If Your Product Has a Problem, Please Don't Wait for People to Die to Report It

When is it time to notify the public that there is a potentially serious flaw in your product? If you've read even one of these posts before, you'll know that my answer is: Immediately.

But that almost never happens, right?

Another "great" example has just surfaced in the last two days, as General Motors doubled the size of a recall -- to nearly 1.4 million cars in the US alone -- for an ignition switch defect.

As reported yesterday by The New York Times' Christopher Jensen (full article, here), "The recall is aimed at vehicles with ignition switches that could inadvertently turn off the engine and vehicle electrical system -- disabling the air bags -- if the ignition key is jarred or the vehicle's operator has a heavy key ring attached to it."

Quick: check how many keys are on that same ring with your car's ignition key.

The cars being recalled are all no longer in production: 2005-2007 Chevrolet Cobalts, 2007 Pontiac G5 models, 2003-2007 Saturn Ions, 2006-2007 Chevrolet HHRs, 2006-2007 Pontiac Solstices, and 2006-2007 Saturn Sky models.

But what's particularly distressing is the speed, or lack thereof, in making the recall happen. In the interim, there have been 13 fatalities.

According to a USA Today timeline (full story by James R. Healey and Fred Meier, here), GM first became aware of the problem in 2004, before the first model rolled off the assembly line:

GM concedes it knew in 2004, before launching the 2005 Chevrolet Cobalt, that the ignition switch might inadvertently move from "run" to "accessory," stalling the engine and cutting power to safety systems. A company engineer had the problem testing the soon-to-be-launched car and engineers proposed several solutions, but because of "lead time required, cost, and effectiveness of each" solution, none was adopted and the car went on sale with the faulty switch.

Really?!?! "Lead time required, cost, and effectiveness" are equal to the value of 13 lives lost?

Here's the rest of the timeline, as developed by Healey and Meier:

2005: An engineer proposes a "partial fix", but it's not adopted. Later that year, GM realized that "it wasn't just the Cobalt with the risk."

Engineers suggested a simple change in the key from a slot to a round hole to ease stress on the switch... but initial approval later was canceled. It did send dealers a bulletin telling them to modify existing keys with an insert and to tell owners to take extra items off their keychains -- but only if customers came in complaining of stalling problem on the Cobalt. 

2006: "GM OK'd a better switch, but didn't make it so mechanics could identify it." In other words, GM gave the better switch the same part number that the old switch had.

2007: GM "believes" that the new switch was used exclusively in new cars. And officials from the National Highway Traffic Safety Administration (NHTSA) provide GM with reports of the first fatal crash. "And only then does GM assign an engineer to track Cobalt crashes where the airbags fails."

2009: GM "finally" shifts to the switch design proposed four years earlier.

2013: "GM determines the original switch wasn't made right."

2014: GM announces a recall of Cobalts and Pontiac G5s, and then, almost two weeks later, doubles the number of vehicles to be recalled.

According to the Associated Press, NHTSA is now investigating whether GM acted "quickly enough" (story, as published by the New York Times, here); the agency "has the authority to fine GM as much as $35 million under legislation that went into effect late last year."

Clarence Ditlow, executive director of the consumer advocacy group Center for Auto Safety, told CBS News, "This is a case where both GM and NHTSA should be held accountable for doing a recall no later than the spring of 2007....GM did this because they had no other choice, the spotlight of public attention was on.GM is just cutting its losses, doing the right thing, but why were 13 lives lost?" (full story, here)

I'm still hung up on the lack of response to the engineering solutions proposed even before the models were first sold, but that no action was taken after the first fatality is appalling.