GM and NHTSA’s “Magic Formula”

Tomorrow, the heads and NHTSA and GM will head into the House committee for a three-Bromo-seltzer morning on the topic of: What Did You Know and When Did You Know It?

We, at The Safety Record, are most interested in understanding why NHTSA declined to investigate the defective ignition modules in early model year Chevy Cobalts and other models, after two Special Crash Investigations, 29 complaints, four deaths and the considered opinion of Defects Assessment Division (DAD) Chief.

According to a briefing report prepared by Majority Staff of the U.S. House Committee on Energy and Commerce’s Subcommittee on Oversight and Investigations, the decision point for the agency was the fall of 2007:

In September of that year, the DAD Chief “emailed other ODI officials and proposed an investigation of “frontal airbag nondeployment in the 2003-2006 Chevrolet Cobalt/Saturn Ion.” The Chief of the Defects Assessment Division went on to state that the “issue was promoted by a pattern of reported non-deployments in VOQ [Vehicle Owners’ Questionnaire] complaints that was first observed in early 2005. Since that time, [the Defects Assessment Division] has followed up on the complaints, enlisted the support of NCSA’s Special Crash Investigations (SCI) team, discussed the matter with GM, and received a related EWD Referral. Notwithstanding GM’s indications that they see no specific problem pattern, DAD perceives a pattern of non-deployments in these vehicles that does not exist in their peers.”

Two months later, an “ODI IE panel reviewed the proposal to open an investigation into non-deployment of airbags in 2003-2006 Cobalts and Ions. A PowerPoint presentation prepared by the DAD and dated November 17, 2007, states that its review was prompted by 29 Complaints, 4 fatal crashes, and 14 field reports. During a briefing with Committee staff, ODI officials explained that the panel did not identify any discernible trend and decided not to pursue a more formal investigation.”

The Safety Record has long observed that we can find no “discernible trend” in NHTSA’s investigation decisions. In a March 8, 2014 New York Times story on the GM debacle ODI Chief Frank Borris said that that calls are made by “really well-seasoned automotive engineers who leverage a lot of technology and lean on past precedent about when to open, when to close, and when to push for a recall. It’s no magic formula.”

Take out the word “magic,” and for once, we agree with Frank.

In February, Safety Research & Strategies submitted comments to NHTSA’s 2014-2018 Strategic Plan docket pointing out this perennial problem, well-documented in a series of Office of Inspector General reports going back to 2002:

– NHTSA uses an unstructured process for determining defects and inconsistent or nonexistent criteria for initiating defect investigations.

– NHTSA makes poor use of available data and refuses to consider information from sources outside the agency or the manufacturer.

– NHTSA focuses on defects that are easily and inexpensively remedied, frequently ignoring more complicated and dangerous defects.

Case in point: In December 2012, The Safety Record wrote a blog about a Ford Fusion floor mat entrapment investigation that was opened in May 2010, based on three consumer complaints of pedal entrapment. Three days before the date on the Preliminary Evaluation Opening Resume, the agency got this VOQ:

Several weeks before the incident, I removed the driver’s side all-weather floor mat (the factory-developed optional one) from our 2010 ford fusion hybrid and tossed it in the trunk. In the days leading up to this incident, the car had been to the body shop and the car wash. As near as I can tell, an attendant at one of those places must have discovered the mat in the trunk and “helpfully” reinstalled it.    But I didn’t notice this right away. When commuting to work on April 16th, I pressed the accelerator to the floor as I moved into a much faster traffic lane. Upon completion of this maneuver I eased up on the gas, but the pedal stayed down and the car kept accelerating, unintentionally at this point. At first I wasn’t sure what was going on, but after about 3 or 4 seconds I heard a “click”, the pedal returned to normal and the car slowed. It was then that i thought of the all-weather floor mat, and when traffic cleared I tried another experimental full-throttle dab to see if I was correct. It happened again (but I was ready for it), with the pedal releasing on its own after a few seconds, just like before. At this point, I was careful not to disturb the mat’s position so I could take video of the problem once I arrived at work.     Once in the garage, and with the engine off, I was able to replicate the floor mat-throttle interference on-camera. As before, the pedal stayed stuck for about 3 seconds after i pressed it to the floor. I was able to repeat this several times. But in one trial the stuck throttle persisted for over 30 seconds, with the pedal staying stuck despite vigorous stamping in an effort to free it. The only way it came loose was by lifting my heel which, in the normal driving posture, had apparently been pinning the mat down and making it less flexible. With the heel pressure removed, the mat bent up out of the way and the pedal began a slow return.     A YOU TUBE VIDEO LINK IS AVAILABLE UPON REQUEST

By 2012, NHTSA bumped the investigation up to an Engineering Analysis, and just for fun The Safety Record decided to read all of the Ford Fusion speed control complaints to understand how this floor mat entrapment problem fit into the bigger picture. We found seven consumers who unambiguously reported that they observed the floor mat entrapping the accelerator pedal; 48 who had unintended acceleration complaints – from surges to highway-type events; and about 92 complained to NHTSA that the Ford Fusion suddenly decelerates.

So what was the formula there? What technology did NHTSA leverage? YouTube? What past precedent did the seasoned investigators lean on? The precedent of the Toyota UA controversy that also landed NHTSA on the wrong side of a table in a hearing room?

So here we are again, gather together to be called into account. NHTSA now says that for the last two years, it’s been using IBM software to augment the opinions of the seasoned investigators.

We’d love for the committee to get more details about that.