Uber and car makers jockey for Nokia's 'HERE' maps – report Transport's enfant terrible, Uber, has reportedly put in a bid to acquire Nokia's HERE mapping unit. The New York Times spoke to an anonymous troika of insiders, who whispered that a bid of up to $3bn had been tabled. Three billion simoleans wouldn't be a big payday for Nokia, which bought HERE for $8.1bn in 2007 but has …
I don't understand why that would be a threat. Cars aren't commodity items like PCs and never will be. Today the automakers do very well selling cars that can't drive themselves. If everyone used a freeware "OS" for the car they'd compete in exactly the same way as they do today.
The only threat is that if one or two of the automakers got way ahead of the others in making self driving work well, but the freeware was "close enough" to the best software it would dilute the value of the investment those companies had made to get ahead.
I'd really hate to contemplate what you'd have to give to Google in terms of personal data or watching ads to get free self-driving software. Will they collect data on how many fast food drive throughs you go through for health clubs to push ads at you, while McDonalds tries to make sure your next drive through has a big yellow M over it?
Self driving cars would be a threat to auto manufacturers because not as many cars would be built- after a car has dropped you off, it would then go pick somebody else up. More people would travel in one car, because lift-sharing would be easier - unless of course you pay extra to avoid the fellow plebs. Fewer cars would be damaged in accidents. Cars would require less maintenance because they would have fewer cold starts, their engines would spend longer at their optimum rpm, and traffic control systems would eliminate start-stopping at traffic lights.
"More people would travel in one car, because lift-sharing would be easier -"
easier yes, but not necessarily desirable. While in some places where heavy traffic makes effective car-sharing a necessity rather than a personal choice, to others a car is mobile castle, full of bits they keep around them, a fortress of solitude on the way to work, and the ability to leap in one own's castle at a moment's whim and take off to do shopping is more of an advantage than the possible fuel savings of a shared journey.
Self driving cars would be a threat to auto manufacturers because not as many cars would be built- after a car has dropped you off, it would then go pick somebody else up.
Not this kid.... Have you actually looked inside some people's cars? There's whole civilizations growing in that back seat area that full of old fast-food joint trash, empty coffee cups and who knows what else.
... You so KNOW that the average person in India can't afford an Audi, BMW, or Mercedes-Benz but they find a Tata much more affordable? Google being first with the software would allow Tata to be first to open a whole new market in their home country and other parts of the developing world practically zero cost in R&D.
That was a typo but like your friend Murphy, I'm not bitter.
Tata have already sold a 0-star NCAP rated car in the west so there's enough of a market in the west that doesn't care. Plus the developing world that tend to be a little less demanding about safety (Honk OK please).
Most western countries have some minimal standards for crashworthiness that are increased over time. If Tata wants to sell into the US or UK, they'll have to make those changes anyway.
Just because a car was 10x less likely to crash than if I was driving doesn't mean I'd want to have one that would burst into flames if there was a crash. People will still care about safety since a crashproof car can never be built.
I'd really hate to contemplate what you'd have to give to Google in terms of personal data or watching ads to get free self-driving software. Will they collect data on how many fast food drive throughs you go through for health clubs to push ads at you, while McDonalds tries to make sure your next drive through has a big yellow M over it?
The bigger fear to me is a Google-car with AI. "It's been 3 hours since you ate.. I'll be pulling into the next McDonalds since they paid us for an ad or two."
Don't know why you're posting anon because you really dont know anything.
The greatest threat is that the tech used for self driving cars is still very immature and dangerous.
To give you a free clue... how do you keep a car within its lanes at all times? ;-)
You think that would be trivial and you'd be wrong.
Posted anon because unlike you... I know something about this as well as about Nokia (HERE). ;-)
"...if your car insides could be switched out by a robot..." -- phil dude
That's an intriguing idea: taking it further, we could all just have caravans with all the personal stuff we want in them, and when we need to go somewhere an autonomous tow-car turns up and takes us there. Autonomous vehicles, nobody owning their own motive power, and caravans everywhere! The entire infrastructure could be powered by harnessing the ensuing rage of Jeremy Clarkson.
sarcasm aside....one of the opportunities that arises with new technologies is the ways in which old things can be redesigned.
Perhaps not caravan style, more "doll house"?
And on autonomous vehicles liability will reside with controlling agent. I have written before , that I fully expect insurance for a person to be much more expensive that a vehicle on its own.
Just floating and idea, not writing policy....
The icon is an aspiration....
P.
I can see why the auto-makers would want control of their own maps but what could Uber do with HERE that would make it worth $3bn? All they need to do is display a map in its app and they can do that with a contract with Google, HERE, Bing Maps, anyone...
Unfortunately both would probably pull the free Android app, only Baidu would keep it going.
There will be a few niche markets for proprietary maps, just like there is a market for say QNX today. Then there will be the formerly dominant actors: Google Maps and possibly HERE. However, as maps turn into a commodity, much the same as operating systems have, there is no real incentive for companies to roll out their own solutions--instead, everyone will use, and adapt, OpenStreetMaps¹ for their own purposes (just like every gadget that need an OS uses Linux today).
You may say: "Ah, but quality control! Pranksters!", etc. That is not a significant problem in practice--it has been known since Ptolemy's days that maps are imperfect², incomplete representations of the Earth (and other celestial bodies nowadays), and that knowledge is built into the solution. E.g., a self-driving car uses maps for strategic planning (e.g., how to get from Glasgow to Edinburgh) not for the actual driving which is primarily based on real-time information gathered from on-board sensors. A sufficiently advanced car system (not an individual car) can even improve and update the map in near real time, e.g., to account for road modifications, long term roadworks, etc.
In short (I have a chatty crystal ball): proprietary maps are not dead... yet. Let us come back and talk in 2030.
¹ If you haven't used it, or haven't seen it recently, I suggest you take a look. And if you are a CS type, read up on its architecture.
² Even the best commercial maps have inaccuracies deliberately put into them as a form of water marking, to detect plagiarism.
Why have a 'crystal ball' when you work in the tech.
Yes, posting anon for the obvious reason...
You need an accurate map.
There are 3 companies... google, HERE, and TeleAtlas which is owned by TomTom if memory serves.
You do remember the issues Apple had with their maps? That was TeleAtlas data. So if you were in Australia... you may get sent 50 clicks in to the middle of nowhere.
So its not just road links, but addressing and then the POIs placed on top of the map.
There's a lot more but the self driving car does need accurate maps as well as a good clock and decent GPS among other things. There's more, a lot more which means that trying to have a driver less car means you should end up taking the bus. ;-)
> Why have a 'crystal ball' when you work in the tech.
As it happens, my background and former career is closely related to this. :)
Just a couple clarifications.
> You need an accurate map.
You need:
a) Sufficiently accurate (and precise) cartographic information (of which a map is just one possible representation).
b) A way of quantifying said accuracy (and precision) in order to verify that your requirements are met.
> So its not just road links, but addressing and then the POIs placed on top of the map.
That's cartographic information.
> the self driving car does need [...] a good clock and decent GPS among other things.
By "a good clock" you mean one with desirable long or short term drift qualities? I do note that GPS is a timing and ranging system, which means that wherever you have GPS you have excellent timekeeping (except at the submilisecond level).
> There's more, a lot more which means that trying to have a driver less car means you should end up taking the bus. ;-)
The immediate automotive industry goal is not to have driverless cars, but self-driving cars. Not the same thing at all. And I do note that neither current nor planned¹ self-driving cars rely on external inputs for piloting (as opposed to navigation): all the information comes exclusively from onboard sensors.
Not sure whether you meant to imply that you work on the mapping or car industry, but here is what I think is a decent article (much better than what can be found at El Reg in any case) about self-driving cars: http://www.engadget.com/2014/12/18/audi-self-driving-rs-7-concept-test-drive/. Be sure to read the full thing.
Going back to the original point: I speculate that your 2030 model car (or closest French equivalent) will be using OSM data for navigation, not Google Maps, HERE, or anything of the sort. Being pure speculation, I may of course be wrong, but we'll see in fifteen years.
¹ On the planned front, inter-vehicle communication systems would act as augmentation systems that enhance piloting capabilities, but they are not required by any self-driving car project that I am aware of.
Posted Anon for obvious reasons...
There are a lot of rumors floating around.
First, HERE aka Navteq made money from maps. The sucking sound is the cost of the R&D once you get past the core maps. (Depending on where the money is being spent. )
LIDAR and 3D is one area. Faster map generation and updates is another. Redesigning how the map data is stored and segregated is another.
Then you have special projects where HERE has partnerships with auto manufacturers. Lots of proof of concepts have been going on. ;-)
The truth is that FB could shell out $$$$ because they want maps for mobile so they can compete w Google and keep your eyes on their web real estate longer. (And to capture more data about your)
But as long as they can guarantee the map data is available, they don't need to buy the company.
The automakers... ;-)
Free clue. BMW opened up an office in the Boeing Building in Chicago. Care to guess who else is in that building?
My bet is on the auto manufacturers.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
