Comments on: Map making and the Battle over Local Search http://garrygolden.net/2007/08/02/map-making-and-the-battle-over-local-search/ ---A Futurist looks at changes ahead, and reasons to be hopeful------- Sat, 30 Aug 2008 14:45:28 +0000 http://wordpress.org/?v=2.1 By: Jeremy Guillory http://garrygolden.net/2007/08/02/map-making-and-the-battle-over-local-search/#comment-90 Jeremy Guillory Mon, 06 Aug 2007 21:30:12 +0000 http://garrygolden.net/2007/08/02/map-making-and-the-battle-over-local-search/#comment-90 Excellent post, Garry. As I read this article, I got an email confirming the delivery of a Bluetooth GPS receiver that I will use with my Palm Treo. After using Google Maps on the Treo, I realize that mapping is essential on mobile devices. I can’t imagine buying phone without maps and local search, and I don’t understand my parents when they say they don’t need it. The customer needs for precision indoor and urban location will likely drive extensions of GPS to get that functionality. Assisted GPS (AGPS) addresses these needs, but imperfectly. Historically, one of the bottlenecks with AGPS has been processing power, which is a problem that continues to go away. If we were designing a localization system now from the ground up, it would be tower/base-centric rather than satellite-centric. It’s cheaper, and all of the money is in urban and suburban areas where leverage for towers is high and strong infrastructure exists. Cellular frequencies, WiFi, cameras (on mobile phones) and the upcoming 700MHz band are all options for new localization hardware. Placelab.org has an interesting WiFi solution for localizing. Multiple phones trying to localize in the same area could increase their accuracy by sharing information with each other or the base station (good Bayesians never throw data away). Excellent post, Garry. As I read this article, I got an email confirming the delivery of a Bluetooth GPS receiver that I will use with my Palm Treo. After using Google Maps on the Treo, I realize that mapping is essential on mobile devices. I can’t imagine buying phone without maps and local search, and I don’t understand my parents when they say they don’t need it.

The customer needs for precision indoor and urban location will likely drive extensions of GPS to get that functionality. Assisted GPS (AGPS) addresses these needs, but imperfectly. Historically, one of the bottlenecks with AGPS has been processing power, which is a problem that continues to go away. If we were designing a localization system now from the ground up, it would be tower/base-centric rather than satellite-centric. It’s cheaper, and all of the money is in urban and suburban areas where leverage for towers is high and strong infrastructure exists. Cellular frequencies, WiFi, cameras (on mobile phones) and the upcoming 700MHz band are all options for new localization hardware. Placelab.org has an interesting WiFi solution for localizing. Multiple phones trying to localize in the same area could increase their accuracy by sharing information with each other or the base station (good Bayesians never throw data away).

]]>