February 12, 2016
Without even waiting for Fiat Chrysler to reply to its Information Request, the National Highway Traffic Safety Administration has bumped a Preliminary Evaluation into the electronic shifters in 856,284 late model Jeep Grand Cherokee, Dodge Charger and Chrysler 300 vehicles up to an Engineering Analysis investigation.
More than 300 consumers have complained to FCA and NHTSA about the difficulty in shifting the gears or that their vehicles rolled away after they placed the transmission in park, crashing into buildings and trees, causing thousands of dollars of damage, and, in some cases, injuries. Thirty individuals, either caught by surprise, or trying to re-enter the rolling vehicle to stop it, have sustained fractured pelvis, a ruptured bladder, fractured kneecap, broken ribs, a broken nose, facial lacerations requiring stitches, sprained knees and severe bruising when they were run over.
What exactly is NHTSA investigating in 2014-2015 model year Jeep Grand Cherokees and the 2012-2014 Dodge models? In August, at the opening of the Preliminary Evaluation, the agency was looking into: “Complainants allege incidents of rollaway after the vehicle has been shifted to Park.” Seven months later, the Office of Defects Investigation is looking into: “Drivers may exit the vehicle when the engine is running and the transmission is not in Park, resulting in unattended vehicle rollaway.” See what they did there?
The vehicles’ Monostable electronic gearshift assemblies do not physically move the gearshift into a detent, but send a gear request from the driver via the Controller Area Network (CAN) bus to the Transmission Control Module, which makes the requested shift. To change gears, the driver depresses a button on the shift lever and moves it to the gear position, then the lever springs back to a centered/neutral position. The gear is displayed on the lever and on the dashboard. The driver’s foot must be on the brake to shift gears. (You can watch the FCA video below.)
This investigation has been underway since August, but there is little in the public file beyond a fully redacted technical presentation FCA made to NHTSA in September, the Information Request, and agency documentation of the existence of these investigations. In the Opening Resume of EA16-002, NHTSA mentions that it tested the shifter and found its performance “not intuitive and provides poor tactile and visual feedback to the driver, increasing the potential for unintended gear selection.” This testing was not memorialized in the public investigation file. Now as we know, humans – and especially humans who work in the Office of Defects Investigation – like to boil automotive problems down to a single cause. (As The Safety Record likes to say: “There’s no all-ya-gotta-do-is…) But, if recent defect history is any indication, many problems are multi-root cause. Toyota Unintended Acceleration, Takata airbag inflators, and wandering GM ignition switches are all cases in which millions of vehicles have been recalled and “remedied,” only to have the problems persist, suggesting that not all of the root causes, or the wrong root cause was identified.
In reading the more than 70 Vehicle Owners Questionnaires (VOQ) reported to NHTSA and the description of ODI’s “testing” regime (which probably consisted of sitting in a Jeep Grand Cherokee and playing with the shifter), we can see scenarios that suggest a human factors problem, an electronics problem, a mechanical problem or any combination thereof. (We’ve assembled the complaints cited by NHTSA here.)
The Human Factors Issues
Many of the defect descriptions in the VOQs suggest that some drivers believe they had pushed the gear shift all the way forward to the Park position, but may have actually stopped at the Reverse position next to it. The driver thinks the transmission is in Park and hits the Start/Stop button to turn the vehicle off. But, since the vehicle is not really in Park, the engine continues to run. The engine is quiet, the telltale is inadequate, and so the driver doesn’t notice upon exit that the Jeep is still running and in gear. VOQ 10672445 sums it up nicely:
“When I put the car into park, it pops into reverse. Then I hit the engine off button, but since it is in reverse, the engine stays on. Then I open the door to get out, thinking the engine is off and the car is in Park, and it starts rolling backward. This has happened 6 times.”
This problem is the legacy of the agency’s decision to allow the introduction of keyless ignition without an appreciation for the unintended consequences of re-defining the “key,” for regulatory purposes, as an invisible electronic code, while allowing manufacturers to give consumers the distinct impression that the fob is the key.
First, note the bewilderment in this complaint, from a consumer in Milford, Michigan who could not understand how his vehicle could not be in Park, if he had the fob in his hand:
“Upon exiting the vehicle, both my passenger and myself, the vehicle was left running, but I took the key fob with me. I then proceeded to walk around and behind the vehicle to the other side of the car. The gas station was having construction work, and in front of the vehicle were cones and the area was blocked off. Approximately 30 seconds later, while both my passenger and myself were inside of the gas station, a witness came running to the gas station window to alert the attendant, that a Jeep Grand Cherokee has just taken off across the parking lot. At this point, I dropped what was in my hands, and ran to look. The Jeep Grand Cherokee, 2014, had crashed itself. One would not think this is even a possibility. Please note, the keyfob was not in the vehicle, the vehicle was placed in park prior to exiting, and the doors were locked.” (VOQ 10763284 )
Here’s another consumer, from Montana, trying to explain it to NHTSA:
“I have huge problems with the keyless ignition and the engine not turning off in two circumstances. First, the engine is very quite [sic] and the radio does not turn off when the car is turned off. If you miss the shut off button the car continues to run even when you leave with the key. There is no warning when you leave the still running car with the key. You come back hours later and the car is still running.” (VOQ 10823099)
Second, note the agency’s description of the Jeep Grand Cherokee’s engine shutdown logic:
“In addition, the engine Start/Stop push-button control logic does not permit normal engine shut-off when the transmission is not in Park. This logic may provide feedback to drivers who attempt to turn the engine off when the transmission is not in Park.”
The irony here is that FCA vehicles are among the few in the new classes of keyless ignition vehicles that actually follow the well-established regulations under FMVSS 114 (not that it seems to matter). In 1990, NHTSA issued a final rule requiring vehicles with automatic transmissions that have a Park position to have a key-locking system that prevents removal of the key unless the transmission is locked in Park or becomes locked in Park as the direct result of removing the key. This was a specific countermeasure against rollaways.
In a vehicle with an old-school ignition using a metal key, you simply could not leave a vehicle running if the key was in your hand, and the transmission had to be in the Park position.
FCA has designed a system that complies with this regulation. In fact, in September 2008, Chrysler recalled 6,636 MY 2008-2009 Dodge Challenger vehicles equipped with automatic transmissions and “Keyless Go” option, because a driver could depress the stop/start button and turn off the engine when the vehicle was not in park, take the fob and exit the vehicle. Chrysler recognized this sequence of events as a clear violation of the standard, which “specifies vehicle performance requirements intended to reduce the incidence of crashes resulting from theft and accidental rollaway of motor vehicles.” Chrysler remedied the defect by reprogramming the Wireless Ignition Node module so the engine can only be turned off when the transmission and gear selector is in the "Park" position.
Today, many manufacturers have vehicles with keyless systems that allow engine shut down in gears other than “Park” and drivers to exit with the fob (which is not the “key”). In January 2014, NHTSA opened a keyless ignition compliance investigation involving Toyota, Ford, General Motors, Nissan, Mazda, Hyundai and Kia 2012 and 2013 model-year vehicles, based on tests of how their keyless ignition systems operate under different scenarios. Almost all of the tested vehicles failed to automatically shift the transmission into “Park” at shutdown. Technicians started the vehicle, shifted the transmission out of park, turned off the vehicle, and waited 30 seconds before exiting the vehicle. Then, the technicians pushed the vehicle to determine if it had automatically locked in “Park.” Of the 34 vehicles NHTSA tested 25 (73 percent) did not automatically lock the transmission in Park – all could roll.
(NOTE: The Safety Record has labored diligently since then to discern how this compliance investigation ended. In July 2014, we asked NHTSA for the documents regarding the conclusion of this investigation. Our Freedom of Information Act request is still pending some 580 days later.)
The importance of human factors research in designing automotive controls has been long recognized, and researchers understand the concept that drivers have ingrained expectations for the operation of vehicle controls, and that changes to those controls induce mistakes. The 1987 Human Factors Research on Automobile Secondary Controls: A Literature Review by researchers at the University of Michigan Transportation Research Institute examined studies from the 1960s forward and summarized basic principles that hold true today. For example, they concluded that previous studies showed: “In designing and positioning controls for cars, driver expectancies for control location, method of operation, and switch type should have a major influence on the designer's decisions. When controls are not placed where people expect them or operate differently than expected, it takes drivers longer to use them and they make more mistakes in doing so.” We also know that manufacturers give this element of the design process short shrift. (As the announcer in the embedded FCA video notes: “now the electronic shift lever with this transmission is a little different…”)
Newsflash to NHTSA, the Alliance of Automobile Manufacturers, and Automotive press I-agree-with-NHTSA: E-shifters-stink-but-they're-not-defective News – if the people who buy your product consistently fail to use it properly, then they are not stupid, the engineers who designed it and everyone who signed off on it are the root cause.
The Case for Electronic or Mechanical Issues
Now, before you say: Well, it’s an obvious case of driver error – there are complaint narratives that don’t fit the didn’t-push-shifter-all-the-way-forward concept. There are cases where the driver says that the vehicle PRNDL telltale or other indicators showed that the vehicle was in Park at the time of the rollaway:
“Vehicle rolls forward while in park. Shift selector is in park and the vehicle rolls. Marked crash due to one instance where it rolled in to the house deck. No damage to either vehicle or deck other than minor scratches. This has happed several times and I have a video of the vehicle rolling while shift selector is in park.” (VOQ 10822442)
“While in “Park” and idling, the vehicle rolled forward and struck objects 20 yards away causing $1500 worth of damage to the right front bumper. My wife parked the car and had exited the vehicle, when after about 30 seconds, it rolled forward and struck headstones in a cemetery. The car still indicated it was in "Park" when my son re-entered the vehicle.” (VOQ 10787576)
“I had a roll away incident back around August 25th 2015. I pulled into my driveway and put the car (2014 Jeep Grand Cherokee Overland) in park. I turned it off and exited the car. It started rolling forward so I jumped back in the car and put on the brake then moved it backwards a little and exited the car again. Again, the car was in park and shut off. As I was walking away from the car it rolled forward again about 8 feet. I know it was in park, when it was not in gear. I got out of the vehicle and it started to roll back on me. I quickly jumped back into the vehicle to stop it without accident, this time. Now I put a block in front of the tire so it won’t roll away. I won’t park it on a hill. Dealer can’t seem to find a problem.” (VOQ 10759622)
“Car rolled even though vehicle was in park and vehicle was not running. 1st incident happened within 10 min of walking away from car, rolled into another car. 2nd incident happened nearly 2 hours after being parked, and 3rd incident happened within 1 hour of being parked on slight incline. Dealer unable to recreate incident. Factor rep called Dodge, Dodge called transmission mfg and had valve com replaced.” (VOQ 10508134 )
If the vehicle is rolling even when the P is illuminated, that’s a different kettle of fish. The defect could also have an electronic cause. As we have seen in other defect scenarios, a driver’s requests for basic operational functions – accelerating, braking, and steering – as transmitted via microprocessors are not always obeyed – for reasons that include poor software design or hardware problems caused by electrical shorts or poor contacts. Or it could be mechanical issue with the pawl – or some interaction between the electronic and mechanical systems. At least one consumer reports having a part replaced.
NHTSA (People Blaming People ™) and automakers love to conclude these little get-togethers with an agreement that consumers are just doing it wrong. There’s no fight over if there will be a remedy, or what the remedy will be, or there’s some cheap, ineffective fix like new floor mats, and everyone can get back to business. The agency’s response to Toyota Unintended Acceleration had a lot more bells and whistles – significant fines and a half-assed investigation – but still followed the playbook. NHTSA stuck with the mechanical causes of Unintended Acceleration – errant floor mats or pedal misapplication – even though there were clearly many complaints that simply could not fit into those narrow categories. For example, some Toyota drivers reported high-speed UA events in which the vehicle only had the standard carpet mats – not the heavy all-weather mats declared to be the culprit in floor mat entrapment scenarios. A driver already underway at highway speed is not misapplying the pedal. NHTSA just threw out complaints that didn’t fit the narrative NHTSA and Toyota wanted to push.
FCA obviously knew the gear shift design sucks – it changed it in the Dodge vehicles in 2015 and in the Jeeps in 2016. That doesn’t help the poor saps who are stuck with it now. And we hope NHTSA doesn’t cherry-pick the data, without exploring the possibility that they are rolling even when the vehicle tells the driver that the vehicle is in “Park.” So far, these electronic shifters have only produced injuries, instead of the intended shift response.