[Project] Autonomous bio-inspired escape with the Crazyflie
Posted: Sun Jul 13, 2014 1:41 am
Hi everyone,
I wanted to share a recent robotics project that I did based on the Crazyflie. From a high level, I explored mimicking how fruit flies detect and respond to moving threats like swatters for lightweight and onboard threat evasion in a quadcopter, with the ultimate goal of making flight (for, e.g., search and rescue missions) truly autonomous and scalable. Using the Crazyflie as a development platform, I created a computationally and physically lightweight sensor module inspired by the fruit fly's rudimentary visual system. I also modified the Crazyflie's firmware to implement onboard algorithms to escape from approaching threats by mimicking fruit fly escape behaviors.
You can read about my project in more detail at http://goo.gl/zE9P9R. Specifically, I discuss the hardware and software modifications that I made to the Crazyflie (for the sensor module and escape algorithms, respectively) in Methods, and present a slow-motion video and graphs of telemetry data from an escape attempt, along with a still-image comparison to fruit fly escape, in Results.
As a result of this project, I was selected as a Regional Finalist and a Computer Science Award nominee in the 2014 Google Science Fair. This is extremely exciting because there are only 90 Regional Finalists and 15 Computer Science Award nominees picked from around the world.
I would be happy to answer any questions about the project. I really enjoyed working with the Crazyflie (the hardware expansion header/open source firmware/etc. were amazingly helpful), so thanks to Bitcraze and this community for an excellent development platform!
I wanted to share a recent robotics project that I did based on the Crazyflie. From a high level, I explored mimicking how fruit flies detect and respond to moving threats like swatters for lightweight and onboard threat evasion in a quadcopter, with the ultimate goal of making flight (for, e.g., search and rescue missions) truly autonomous and scalable. Using the Crazyflie as a development platform, I created a computationally and physically lightweight sensor module inspired by the fruit fly's rudimentary visual system. I also modified the Crazyflie's firmware to implement onboard algorithms to escape from approaching threats by mimicking fruit fly escape behaviors.
You can read about my project in more detail at http://goo.gl/zE9P9R. Specifically, I discuss the hardware and software modifications that I made to the Crazyflie (for the sensor module and escape algorithms, respectively) in Methods, and present a slow-motion video and graphs of telemetry data from an escape attempt, along with a still-image comparison to fruit fly escape, in Results.
As a result of this project, I was selected as a Regional Finalist and a Computer Science Award nominee in the 2014 Google Science Fair. This is extremely exciting because there are only 90 Regional Finalists and 15 Computer Science Award nominees picked from around the world.
I would be happy to answer any questions about the project. I really enjoyed working with the Crazyflie (the hardware expansion header/open source firmware/etc. were amazingly helpful), so thanks to Bitcraze and this community for an excellent development platform!