A single image, from about 500 feet high
Along with 462 other people, I backed a Kickstarter project earlier this year. The open hardware project sold balloon mapping kits to tinkerers like myself, and cost $85 for a package that included everything but the camera and the helium. I finally got an opportunity to give the kit a chance, and this article is going to run through my experience. We’ll cover the gear I bought, but the more important subject is that of hobbyist balloon mapping (a very important subject indeed).
Balloon mapping is a pretty simple concept. You fly a weather-sized balloon (about 5-feet in diameter) in the air and dangle a camera from it. If done properly you’ll get any number of overheard shots of a small parcel of land, which you can stitch together into a map without too much difficulty. It’s best thought of as a poor man’s aerial or satellite imagery, as you get the same effect, with a total investment of under $300. And due to the proximity to the ground, you can get great results with very inexpensive equipment. It’s not so much a replacement for satellite/aerial imagery as it is a low-cost, high-resolution alternative.
Upon receiving the balloon mapping kit, and resisting the urge to assemble it in my apartment, I realized that I’d need something to inflate the balloon with. In this case it’s helium, which you can get at any party store, but it won’t be as simple as you expect. I knew my balloon required 80 cubic feet of helium, but the worker at the store measured their helium tanks in terms of how many (party) balloons they could fill. In my case a small tank, about 2 feet tall, could handle 60 balloons or a more reasonably-sized tank, at about 4 feet tall, was good for 120 balloons. Opting for too much as opposed to too little I got the larger one, which cost me $50 plus a $200 deposit. The smaller size, which I now think would have been fine for a single inflation, would have been $30.
The other component mappers will need to think through is the camera. In my case I have an old Canon point-and-shoot setup with the CHDK firmware. This means I have an intervalometer on it, which I can set to shoot however I’d like (such as one shot every 3 seconds with a 5 minute delay on the first shot). This is the high-tech solution, the alternative would be any camera with continuous shooting and a rubber band holding down the trigger. This doesn’t sound like a great idea to me, but I suppose it could work.
With the hardware out of the way, it was time for the setup. There is some knot tying and light construction, but essentially you’re filling a 5.5-foot diameter balloon, hanging a camera from it, attaching a rope, and then letting it go in the air (while holding the rope). My kit included a 1000-foot spool of high-quality kite string, which did the job nicely and was totally tangle-free so long as I was careful. The documentation talked about a rope length of 1000 meters so I was disappointed with the amount of elevation I got, but I will remedy this before I balloon map again.
Once the rig is up in the air, most of your work is done. At this point it was time for some light mathematics (basically figuring out how long the memory card will last) and then directing the balloon. The balloon is completely subject to the wind so you can only direct it as far as you can walk in in a given direction. The more important part of is keeping the string away from trees and pesky power lines.
While flying, there were two things that surprised me, though both should have been completely obvious. The first was just how windy it is above the tree line. I was mapping on what seemed like perfectly still day (winds of 1-2 mph) but the whole time the balloon was being pushed to the northeast by the wind. What was nonexistent where I was standing (the tops of the trees weren’t even moving) was enough to have me mapping about 500 feet away from my intended target. There was not one point in the day where the balloon was close to directly overhead, and again, this was a very still day.
The other factor that surprised me is the weight of the rope. Even very light Dacron line weighs over a pound when you have 1000 feet of it, so when using a balloon with under 10 pounds of lift the length of the lines has a major impact. There was a noticeable deflection in the line between the endpoints, all of which was caused by the weight of the line. So it was possible for me to pull in line and have the balloon actually go higher. While I’m very happy with the Dacron, I’m considering something lighter, possibly a braided fishing line (though I have yet to explore this sufficiently).
The main downside with the kit I bought was its method for supporting the camera. It proposed the use of the top half of a 2L soda bottle, with the camera tethered inside. The rig is easy to set up, cheap, and light, but also flimsy. During my first flight I only got about 50 out of 500 shots because my camera shifted after being hit with a gust of wind. After this I redesigned the camera rig, using some of the parts I had lying around. I ended up using an insulated faucet cover, which is much more secure, protected the camera better, and is more stable in the wind. At $1.50 it’s barely more expensive than a bottle of soda.
Once you have all your pictures taken, your quest has only just begun. At this point you need to stitch them together, a task which you’ll pretty much have to do manually. MapKnitter.org has a very nice tool for this, but it won’t do the work for you, it will just give you the tools to lay out and contort your image (remember, the camera moved side-to-side a lot, so not all your images will have a perfectly flat perspective). Ultimately if you took enough pictures, and you edited through them diligently, you’ll end up with a map that you can save, print, or expand upon.
With it all said and done, I can say that I’m quite happy with my balloon mapping experience. I’ll be doing it again soon, with a few changes to my routine. Next time I hope to have more line (1000 meters maybe?), be more methodical about my mapping coverage, get the smaller tank of helium, start earlier in the day in order to avoid winds (6am or so), and I’ll attach a GPS tracker to the rig. The GPS won’t make a huge difference as the areas are easy to approximate, but I’ll appreciate having the extra data to work with.
The result of my mapping were mixed, but mostly I was impressed with what I got. The shots were of a location that will remain undisclosed (it might or might not be a family member’s home), but next time I’ll be sure to make a map available. For now you can view a single image I took (at the top) or a sample from MapKnitter directly below.
Sample map: Baseball field by anonymous
The balloon mapping kit is still available for purchase through the organization behind it all, the Public Laboratory, if you’re interesting in giving it a shot.
- Baseball field by anonymous
[kit image via public laboratory]
The other thing to consider is that Google might not be planning to release devices at all. If you’ve been paying attention to the speakers at this year’s AnDevCon, you’d see that there are no shortage of Googlers with beta builds of Jelly Bean. Several developers in attendance at that event have made note of it and it’s not surprising. Google tests their OS in every stage of the build process in very much the same way. In fact, Google has private developer space in Mountain View for OEMs and other partners that would like to come an work with them on the next version of Android.
