latest generation components.”

This use of what appeared to be boilerplate language surprised the taxi driver. Why didn’t Tesla simply describe what had happened to him in the space designated for just that? Why was a repair following a potentially dangerous incident being described as an “update” intended to “benefit” his car, and attributed to Tesla’s remote diagnostics? The taxi driver became suspicious.

It turned out that a number of Tesla owners had reported remarkably similar incidents in online forums, and that the failure of high voltage contactors was a fairly well-known issue. An owner who posts on the Tesla Motor Club forum under the username mknox describes an almost identical failure, complete with the loud bang and warning messages, and reported that Tesla service subsequently replaced his car’s high voltage contactors. Another owner, posting on the official Tesla.com forum under the name bobgriswold, reported the same malfunction while accelerating on a highway, which was also traced back to a faulty contactor. A TMC poster going by wk057 also lost power on the road, and after initially being told by Tesla engineering that the car’s firmware was to blame, the problem seemed to sort itself out. Then, about a month later he reported receiving an email from Tesla that was identical to the message that appeared on the French taxi driver’s invoice:

“As part of providing peace of mind and a great ownership experience, Tesla vehicles are equipped with telematics systems to provide remote diagnostics support. We have been notified this vehicle has been remotely diagnosed that the Power Switch and Power Supply would benefit from the latest generation components.”

Variations of this message have been reported by Tesla owners around the world. An owner in Washington State reports receiving the following email:

"Engineering has identified your car as potentially benefitting from a switch and power supply update. The technicians will evaluate your high-voltage system and determine whether it would benefit from having the latest generation power switches installed. If they determine that it would, we will perform the installation.

Over a dozen owners from around the world also report being contacted by Tesla and told that they would benefit from a contactor update. One owner recounts such a phone conversation:

“I got the call about a month ago, but they were not specific and required the car to be in the SC for two days, they did not provide details why needed to be changed, they keep insisting that nothing is wrong and is a ‘proactive’ fix.”

Tesla’s emphasis on the proactive nature of this campaign and the “non-safety related” nature of the issue was critical. If Tesla ever admitted, in any form of customer communication, that a contactor failure could result in danger to the driver or other motorists it would be required by law to issue a recall on the part. The TREAD Act of 2000, passed in the wake of the Ford/Firestone recall scandal, requires automakers to report any defect to NHTSA within five days of determining that it affects safety, after which it must order a recall. The same law also requires automakers to report to NHTSA

“Notices, bulletins, customer satisfaction campaigns, consumer advisories, and other communications sent to more than one owner regarding any defect in its motor vehicle equipment, including any failure or malfunction beyond normal deterioration in use, or any failure of performance, or any flaw or unintended deviation from design specifications, regardless of whether or not such defect is safety related.”

Rather than report and recall the contactors as a safety-related defect, Tesla issued a technical service bulletin (TSB). It is unclear when exactly the original TSB was issued, but the first revision was issued in May of 2013 and a second revision was issued in June of 2014. TSBs are only an accepted alternative to a recall by NHTSA if the defect it repairs is not safety related, and Tesla was careful to avoid giving the safety regulator any hint of a risk to drivers. For starters, Tesla called SB-13-44-003 a “HVIL [High Voltage Interlock Loop] Connector Upgrade,” implying that it was intended to improve performance rather than fix a defect. There wasn’t a hint in the bulletin that faulty contactors could cause sudden power loss, just the possibility the defect might make a car fail to start.

By calling the bulletin an “upgrade” and by suggesting that the worst case scenario was an inability to start the car, just four months after the Edmunds car had lost power on a highway onramp, Tesla buried the problem. It was a tactic that had worked before: according to a 2015 report by the Department of Transportation’s Inspector General [ PDF ], NHTSA investigators missed an opportunity to catch the GM ignition switch defect because GM filed a TSB that conspicuously failed to mention the possibility of a vehicle stalling.

Concealing the potential for a defect to cause a vehicle stall is well outside the norms of US auto safety reporting. NHTSA has considered such a fundamental loss of performance to have intrinsic safety implications since the 1970s and between 2004 and 2013, some 91 recalls were carried out to address defects that caused stalling. One of these was even for an electric vehicle: in November of 2013, Ford recalled its Focus Electric car for a problem involving a sudden loss of power that was eerily similar to Tesla’s contactor issue.

Even closer to home, both of Tesla’s (now-former) automaker partners, Mercedes and Toyota, have recently recalled their Tesla-powered vehicles for sudden losses of drive power. Mercedes recalled its electric B Class in 2015 for a problem involving an incorrect signal regarding the status of the high voltage contactor in its Tesla-sourced drivetrain, which could cause the vehicle to lose power. The problem was solved with a firmware update, but Mercedes still went through the recall process that Tesla has strenuously avoided. In 2015 Toyota recalled its own Tesla-powered electric vehicle, the RAV-4 EV, for a software issue that caused the vehicle to unexpectedly shift into neutral.

Why would Tesla not recall vehicles for sudden power loss, when there was no shortage of owner reports of the phenomenon and when both of the automakers who used its drivetrains did? Perhaps because they could. These decisions were made before NHTSA knew that Tesla made owners sign a non-disclosure agreement before performing “goodwill repairs” on defective vehicles or buying them back. The French taxi driver eventually signed an NDA as part of a buyback deal, and at least one online report of lost vehicle power appears to have concluded with an NDA. If owners weren’t independently reporting defects, NHTSA’s only other source of data about defects in Teslas was Tesla itself.

Had Tesla owners been more diligent in reporting sudden power loss to safety regulators, or if NHTSA safety investigators regularly read the online forums where Tesla fans do report defects, this might have been caught years ago. Because so many Tesla owners are also investors, forums often discourage defect reporting for fear that it might lead to bad publicity. So even though Tesla owners were able to connect reports of sudden power loss to the contactor flaw, their knowledge never made it to the safety regulators.

While there are no reported cases of injury or death as a result of this vehicle failure, Tesla repeatedly avoided the regulatory framework created by NHTSA to protect drivers. As the GM ignition switch case and many others have shown, the temporary embarrassment of a recall is far better than burying a defect with potentially serious safety implications. Tesla’s apparent attempts to avoid standard safety regulatory procedures are troubling.

At the time of publication, Tesla had not responded on the record to requests for comment.

Neither NHTSA nor the Department of Transportation responded to requests for comment.

April 2017