Slashdot Mirror
BMW To Compete With Google To Build Software For Self-Driving Cars (reuters.com)
SmartAboutThings writes: Google is about to get some serious competition in the self-driving car race from none other than BMW, one of the most prominent names in the car industry. Speaking to Reuters, BMW's Head of Research and Development, Klaus Frohlich, said the following while present at the Geneva Motor Show: "For me it is a core competence to have the most intelligent car. Our task is to preserve our business model without surrendering it to an internet player. Otherwise we will end up as the Foxconn for a company like Apple, delivering only the metal bodies for them." BMW believes its competitors in the future will include internet taxi service Uber and sales website TrueCar. As the company is approaching its 100th birthday, the company is now on a quest to build the "ultimate driving machine." The company is preparing for a world in which its customers will be mere passengers, and the cars will do the driving themselves.
90% of the reason I buy BMWs is because they ar enjoyable to drive. If it drives itself then they are alienating their fan base, who may as well buy from other manufacturers.
Google is about to get some serious competition in the self-driving car race from none other than BMW
Google has been working on this AI problem since probably 2008 or so and been road-testing self-driving cars since 2011. As far as I know BMW has no development at all on this concept.
What makes BMW think they have some great insight into the artificial intelligence problem, that will make it possible for them to emerge as a serious autonomous vehicle competitor today, if Google is already 5 years ahead in learning and development of the software, data collection, and STILL does not believe their fully autonomous vehicle can be fully productized until 2020?
Even Tesla has a head start on BMW.
Autonomous driving is not a simple solved problem. It's not like you can just slap a clone together and have it out within a year or something....
Research and development is an extremely long process, there, and as far as I know BMW's R and D is focused on building cars.... they probably don't have people who even know about robotics, let-alone AI for the self-driving problem.
I say this as somebody who owns and loves BMWs... they suck ass at software. They can't even get things like their console information / entertainment system to be reasonably bug-free, so how the hell do they think they can actually do difficult and important stuff?
Help save the critically endangered Blue Iguana
BMW will add some cool features that the GCars won't have:
1: A weighting for their e-cars to slide into narrow gaps between cars in traffic, then panic stop.
2: No turn signals.
3: Auto detection of parking line spaces, and automatic parking to take up as many as possible.
4: A middle finger sign that pops out, when a lane change is successfully executed.
5: e-Ink on the sides of the vehicle to mimic being keyed by unhappy people after #1-4 are done.
Look at Here Maps (maps.here.com) , and there offline app navigation. Really you'd think nobody could compete with Google maps, and yet Here is far superior.
Less worthless bells and whistles and noise and more the "where is X" fundamentals of maps.
BMW has shown self driving prototypes since 2011, European cars already have features like lane follow, auto park, sign recognition, collision avoidance and so on, so they're already rolling out key components in existing cars.
My expectation from BMW is that it will make cars that drive from A to B with cameras watching the roads.
My expectation from Google is that they will make cars that drive from A to Sponsored C to Sponsored D to B with cameras watching me.
If you think that's a joke, look at what their glorified Android radio does (it requires all the internal engine data and position, be sent to Google server), so if you're driving a car with a Google android 'radio/navigation' its tracking you and every detail of your driving even when its off.
"...delivering only the metal bodies for them."
No, they won't even be doing that- Google & Uber will not want their metal body or their programming expertise. The mystique of the quality German car quickly evaporates when you own one or crunch the numbers on the cost of purchase, maintenance, parts and accessories.
I'm on my third BMW motorcycle; each new purchase 20+ years after the previous. During that time I've owned a couple dozen other motorcycles. Like the Harley, BMW motorcycles have made only minor changes over the last 100 years, and like the Harley they still can't get it right. Meanwhile the Japanese can snap their fingers and come up with a totally new design that is almost flawless and it's relatively affordable.
It's not just motorcycles; Consumer Reports can't find a single excellent car in the 2015 BMW lineup. It's sad, I'd love to own a vehicle that actually lived up to such a reputation but they don't seem to come from Europe or America.
...omphaloskepsis often...
Yes, I was also wondering if BMW self-driving cars would also drive like assholes.
You are welcome on my lawn.
So its BMW's programming team verses Google's, I know where I'm placing my money...
for driverless cars to become popular. I'll be able to take advantage of their dialed-up safety protocols and be able to cut my way through traffic faster with less risk.
And as an avid driver, I can tell you that avid cyclists are the biggest assholes on the road, by far. It's not even close.
You are welcome on my lawn.
> or a gun person (amazing how those seem to go hand in hand) then you have to make stupid comments about how they'll never be able to make a smart gun that works
Your ignorance of the topic is glaring. I'm making this post via my smart phone. Often, as I post on Slashdot, my smart phone suddenly hides the keyboard, so that as I begin to press the "h" or "j" keys, they vanish and the press instead goes to the Submit button which was underneath. I end up posting half a post because the smarts in my smart phone decided it was time to remove the keyboard from my screen.
Sometimes, my smart phone phone randomly locks up for a few seconds, which is no big deal because it only lasts maybe ten seconds before it starts working again. That kind of thing happens with smart devices.
In 2011, the model 1911 (introduced in the year 1911) was still the world's most popular handgun because it's been proven reliable (partially because it has only about foir moving parts that are critical). When your smart gun locks up for ten seconds, that's enough time for the bad to put ten bullets in you.
In a gun fight, a snafu with your hardware means you are dead. We carry the hundred-year-old 1911 model because reliability means life. A bug, an imperfection, an "undocumented feature" with your handgun means your kids no longer have dad, so that's why we're not interested in nifty new designs with cool new features.
The upgrade cycle on handgun designs for serious users is about once every hundred years for this reason. After 30 years of torture-testing the Glock design, including testing it after burying it in sand, in swamp, etc., we've started to trust that it's reliable, and it's taking it's place as the handgun design for this century. We pay $650 for a Glock rather than $250 for a very similar Kel-Tec model because we trust the Glock's reliability just a little bit more. In defense guns, reliability is life, so we'll absolutely pay more than twice as much to get something 2% more reliable.
Go bury your smart gun in wet sand, in a swamp, drop it down the stairs, then test it. Keep doing this for 30 or 40 years and you'll meet the testing standards we've always required of any new design.
To illustrate a bit further for those who don't "get it", virtually every gun made today, from pistols to machine guns, is a John Moses Browning design. People have tweaked things here and there, adjusted the size and shape, but they mechanisms for all types of guns are John Browning's designs. Second place is Gaston Glock, for figuring out which plastic could be used to make a frame for a Browning breach-lock design.
Basically, John Browning is the one guy we trust with our lives. We trust his designs. When he designs a smart gun, I'll buy one. That really is holding "smart" guns to the same standard we hold traditional guns - we trust their design only if John Browning signs off on the design. Of course he's been dead since 1926, so that's unlikely.
Actually any "smart" gun would almost have to be the opposite of Browning's genius - he's THE gun designer because he was a master of simplicity, of having the fewest possible moving parts. Simplicity, few parts to go wrong, is what makes his designs so reliable. That's the opposite of "smart".
Your smart thermostat is so complex and therefore difficult to FULLY understand that it can't be made secure or reliable; they get hacked and crash. That may be okay for a thermostat.