Building a Peaceful Geodata Marketplace
Is Free Data Always a Good Idea?
ESRI seems to promote the idea that software should cost money, but data should be free. At the same time OSGeo folks seem to think both software and data should be free. Is anyone asking whether free data is always a good idea ?
If you’re like a lot of GIS professionals, a big part of your job is adding value to geodata. Free data seems like a good thing at first, but without incentive for data to be collected, opportunities to add value are limited.
Hyperspectral Cube Image stolen from Duke University Math Dept.
Case in Point: Hyperspectral Imagery
In its raw form, hyperspectral imagery (HSI) is a pain to deal with. The files are huge, and difficult to visualize. To be at all useful, it must first be processed. Processing HSI is not trivial – graduate training is needed plus specialized software. So let’s suppose you take out a few college loans, get an advanced degree in remote sensing, and learn to glean information from HSI that can save the planet.
With HSI supporting so many different types of environmental analyses, you would think the world needs more HSI experts. But once you start looking for a job and get only get 4 hits when searching Dice.com for hyperspectral, all of which are military projects, you may begin to think otherwise. There’s a disconnect somewhere. I think the problem is the market, or rather the lack thereof.
HSI at USGS
USGS sells data collected by the Hyperion satellite. After using the rather lame EarthExplorer, you must purchase an image in its entirety ($250 each). I’m not sure how big the files are, but rest assured, they are big … purchase is delivered on DVD, not available via download. Just because you can buy something at there website doesn’t mean its a marketplace though. A market is a place where you can sell stuff, like eBay.
Take a look at WeoGeo where buyers are able to discover and purchase an HSI with just the bands they need for an area of interest. Perhaps as importantly though, users can also participate as sellers. So after processing the imagery, a user can then log in as a seller and list a derived product for sale, choosing how royalties are to be paid back to the vendor of the parent (raw) data.
I’ve used hyperspectral imagery as an example, but I think the same forces apply to any dataset requiring expertise to process. Sensors are becoming more sophisticated and UAVs more affordable. Technology alone does not an industry make. We need to be building marketplaces for the new types of data these new platforms will be collecting.