Third major iteration. It is about 1/3 the weight of my first attempt.
This is still a work in progress. The motion-based patterns using the gyroscope and accelerometer aren't fine tuned for this new version of the hardware so I didn't want to show the motion patterns, only motion colors.
The microcontroller bundle is 21 grams, the light strip is 71. The mid-vent Djinn started at 237 grams, and was 341 grams when fully assembled, so the tape adds about 12.
The battery life depends on how bright the lights are, since they can each of the 300 lights draw anywhere from 0-60mA ranging from off to full bright. The 350 mAh battery lasted around 45 minutes, better than I expected and longer than some of my tests at home.
The wind was variable and turbulent, the Djinn performed wonderfully. I chose the mid-vent because the forecast called for 8-10 with gusts to 15, with a low pressure system moving in. Winds were good when we set up, but then dropped to almost nothing shortly after setting up. I choose short lines and feared I'd be doing more training with light wind practice so I used 30 foot lines. With almost 50% of the weight added extra, and such variable wind, it flew wonderfully.
I can only imagine how a group fly with the directional-based colors would look. As each kite followed along the line their colors would match each other as they flew through each position. I've programmed that color palette to be based both orientation and compass direction, so it would vary as they flew but be the same at each point in space.