Andrew Turner, Mikel Maron - Trends and Technologies in Where 2.0
As people are going out and gathering information on their own we’re collecting a lot of geo-aware data. This is becoming a really hot area. Nokia and TomTom just made big acquisitions. Every Web 2.0 service is starting to add location. You can start mining this information with tools like geocodr.
How do you start gathering them together? We started a company called geocommons. We’re taking this massive amount of data and trying to pull it all together. It’s an open database of freely available data with creative commons license. You can see where it came from and who posted it. You can search the data.
What about when your communities are supplying a lot of data? In Detroit the city is geocoding walking trails. With Hurricane season there are lots of people geocoding where shelters are. A local NYC company social lite is doing place marking with bars using mobile web. Android has a lot of applications which are innovating on the geo aware capabilities of the phone.
Mapvertising is one way in which people are trying to make money in this space. But it’s hard. You don’t want to do a search for a romantic restaurant near you and get back a Hooters advertisement.
Once people are sharing all of this there is a problem of privacy. Flickr is looking at casual privacy where you can set who is allowed to see where your photos were taken. Fire eagle is a location brokering system. If you trust Yahoo! they can be the trusted holder of your location. You can specify which sites get which granularity (only zip code, for example) of knowledge of your data.
NeoCartography sites like EveryBlock and is trying to focus on the data as opposed to the street. You can look at and understand the area based on data. OpenCycleRoute allows you to re-render a map with the best routes for bikes.
We’re launching GeoCommons Maker in a couple of weeks which allows you to create proper maps.
Burning Man Experiment
You may think it’s a bunch of naked hippies in the desert blowing things up, and it is, but it’s a whole lot more than that. You show up in a desert and within a week you have a miniature city. It provides a canvas for trying out upcoming geocoding technologies.
We collected over 100 gigabytes of data over the week. These are early results. Why is this important for 2.0 Expo? This is a look at what these technologies can enable. Burning Man Earth was a really interesting iteration in Where 2.0.
We took remote sensing data every day. We used pictearth.com, diydrones.ning.com, openaerial.com. We got a small plane with a camera. We were gifted 200 Gallons of Fuel. We took a flight path every morning. Really cool pictures of how the event evolves day after day.
Processing with ERMapper, ESRI, Photoshop, Sweat. You also are recording with a GPS device which gives you where each photo is centered. Takes a lot of sweat to get everything lined up.
In the Future we’re going to turn this into a social application.
OpenStreetMap (OSM) is a free map for the world like Wikipedia for maps. We export from GeoDjango to OSM XML. Import into OSM through its REST AP. Mapnik + mod_tile. The output is a tiled map. Flickr asked if they could take our tiles and use it for people to tag photos. Was really easy to do because we used basic tiles.
We need to start making our map tiles and our geotags time aware because the earth changes. By using OpenStreetMap you can get Garmin maps for free. There’s a freeware product called cGPSMapper. We used Garmin Rinos because they have radio built in so you can see where your friends are. At burning man it was incredibly useful. We also did vehicle tracking by sending packets over ham radio and APRS. Signal picked up by a digipeater which sends the data on. Some which take that positioning data and post it to the internet.
OpenViewProject.org – we were at WhereCamp at Yahoo and hacked Google street view’s data. Google sent a cease and desist. So a friend of mine bought a lot of gear and a tricycle so that he could do it himself.
Gigapans – a gigapan is a gigapixel image. Greater than 500 megapixels. NASA designed a little, sub $200 robot that captures panoramas and their software stitches it together. There’s a site Gigapan that allows you to view these massive photos and zoom in.
Kite Aerial Photography – you can script photos for cannons using their developer kit.
Google Earth & SketchUp models – Andrew Johnstone would take photos of art and texture models made in SketchUp.