A 120 camera rig of networked Canon DSLRs driven by RaspberryPi with a custom GPIO backpack.
PEDG approached us with the monster task of controlling 120+ DSLR cameras in real-time, as part of the Nike Speed Track activation to be launched in NYC for 2 weeks during the World Cup. Because we not only needed to trigger captures with extreme precision ( < 5ms latency), but also download the images locally in order to generate the composited video, we devised a networked solution with Raspberry Pi backpacks on each camera.
I designed a PCB to interface with the RPi's GPIO pinouts which triggered and autofocused the DSLR via it's shutter port, read an analog input from 3rd party laser/motion sensors, and provided LED status feedback. The RPis run a Node app listening for networked UDP commands, and transferring image data after it was downloaded from it's camera.
Immediately after a set of images are captured, they are sent to an automated image processing workflow, overlaid with the event branding, and then sent to the user via email.
ROLE
Lead Hardware, Software,
Project Management
TECH
Raspberry Pi, NodeJS
TEAM
Incredible Machines, CFS, PEDG, Nike
YEAR
2014
