Azavea is a rare company. One of the benefits that we (full-time employees) earn is the ability to define our own research project after 6 months.  There is a list of active research projects here.  My personal interests took me to working with C. Dana Tomlin, and implementing a radial propagation tool for ArcToolbox, as well as collaborating with the MIT Media Lab’s Tangible Media Group on the project http://www.sourcemap.org/.

I won’t get into what sourcemap is (that’s already been done), but I thought it would be cool to mention some of the technical challenges that the project was/is facing, and what we’re doing about it.

First, the project started out with maps from a free, proprietary mapping API. This was a great stepping off point, but as the maps and networks became more complex, that API started to break.  I advocated that the project migrate to an Open Source JavaScript mapping library, by the name of OpenLayers.  This helped us overcome that barrier to complex maps.

Next, we wanted to change the cartography from simple lines between points to something more ’stylish’.  The problem that we experienced was that it was difficult to discriminate the paths of colinear parts of an object. To solve this, we added a slight arc to the lines.  This technique makes it easier to discriminate the connections between colinear parts and an object.

Time to bust out the trigonometry again!  I doodled the following sketches on a mousepad weekly planner.  I wanted an offset of the line in the vertical direction for most horizontal lines.  This calculation got me started, but then I realized that it would be weird if the arcs suddenly swapped sides if you passed the Y axis (e.g. from 359 degrees to 1 degree).

So I added another step, which attenuated the arcing, starting at 45 degrees from either side of the Y axis, and gradually decreasing to no deflection in the Y direction.

The result?  A cool looking network of lines that are clearly distinguishable! You can see it in action here.

We’ve got even more improvements on the way for the map, and I think it will make it a really compelling, and intuitive visualization of global supply chains.  Watch this space for the next installment.