100 Hackers Walk into a Planetarium…
Science Hackday Chicago blew through the Adler like a hurricane this past weekend.
Over 100 hackers, electronics gurus, artists, scientists and educators came to see what can be accomplished in 30 hours of dedicated tech crunching. I suggested two projects: a controlled descent system for our missions using a parafoil kite and a user-friendly prediction/tracking software interface. I got a lot of takers on the parafoil kite idea.
A couple of folks were interested in the tracking software. In total 6 people joined the projects. After hashing out peoples’ skills and interests we focused on what was “doable”. We had mixed success but the most important outcome may be the valuable lessons we learned that can apply to a working design in the future. Just getting the payload in the air was quite an accomplishment but we were able to test real hardware and software options as well.
Two designers, Wei and Noah started working on a visual interface for the tracking program. Wei left in the evening and came back in the morning (other than her everyone else worked all through the night with only a 1-2 hour break for sleep!). We didn’t get any takers on the programming end of the software but Noah designed a really nice console for selecting a range of launch variables that can be used to run multiple predictive tracks while displaying the range of flight options with overlays color coded for ascent speed or max altitude.
There were two aspects to the controlled descent idea: Using the internal GPS to pick a safe target landing site and controlling the flight. On the controlled descent team, Brenda, Diego and Tim worked on using internal GPS data but realized the more important (and realistically testable) part was parafoil control. They used an accelerometer to run a servo to pull the kite cords – ostensibly to keep the flight level. They figured once we can show some ability to control the kite from the payload, it would be a small step to control direction. With that accomplished, adding the GPS for targeting would be a relatively easy final step.
They were able to get the GPS to inform the Arduino of the payload’s location while Tom and I built the mechanical system. We machined an aluminum axle that attached to the servo with two 5″ diameter v-grooved wheels on it. The wheels wound and unwound the kite string with a 6″ max take-up (6″ is what Tom tested to be sufficient to get a significant turn out of the kite).
We went through many tests and made 2 or three major redesigns by the morning. We did get intelligent results from the accelerometer/take-up wheel system. When the accelerometer was tilted, the wheel took up the control lines as intended. We discovered quite too late that the servo we were using to drive the take-up wheels did not have enough power to pull the 6 pound payload.
In a last ditch attempt to see some working results we decided to cut down the wheels and try and use them as a pendulum. Though the concept may be valid, we determined a step motor would probably work better than a servo in driving the mechanism.
Here’s a shot from ~5am of our first test getting the payload in the air. Hmm, reminds me of something. The first valuable lesson learned was that the payload needs to be fixed much closer to the parachute because as it pendulums it leads to a “feedback loop” from which it is hard to recover. Thus the first time we needed to rebuild the payload box!
After a mere 26 straight hours I had to bow out to attend to Mother’s Day duties and the team continued on refining for the final hours. They were able to run a final flight test. You can see their results here on YouTube.
The best part of the whole experience was meeting and working with so many enthusiastic and talented people. This is just the beginning of a relationship with all those involved. I’ve extended an invitation to everyone on the team to join us for an upcoming launch. They were all very interested in that. So, hopefully we’ll see more of them in the coming months.