Re: Hell, yeah!
@Joe Werner,
"If the car is not capable of real autonomous driving it should not be advertised as such. Though... automobil means self-driving... ;)"
Even though I'm writing from the eastern side of the Atlantic, I'm quite impressed by the manner in which the State of California is handling these things. They're being permissive, sensible and, in the interests of the public good, firm.
"Wanna test a self-driving vehicle? Go ahead, but here's the rules. And we'll publish the test results too, even if you don't like them". All in all, highly laudable. Someone has to stop self promoting mega rich inexperienced young hipsters from doing/selling dangerous things on a large scale to the detriment of the public, and at least on the topic of self driving cars the State seems to be doing that, and setting a good pattern for regulatory environments the world over. And by letting said hipsters learn from their mistakes, perhaps they'll turn into wiser, if poorer, adults. Everybody wins!
And now they're forcing some naming sanity on the industry too, quite quickly too after the accidents involving Teslas.
Rumour reported elsewhere on The Register has it that Google's self driving car team is running into team problems following the unsatisfactory results that CA published. Not surprising, a big group of exclusively young engineers are given a large pot of money to go and do a "cool thing" are being shown the error of their ambition by their own results and don't know what to do about it. Well, tough, that's life, get used to it.
Successful engineering (and I am a successful engineer) is all about brutal honesty with oneself and with others. Unless you absolutely know for certain that something is going to work, then it is merely speculative experimental development that may, if one is fortunate, produce some interesting results, but it is not a sound basis for an entire corporate strategy on. If you're not brutally honest with oneself, Nature (who as Feynman pointed out cannot be fooled) will be brutally honest with you on your behalf. That's what's happening at Google and Tesla (and probably Uber in due course too).
So Google and their self driving car? Huh. Count the corner cases involved in a driving algorithm. Can't count them? Well then, you can't write the algorithm. You can't even completely write down in detail what "driving" is. Ergo, no self driving car will happen. Don't even bother to start up the development program. It's less fun that way, but it does mean that you will end up doing a different project that can work and you will meet with success sooner.
SpaceX, according to rumour, are somewhat similar. Their only merit is that they're "cheaper". But that means they think that all the other launch companies that have ever existed have been needlessly pissing large quantities of cash away on some pointless things. One wonders what those things might be. Afterall, things like workforce size/reward/morale, quality control and profits for shareholders are just as vital to a successful endeavour as a good design and advanced manufacturing techniques. You can't take away any of those things and still have a successful business. And for SpaceX, not even the fabled re-usability is a massive cost saving; practically all you save money-wise is the cost of the raw materials. If you cannot achieve a high re-use count, everything else (the staff, procedures, facilities, supply chain, etc) has to stay in place and remain competent regardless, and that's where the bulk of the money goes.
"Tbh I like them Teslas and such, but they are still scary: silent cars. Never realised I rely on my ears that much..."
Me too, having been in a friend's one. Nice performance, terrible range (for my requirements), though I don't like the size of the display screen; far too large.
Autopilot is a needless distraction for Tesla that may have helped them sell a load of cars but could end up costing them dear. Also, judging from the reactions on social media about Teslas, I'd judge that a lot of their customers are hipster types more interested in the Autopilot than the battery power (though the torque is addictive in the P models). If Autopilot stops being a thing that sells cars, Tesla may lose a bunch of customers and find they've overreached themselves.