The question is how many total miles of streets are there in the U.S. and how many vehicles would it take to cover them over what period of time in order to make this effective enough to be used in an application or product in a reasonable time frame? And does Apple have that number of vehicles out there?
Assuming that each vehicle only has to traverse each street once regardless of how wide the street is, I estimate it would take 5.3 weeks for one vehicle to cover the streets of New York City, not including highways. That's based on 6400 miles of streets and a vehicle having an average effective speed of 20mph capturing images 10 hours a day, 6 days a week with 100% efficiency: never traveling on any portion of the same street more than once.
And that of course doesn't include returning to streets on a regular basis, especially in commercial districts, to record changes in the streetscape. And that's only the U.S. Apple really has to cover every major locale in the world.
So I have to wonder whether this is a real effort to gather 3D mapping info comprehensively or whether this is some kind of test.
Apple should have set up a shell company to collect this data 5 years ago. Starting now?
Today's technology is likely more capable than what was available 5 years ago for this type of 3D mapping. Remember, there weren't even any consumer-grade drones in 2010.
It makes more sense for Apple to prioritize the accuracy of their 2D map data before tackling 3D imaging, particularly at street level. Apple Maps (with their own proprietary dataset) is just over two years old.
Plus, how are you going to generate accurate 3D data if your own 2D data is untrustworthy? Apple has had to work very hard to clean up its 2D data deficiencies in the past two years.
Barring some sort of major breakthrough, Apple Maps may always be playing catch-up to Google Maps, but it's not impossible for them to get very close. After all, Apple Maps are now better than veterans Mapquest and Yahoo Maps by a fair margin.
Apple should have set up a shell company to collect this data 5 years ago. Starting now?
In an ideal world. perhaps. But some things are long term. I'd say map services are and Apple maps is very good for many purposes right now - in fact I use it 98% of the time. I use the satellite view much more often than I ever did street view in Google maps because Apple satellite view usually serves the purpose I need (verifying locations).
Apple is not usually concerned with being first but rather with being best. In the case of self driving cars - if Apple ever brings one out it will not be the first but it will be one that looks fantastic, works well, when the market is mature enough to make an entry worth while.
Originally Posted by zoetmb The question is how many total miles of streets are there in the U.S. and how many vehicles would it take to cover them over what period of time in order to make this effective enough to be used in an application or product in a reasonable time frame? And does Apple have that number of vehicles out there?
My guess is that Apple will focus on the most heavily trafficked areas. They know where & when their devices are being used. It makes more sense to street map Times Square in NYC versus some dirt road in Montana.
Street mapping the entire U.S. doesn't make much practical sense. It would be more important to capture the top 90% of passenger-miles/pedestrian-miles and focus on resurveying the top 10-20% on a regular basis.
Note that Apple needs to consider the usage of foreign visitors, not just U.S. residents. If international tourists are visiting certain spots, recent/accurate street level data for those sites would be particularly useful.
I think a van covered in cameras should have some identifying marks on them... seems awfully sneaky to be driving an anonymous van around recording everything it sees.... Google StreetView cars are highly visible. How about the other companies?
Kind of like those scanner-bots in the movie Prometheus - send them out, let them criss-cross over a range of terrain then return to home.
Maybe not cheaper at first, but you're going to be doing this repeatedly over time.
Then again people would freak out at the idea of "iBots" flying over streets collecting data.
Just a matter of time.
Once drones are legal, perhaps, until then they'd be VERY expensive... OTOH for street level data collection a ground vehicle makes more sense than a flying object who isn't going to legally EVER be allowed to fly down Fifth Ave, say, at seven feet off the asphalt...
Of course for Flyover data updating to keep it more current they'll be excellent, far cheaper than manned aircraft.
I think a van covered in cameras should have some identifying marks on them... seems awfully sneaky to be driving an anonymous van around recording everything it sees.... Google StreetView cars are highly visible. How about the other companies?
Let's rein in this posse before it hurtles off into the canyon of unsupported assumptions. Maybe I missed it, but where in this article is there an authoritative source that this van is connected to Apple? Where is the proof that these vehicles are leased to Apple? What if it's updated Street View equipment for Google?
Let's rein in this posse before it hurtles off into the canyon of unsupported assumptions. Maybe I missed it, but where in this article is there an authoritative source that this van is connected to Apple? What if it's updated Street View equipment for Google?
The vans are registered to Apple. Has that not been proven? Maybe not.
The vans are registered to Apple. Has that not been proven? Maybe not.
The now-departed City News Bureau of Chicago was a great training ground for aspiring journalists. The longtime editor famously berated his cub reporters: "I don't care if your mother tells you it's so ... check it out!"
Apple should have set up a shell company to collect this data 5 years ago. Starting now?
Google Maps started in early 2005 and started its Street View project about two and a half years later. Apple Maps started about two and a half years ago... and its starting its own Street View project now. Maybe there's a natural timeline to these things that armchair commentators on the internet aren't privy to because you aren't in this field of business.
Comments
5 years ago Apple was caught off guard by Giggle's betrayal.
They're slowly providing superior alternatives to Giggle one step at a time. Which I think gives them more focus.
Apple should have set up a shell company to collect this data 5 years ago. Starting now?
I'm curious, how do you know when they started?
Quote:
Apple should have set up a shell company to collect this data 5 years ago. Starting now?
Why didn't you just come out and say APPLE IS DOOMED!?
He's going to wait 5 years...
(and I think the Google van looks like the Monitor)
Wish Apple van looked like the Merrimack. It'd be tense when the vans passed each other.
Apple should have set up a shell company to collect this data 5 years ago. Starting now?
Today's technology is likely more capable than what was available 5 years ago for this type of 3D mapping. Remember, there weren't even any consumer-grade drones in 2010.
It makes more sense for Apple to prioritize the accuracy of their 2D map data before tackling 3D imaging, particularly at street level. Apple Maps (with their own proprietary dataset) is just over two years old.
Plus, how are you going to generate accurate 3D data if your own 2D data is untrustworthy? Apple has had to work very hard to clean up its 2D data deficiencies in the past two years.
Barring some sort of major breakthrough, Apple Maps may always be playing catch-up to Google Maps, but it's not impossible for them to get very close. After all, Apple Maps are now better than veterans Mapquest and Yahoo Maps by a fair margin.
I predict 20 billion once the ?Watch ships. But who's counting?
Apple should have set up a shell company to collect this data 5 years ago. Starting now?
In an ideal world. perhaps. But some things are long term. I'd say map services are and Apple maps is very good for many purposes right now - in fact I use it 98% of the time. I use the satellite view much more often than I ever did street view in Google maps because Apple satellite view usually serves the purpose I need (verifying locations).
Apple is not usually concerned with being first but rather with being best. In the case of self driving cars - if Apple ever brings one out it will not be the first but it will be one that looks fantastic, works well, when the market is mature enough to make an entry worth while.
5 years ago Apple was caught off guard by Giggle's betrayal.
They're slowly providing superior alternatives to Giggle one step at a time. Which I think gives them more focus.
Hehehe... Giggle.
The question is how many total miles of streets are there in the U.S. and how many vehicles would it take to cover them over what period of time in order to make this effective enough to be used in an application or product in a reasonable time frame? And does Apple have that number of vehicles out there?
My guess is that Apple will focus on the most heavily trafficked areas. They know where & when their devices are being used. It makes more sense to street map Times Square in NYC versus some dirt road in Montana.
Street mapping the entire U.S. doesn't make much practical sense. It would be more important to capture the top 90% of passenger-miles/pedestrian-miles and focus on resurveying the top 10-20% on a regular basis.
Note that Apple needs to consider the usage of foreign visitors, not just U.S. residents. If international tourists are visiting certain spots, recent/accurate street level data for those sites would be particularly useful.
I think a van covered in cameras should have some identifying marks on them... seems awfully sneaky to be driving an anonymous van around recording everything it sees.... Google StreetView cars are highly visible. How about the other companies?
Apple should have set up a shell company to collect this data 5 years ago. Starting now?
Noticed now.
Seems like drones would be cheaper.
Kind of like those scanner-bots in the movie Prometheus - send them out, let them criss-cross over a range of terrain then return to home.
Maybe not cheaper at first, but you're going to be doing this repeatedly over time.
Then again people would freak out at the idea of "iBots" flying over streets collecting data.
Just a matter of time.
Once drones are legal, perhaps, until then they'd be VERY expensive... OTOH for street level data collection a ground vehicle makes more sense than a flying object who isn't going to legally EVER be allowed to fly down Fifth Ave, say, at seven feet off the asphalt...
Of course for Flyover data updating to keep it more current they'll be excellent, far cheaper than manned aircraft.
Thank you for pointing out "Driver" in the photo. That was very helpful. (sarcasm)
hahahahah - but to be fair - i guess that point is only made to show the autonomous driving to be void in this model car.
Let's rein in this posse before it hurtles off into the canyon of unsupported assumptions. Maybe I missed it, but where in this article is there an authoritative source that this van is connected to Apple? What if it's updated Street View equipment for Google?
The vans are registered to Apple. Has that not been proven? Maybe not.
I called it 3D mapping yesterday. Do I get a gold star beside my name?
Vector-based street view?
Apple should have set up a shell company to collect this data 5 years ago. Starting now?
Google Maps started in early 2005 and started its Street View project about two and a half years later. Apple Maps started about two and a half years ago... and its starting its own Street View project now. Maybe there's a natural timeline to these things that armchair commentators on the internet aren't privy to because you aren't in this field of business.