An Engineer’s Eureka Moment With a G.M. Flaw
nytimes.com
nytimes.com
I didn't know that turning off the ignition would deactivate the air bags. I've welded on vehicles before, and have talked to body work guys about it. Everyone says to disconnect the battery so you don't accidentally set off the air bags. It can happen by stabbing the mig wire into the air bag harness, or just by letting the welding lead lay next to it and inductively causing current to flow.
Seems like if you're parked in a parking spot, and someone smashes into you hard enough that the acceleration can trigger the airbags, you would want the airbags to go off. I don't know why you would ever want to disable the airbags (except baby in car seat).
Plenty of ECUs have been ruined by folks getting their exhaust welded when the people doing the welding don't really know what they're doing. Especially if they normally work on older cars which tend not to have sensitive electronics.
For the airbag I think the sensors are passive, but even if they aren't, a grace period could be added (like, it will shut down after someone shuts off the car, then opens and closes the door)
Bluetooth devices actively searching for devices to pair with could plausibly draw significantly more power than that.
[1] http://www.analog.com/static/imported-files/data_sheets/ADXL...
There are pressure sensors, such as those used to determine if the passenger airbag should be enabled/disabled. That should be the reference for both driver and passenger; if seat = occupied, trigger airbag if collision detected, regardless of engine status, ignition switch position, etc
If only automakers could update this sort of logic over the air...
Even with physical keys I would assume that this particular circuit would be controlled via a relay and the car computer would only let the driver to stop the engine by removing the key if the car was not moving.
There are already cars where the ignition switch is completely wireless and software-controlled. For example, many cars offer a start-stop feature which turns the engine off when standing in red lights and automatically starting it again the foot moves from brake pedal to the accelerator. People are going to learn that there's no such thing as ignition switch anymore.
Motorcycles have red kill switches that shut down the engine completely. A school bus or a car could have something like that.
My car, for example, is a Chrysler 300 and there is no real connection between the key and what provides power to the engine.
Here[1] is a diagram showing how the CAN bus works on the Chrysler/Dodge LX platform vehicles. The SCM/SKREEM module on that diagram is where the ignition switch and logic that talks to the keyfob is and makes sure you have a valid key. Here's a recall related to the module which hints a little bit about what it is and does[3]. Note on the recall that this applies to a whole bunch of Dodge/Chrysler platforms.
Based on the wiring diagram[2] (for Dodge trucks, which is one of the platforms that [3] applies to), the SCM/SKREEM (called Node-Wireless Ignition in one diagram and WCM in another) uses a 2-wire CAN bus for its communication. Once the SCM/SKREEM validates the key it sends a message on the CAN bus to the engine controller to tell it to start the engine.
In these vehicles, whether the engine is on or off is purely under software control. That said, I'm not going to try turning my car off while I'm driving.
[1] http://the8thlayerof.net/2013/11/10/canbus-vehicle-network-d...
Since the part was changed on the later models (post '06-'07) without changing the part number, that means there is no way to tell the new ones from the old defective ones. This goes against commonsense, why was this done?
Almost certainly because they knew about their own enormous liability. They settled with a family for this issue in February of 2006, and not long after did the engineering change, albeit leaving the defect out there on millions of cars (insert Fight Club quote).
I suspect this issue is going to continue to grow until someone is charged with criminal negligence causing death.
Which is worth noting won't happen; not under this administration with Holter behind DOJ's steers. Keep in mind, GM was bailed out and saved from bankruptcy by taxpayers money, despite their uproar. Clearly, GM is in bed with administration.
Others have speculated it was done for liability reasons, that's IMO the most likely explanation. The beauty of civil litigation is there's an easily accessible paper trail. It would be interesting to see whether or not GM has a habit of changing designs without changing part numbers.
However, frequently designs get changed by "marketing" without a part number change, even in situations without possible liability. E.g. wireless routers are sometimes totally revised while retaining the same part number. Apple also likes to implement stealth changes.
Most engineers understand "commonsense", most marketing pukes don't.
But- Are the deaths due to the drivers not adapting quickly enough to the loss of power steering? After all breaks and steer are still working?
I have driven a vehicle without power steering and it wasn't that bad, and the vehicles in question (Cobalts) are rather light.
Yes power steering loss makes the car very sluggish, but really you're just coasting at that point and turning the ignition switch quickly is second nature. I'm sure it could be a little more frantic at highway speeds.
I wonder if this is something that should be taught in driver's ed?
(Regardless of my above comments, I do agree GM is responsible for sweeping this under the rug)
Is this use of "because" correct? Sure, why not. He wouldn't have died if his bonus had been paid? Might well be true.
I would personally not be interested in "due to" and "because of". To answer your question, you could have asked instead,
"I wonder if experienced drivers who were used to driving without power steering would have not been adversely affected."
Note: my example is a case where blaming a lack of bonus on the death is clearly wrong - but isn't attributing causative power to someone not being used to a car suddenly changing into a totally different input paradigm equally wrong? Better to stick to the factual question for starters.
I think a lot of these accidents are indeed caused by drivers panicking from being completely out of touch with what the car is doing, but that does not mean there is a straightforward solution. Knowing what a failure of power steering/brakes feels like, I'd just coast over to the breakdown lane and stop with the hand brake/engine. On the other hand, I once had a passenger (in my old automatic transmission car) accidentally bump it into neutral when I was on the highway and it took me ten seconds to figure out what was going on.
Fortunately the road was straight and after about 3 second it came back on as I was trying to see where the edge of the road was, and no other cars were around (although that would probably have helped me see or regain perspective).
It's easy to imagine how I could have rolled off into a ravine regardless if I could forcibly steer the car.
Regardless, while driving without power steering and brakes is possible when you're expecting it, it's quite a shock when you aren't. Especially if you've never done it before, which I suspect is true for many of today's drivers.
I can easily envision circumstances where loss of power could cause an accident. Fast road, sharp corner, surrounded by woods, etc.
Most readers here can either afford a subscription, or know how to read a story-of-interest without sending paywall-triggering cookies.
Some great reporting entities, using information beyond that which we have as readers, have decided such leaky paywalls are better than the alternatives. Against that domain expertise, we don't need HN to adopt some uniform, ideological "boycott all paywalls" policy. If it's so easy to get the same good stories from elsewhere, find them and submit them.