WhereCamp unconference on mapping, geolocation and related topics.
This was my second year at WhereCamp. It's always a bit humbling. I feel like I'm pretty geeky, and I've written a couple of Python mapping apps and I know spherical geometry and stuff ... but when I get in a room with the folks at WhereCamp I realize I don't know anything at all. And it's all so interesting I want to learn all of it! It's a terrific and energetic unconference. I
I won't try to write up a full report, but here are some highlights.
Several Grassroots Mapping people were there again this year. Jeffrey Warren led people in constructing balloons from tape and mylar space blankets in the morning, and they shot some aerial photos. Then in a late-afternoon session he discussed how to stitch the aerial photos together using Cargen Knitter.
But he also had other projects to discuss: the Passenger Pigeon project to give cameras to people who will be flying over environmental that need to be monitored -- like New York's Gowanus Canal superfund site, next to La Guardia airport. And the new Public Laboratory for Open Technology and Science has a new project making vegetation maps by taking aerial photos with two cameras simultaneously, one normal, one modified for infra-red photography.
How do you make an IR camera? First you have to remove the IR-blocking filter that all digital cameras come with (CCD sensors are very sensitive to IR light). Then you need to add a filter that blocks out most of the visible light. How? Well, it turns out that exposed photographic film (remember film?) makes a good IR-only filter. So you go to a camera store, buy a roll of film, rip it out of the reel while ignoring the screams of the people in the store, then hand it back to them and ask to have it developed. Cheap and easy.
Even cooler, you can use a similar technique to make a spectrometer from a camera, a cardboard box and a broken CD. Jeffrey showed spectra for several common objects, including bacon (actually pancetta, it turns out).
|JW:||See the dip in the UV? Pork fat is very absorbent in the UV. That's why some people use pork products as sunscreen.|
|Audience member:||Who are these people?|
|JW:||Well, I read about them on the internet.|
Two Google representatives gave an interesting demo of some of the new Google APIs related to maps and data visualization, in particular Fusion Tables. Motion charts sounded especially interesting but they didn't have a demo handy; there may be one appearing soon in the Fusion Charts gallery. They also showed the new enterprise-oriented Google Earth Builder, and custom street views for Google Maps.
There were a lot of informal discussion sessions, people brainstorming and sharing ideas. Some of the most interesting ones I went to included
- Indoor orientation with Android sensors: how do you map an unfamiliar room using just the sensors in a typical phone? How do you orient yourself in a room for which you do have a map?
- A discussion of open data and the impending shutdown of data.gov. How do we ensure open data stays open?
- Using the Twitter API to analyze linguistic changes -- who initiates new terms and how do they spread? -- or to search for location-based needs (any good ice cream places near where I am?)
- Techniques of data visualization -- how can we move beyond basic heat maps and use interactivity to show more information? What works and what doesn't?
- An ongoing hack session in the scheduling room included folks working on projects like a system to get information from pilots to firefighters in real time. It was also a great place to get help on any map-related programming issues one might have.
- Random amusing factoid that I still need to look up (aside from the pork and UV one): Automobile tires have RFID in them?
Lightning talks included demonstrations and discussions of global Twitter activity as the Japanese quake and tsunami news unfolded, the new CD from OSGeo, the upcoming PII conference -- that's privacy identity innovation -- in Santa Clara.
There were quite a few outdoor game sessions Friday. I didn't take part myself since they all relied on having an iPhone or Android phone: my Archos 5 isn't reliable enough at picking up distant wi-fi signals to work as an always-connected device, and the Stanford wi-fi net was very flaky even with my laptop, with lots of dropped connections.
Even the OpenStreetMap mapping party was set up to require smartphones, in contrast with past mapping parties that used Garmin GPS units. Maybe this is ultimately a good thing: every mapping party I've been to fizzled out after everyone got back and tried to upload their data and discovered that nobody had GPSBabel installed, nor the drivers for reading data off a Garmin. I suspect most mapping party data ended up getting tossed out. If everybody's uploading their data in realtime with smartphones, you avoid all that and get a lot more data. But it does limit your contributors a bit.
There were a couple of lowlights. Parking was very tight, and somewhat expensive on Friday, and there wasn't any info on the site except a cheerfully misleading "There's plenty of parking!" And the lunch schedule on Saturday as a bit of a mess -- no one was sure when the lunch break was (it wasn't on the schedule), so afternoon schedule had to be re-done a couple times while everybody worked it out. Still, those are pretty trivial complaints -- sheesh, it's a free, volunteer conference! and they even provided free meals, and t-shirts too!
Really, WhereCamp is an astoundingly fun gathering. I always leave full of inspiration and ideas, and appreciation for the amazing people and projects presented there. A big thanks to the organizers and sponsors. I can't wait 'til next year -- and I hope I'll have something worth presenting then!
[ 23:40 Apr 24, 2011 More mapping | permalink to this entry | ]