DETROIT -- A General Motors supplier studying an “anomaly” in a crash test on a preproduction Chevrolet Cobalt urged the company in 2004 to take a closer look at the connection between the ignition system and airbag deployment -- a link that eluded engineers and investigators at the company for years afterward.
A July 1, 2004, report by Siemens VDO Automotive analyzed why frontal and side-impact airbag sensors simultaneously shut down less than two-tenths of a second after the moment of impact. It was written a little more than a month before GM began building the first Cobalts.
The report, released this week by a congressional committee investigating GM, examined both the results of the crash test and a series of laboratory simulations run by Siemens VDO to determine how the airbag sensors would respond to a loss of power. The cutoff of the sensors “appeared to be indicative of an ignition cycle,” Siemens engineer Douglas McConnell wrote.
He concluded: “It is recommended that future severe crashes have ignition voltage and [in-vehicle network] messages monitored to determine the root cause of the … Power Off issue.”
The document doesn’t identify the flimsy ignition switch as the culprit in the power loss. But it is significant because it shows that, before the first production Cobalt ever left the assembly plant, a GM-commissioned analysis had flagged a potential connection between a loss of power and airbags not deploying, and recommended that GM seek a root cause.
Critical missed link
GM’s inability to recognize that link was the key reason for its failure for several years to identify the defective ignition switch as a fatal safety flaw, rather than a customer inconvenience, an investigation by former federal prosecutor Anton Valukas found. His report, released this month, said that the connection to airbag non-deployment was not widely understood by GM employees until nearly five years after the 2004 crash test.
Valukas wrote that GM engineers who studied the issue of vehicle stalls over the years “inexplicably … did not recognize that the loss of power from a key moving into accessory caused the airbag system to shut off — a missed connection that led to devastating consequences.”
Officials from the National Highway Traffic Safety Administration have said they did not realize that GM had designed its airbags to not deploy when the ignition was out of the “run” position.
‘A big clue’
Erin Shipp, a consultant at Robson Forensic in Denver and a former Chrysler structures engineer, said the Siemens report should have provided GM engineers “a big clue” about the link.
“This testing basically shows that they should have known: If there’s no power, there’s no airbag,” Shipp said. She and two other engineers reviewed the report at the request of Automotive News. The other engineers said they interpreted the report similarly.
“The recommendation at the end of the report, to search for a root cause to the ignition-off issue, was a good one that would have helped them determine if the key was moving,” she said. “They either didn’t do that root cause analysis, or somehow it didn’t lead to a discovery.”
Shipp’s name appears in Valukas’ findings in a reference to a report she filed on behalf of a crash victim suing GM in 2012. She was one of several outside experts who concluded, before GM did, that the Cobalt’s ignition switch was causing airbags to fail.
Siemens is now part of Continental AG, which also noted the connection between ignition mode and airbag deployment in a report it gave to GM in May 2009.
GM spokesman Greg Martin declined to comment on the crash test report. “As the Valukas report has detailed, there were several missed opportunities for GM to properly identify the problem,” he said.
Predates switch complaints
The Siemens report shows that it was provided to five GM engineers, one of whom has worked for NHTSA since 2007.
That engineer, Matthew Craig, identifies himself on LinkedIn as a former safety performance integration team leader at GM and now NHTSA’s chief of human injury research. Craig referred a request to discuss the report to a NHTSA spokeswoman, who declined comment.
The report predates multiple engineering inquiries into incidents of drivers accidentally turning the Cobalt’s ignition off by bumping it with their knee. Because those inquiries characterized the problem as only a moderate inconvenience to drivers rather than a safety issue, it was treated as a low priority and potential solutions were rejected due to cost concerns, Valukas’ investigation found. Engineers involved in those inquiries told Valukas’ team of investigators that they didn’t believe the flimsiness of the switch to have safety implications.
At the time the Siemens report was prepared, only one of the 13 deaths that GM now links to failed airbags caused by faulty ignition switches had occurred. The second of those fatal crashes happened three days later. Both crashes involved 2004 Saturn Ions, which used the same ignition switch as the Cobalt.
GM started production of the Cobalt in August 2004. The Siemens report wasn’t mentioned in Valukas’ report; neither were any of the five GM engineers shown as receiving it.
GM recalled 2.6 million Cobalts, Ions and other small cars earlier this year to replace the ignition switch. It paid a $35 million fine levied by NHTSA for allowing the defective cars to remain in customers’ hands for years after employees learned of problems with the switch.
The Siemens report was among 83 documents released Thursday by the House Energy and Commerce Committee, which held its second hearing on the ignition-switch recall last week.
The documents also show that Delphi Automotive, which made the faulty switches in Mexico, has turned over information to a federal grand jury investigating the recall. Valukas’ report said Delphi had refused to let its employees participate in GM’s internal probe of the matter.