IARC Mission 7

In summer 2016 I co-founded a team for the International Aerial Robotics Competition as part of Pitt’s Robotics and Automation Society. We worked on Mission 7, which required designing a drone that could navigate indoors while avoiding obstacles and interacting with ground robots. Our team competed at the IARC in summer 2017 and summer 2018, and won awards for “Most Points” and “Best System Design” at the American Venue both years, as well as “Best Technical Paper” at the American Venue in 2018. More information can be found on the project page on the Pitt RAS website, and a technical overview of our design can be found in our 2018 Technical Postmortem.

Our team of around 30 students worked on lots of different technical areas; I primarily worked on:

• Vision-based detection and tracking for target robots with 5 monocular cameras, using both OpenCV and TinyYOLO
• Detection and tracking for obstacles with Intel Realsense cameras
• Vision- and IMU- based state estimation with an EKF
• Simulator for testing perception, controls, and motion planning

I also made some contributions to various components across our controls and planning software. More details on each of these, and the rest of the project, are in the Technical Postmortem.