Embedded Expert: No Pedal Misapplication in Toyota Case

Sounds like a classic case of wooly thinking and over complexity on the part of the software development team. It is self evident that if it took that long to plough through the code and the multiple failsafe, then the software is badly written over complicated and poorly documented. It shouldn't be that complicated.

Just like it said... a simple brake priority over throttle should be pretty much common sense. Unless you are a rally driver and want to heel and toe... but then you can have you chaps re-write the code
Del

"then the software is badly written over complicated and poorly documented."

Let alone, badly tested!

Poorly designed hardware/firmware caused unintended operation, lack of driver training made it fatal.

Unintended consequences arising from 100%-untested codings.

If the code isn't 100%-tested--for EVERY possible contingency--it is "untested."

If the code isn't 100%-tested--for EVERY possible contingency--it is "untested."

Exactly and that's why simplicity is vital.
To test that something does all the things it should do is relaively easy.
But to test it doesn't do all the things it shouldn't do is an almost infinte test set, unless it is very simple.
After all that's why we have modular software and suchlike.
The problems arise when we rely on mathematical models/statistical methods etc to give us test methods and data sets.
You can't beat a good old fashioned curious engineer/technician or idiot to test something... what happens if?I so often meet cases of complete overthink with software.And as engineers I'm sure we all know...
If it can go wrong it will go wrong...
Usually the day after the field trial endsI may be buying a car soon... I will be going for one with as little electronics and software as possible. I shall certaily want one with a mechanical door and ignition switch, I've learned this from bitter experience.Del

Pilots should still be able to, and know how to, fly by wire if needed. My trust in computers will never approach 100%.

You completly miss the point here.

An automobile DOES NOT NEED electronically controlled this and that and thus and such.

Software smothware.

What's wrong with a throttle cable that controls the throttle and an ignition key that controls the ignition?

Nothing, until you try to add engine management systems for fuel economy and traction/stabilization control systems where eGas provides advantages.

While my personal preferences lean heavily toward simplicity, there are times where a greater degree of sophistication is not only desired, but warranted.

Maybe you might be able to relate the latest incarnation of the Porsche GT3 track car to this debate. While I may like the earlier version with a manual transmission for its simplicity and fun factor, the 2014 version with the PDK transmission is significantly faster with that transmission than without.

Or, take the transition from old carburetors to EFI. The carbs were simpler, but the EFI beats carbs every time, short of a nuclear winter.

So, while cars do not need EFI, eGas, or smart transmissions, those new cars that are embracing that technology are performing longer, better, and cheaper than their 1960s and 1970s counterparts. When is the last time you had to get a "tune up" for your modern car?

eGas is not necessarily evil, but Toyota's earlier application of that technology was poorly designed and executed.

Granted, cars are more efficient now, but, my last tune-up cost me $800.00 USD.

Throttle and ignition do not have to be electronically controlled to take advantage of engine management systems for fuel economy and traction/stabilization control systems.

If there's no positive, intuitive way to turn the engine off, the car is potentially dangerous. I understand that with the engine off things like steering and braking may become more difficult, but one can still stop the car, eventually.

Modern stabilization systems no longer rely on simply applying brakes to some wheels, but also incorporate throttle adjustments as part of the equation.

At some point there will be more systems that distribute the power to wheels independently depending on conditions as these systems evolve.

Actually, turning the car off by ignition would also lock the steering wheel. Smart systems could override that.

Yes, you could manipulate the switch after it turns off the engine, but the success rate of that under a panic situation is probably close to zero as no one practices that kind of emergency maneuvers anyway.

I doubt we will ever see a engine kill switch on the dash as that sends a negative message to potential buyers.

"Actually, turning the car off by ignition would also lock the steering wheel. Smart systems could override that."

Actually turning the ignition switch to "off" does not lock the steering wheel. That is accomplished by turning the ignition switch to the "lock", full off position.

Granted, in a panic, this distinction may be lost.

I stand by my statement that we've gone too far with drive-by-wire.

I had my fly by wire definition bass ackwards. I was thinking it was the old fashioned cables.

I wonder how you would define "flying by the seat of your pants"?

I can't define it, but I've been doing it for a long long time.

In all the cars I have, the steering doesn't lock until you pull the key out, probably just for this reason.

Sometimes they go overboard, just for the sake of doing it.

I would have been fine with hand crank windows, for example. I hang onto my 86 truck because I know that I can keep it going with my existing skill set, and it's cheap to fix. The gas it drinks isn't cheap though.

My pet peeve is electric windows and door locks.

For two reasons:

1. No battery, no windows, at all. No up, no down. Locks can still be opened, if you have a key, oh unless it's one of those idiotic keyless cars, then you're screwed.

2. The cost to repair either is obscene.

I've changed out electric window motors, and it's no fun at all. Cussing and scraped knuckles from the start. Even my truck has those.

Yep. In an emergency, if power is lost, escape becomes a lot more difficult.

That's an irresponsible statement.

You have no way of knowing the circumstances surrounding the accident.

The evidence stated that there was no pedal misapplication.

One should not have to be a software expert in order to drive a car.

Vehicles today are unnecessarily complex and many of these "improvements" are accidents waiting to happen.

Well said... I expect it won't be long before someone burns to death because they can't get out of the car because of these daft 'lock the doors as soon as you start off' systems....
Yeah yeah yeah... I know why they are there...
Del

My 86 truck has a cable running from the gas pedal to the throttle. When I release the pedal, a spring brings the throttle back to idle. So far, so good.

Some carbs I have seen, have TWO return springs to provide redundancy. One was a torsion spring on the shaft(s) and the other was a familiar coiled tension spring.

I think I've got a throttle return on the carb, and the gas pedal itself has a return spring.