Google Fusion Tables and FME 2012
It’s been over a year and a bit since I last talked about Fusion Tables, and Google has continued to emphasize it as their solution for hosting user maps and tabular data. So, I’m pretty excited that our recently released FME 2012 can get data into and out of Fusion tables in a ton of interesting ways.
What’s Fusion Tables?
It’s a way to share tabular data, visualize it, and then build on it collaboratively. For example, if you have a spreadsheet with locations inside (addresses, place names, etc.), you can turn it into a map and then share it with a link or embed it in a web page, all in a few minutes. Updates are reflected as they’re made, so the map is never stale.
Similarly, there are simple tools to extend or combine other people’s data (which is tagged with refreshingly minimal metadata – a name, description, attribution, and link); the combined results stay up-to-date too.
What’s changed with Fusion Tables since my last post?
Some things have stayed the same. You can still add and share your own places and shapes in Google Maps, but this feature is now called “My Places” instead of “My Maps”. As mentioned above, Fusion Tables is evolving and getting more traction. For example, tables are now included in Google Docs, there are new styling options, and the everything-can-be-commented-on part of the collaboration model is being rethought.
One interesting development is the introduction of Google Earth Builder, which tackles some of the same hosting-user-data challenges but with an enterprise focus and a much bigger scale. The fact that Earth Builder leverages Fusion Tables internally as a key part of its architecture (discussed earlier) is a good sign that Fusion Tables will be here for some time.
What does FME bring to the table?
FME complements the strengths of Fusion Tables by making it easy to load data from any format, data model, or coordinate system. In my earlier post, I used FME to convert some sample data to KML, which I then imported into Fusion Tables; with FME 2012, this can be done in a single step. Similarly, you can export a table to any of FME’s supported formats, transforming the data as needed along the way.
Updates and deletes are supported too, but these are only suitable for small numbers of rows at a time. This is because Fusion Tables can insert several hundred rows per request, but only update or delete a single row (or delete all rows) in one request.
In addition to making it easy to get your data into and out of Fusion Tables, FME also enables new workflows. For example, Don has an example (15 minute video) showing how to configure FME Server to dynamically update a Fusion Table based on sensor data pushed to it in real time. The possibilities are endless!
Here’s a quick video that I put together which gives a taste of what’s possible.
Want to learn more?
If you’re interested in making maps with Fusion Tables, the Wisconsin State Cartographer’s Office has published a fantastic guide. Interestingly, they also note the desire to import other kinds of data like Shapefiles, and discuss workarounds similar to those in my earlier post (convert to KML or use a web service).
If you’ve used or thought about using Google Fusion Tables, what has your experience been?