Spaceland, 3-D Maps in the Browser

Yesterday, Microsoft released an amazing, computer-generated 3-D mapping site via the Virtual Earth team. Spaceland, as it's called, is quite beautiful and available in 15 cities. It can be used for driving directions, searching, and real-time traffic monitoring. It has an API. You can use an XBox controller to steer Spaceland; this works much better than a mouse and keyboard. There's a lot in this release and it's worth exploring on your own. The screenshot below shows downtown Seattle with ads.

Spaceland utilizes technology from the GeoTango and Vexcel acquisitions. Vexcel is a GIS data collection asset for MS. They use their advanced UltraCam to gather the data used in the models. This was the data collection part of the equation. That's right, MS is now gathering their own GIS data. What part of the data stack will be next?

Geotango had products at the time of the acquisition that allowed 3D models to be built based on photos and GIS data (such as rivers and roads). If you were at Where 2.0 in 2005 you would have seen GeoTango demo their GlobeViewer product, an ancestor of Spaceland (Radar post). If you would like more reading on these two companies and their original products Directions Magazine has two great articles: Microsoft and GeoTango and Google and Microsoft: Further Disruption Ahead (focusing on the Vexcel and Sketchup acquisitions).

As has been widely reported, Microsoft is experimenting with ads in its new 3-D GIS playground. These are supplied by Massive, a company that was originally serving ads in games and from Microsoft's homegrown ad platform. The ads are relatively large, but don't detract too much from Spaceland. These are still experiments and this team above any other at MS has shown the ability to iterate and improve. For this space to continue to grow and innovate they need to be able to find ways to pay for them.

Spaceland utilizes the .Net Framework to handle client-side caching and is currently Windows-only. Currently, it is only available for IE, but they are working on supporting FireFox. The lack of a Mac or Linux platform is disappointing and will limit a lot of people's willingness to share. The Mac's share is growing. If MS is truly going to win the internet game they need to recognize this.

The product to compare this to is Google Earth. Both use technologies that were acquired rather than homegrown. Both are 3-D mapping products that are providing amazing data and experiences to consumers via the internet and at great cost to the parent companies, but that is where the obvious similarities end. Spaceland requires an initial download, but is primarily in the browser (and Windows & IE only with FireFox coming soon), while Google Earth (acquired with Keyhole) is a separate application available on Windows, Macs and Linux. Unlike Spaceland, there are no ads in Google Earth. What Google's monetization plan is for the consumer portion of this product I don't know (the Google Earth Pro subscriptions can't make up for the consumer version's cost).

The real difference comes in the vision around the actual 3-D product. While Microsoft has taken the data route, Google has taken the decidedly un-Google tactic of relying on people to build out their 3-D world using the Sketchup software. Google has turned building out their world over to the people instead of algorithms, unlike Google News (compared to Digg), Google Maps (compared to Open Street Map) or Google Search. Microsoft has collected data and is using algorithms to build out their world. Which is the better plan? Microsoft will be able to build out their product in a scheduled manner. whereas Google will rely on the wisdom of the crowds to grow it in spurts. Microsoft will go after the largest, most important areas where they get licenses to gather data. Google will have 10 versions of the Eiffel Tower, instead of one. Which consumers want? Microsoft's big bet is that user won't want to pick which model to view; Google's is that consumers will want more control over this world. It will be interesting to look 6 months from now and compare coverage and ease of use.

What's next? As mentioned in the blog post, there will be more cities. Also GM of VE, Stephen Lawler, shared that they wanted to be able to model the street-level view of cities. They would gather this data throught their partner Facet Technology Corporation (presumably through the use of their SightMap product). I also suspect that we will see more features around user annotations. They have collections, but it would be good to see something beyond pushpins.