Worked on detecting objects using deep learning for UAV systems

Researching cooperative capabilities for multi-agent systems.

Designed a decentralized cooperative resource allocation algorithm for deconflicting airspace thatwas successfully field tested in a 15-UAV mission. The purpose of which was to decrease human involvement during the launch of multiple UAVs and maximize time available for cooperative flight.Rapidly prototyped a GUI during live-fly field experimentation in order to increase situation awareness for safety critical missions.Served in the role of Swarm Operator for record breaking 30-UAV live-fly field experimentation.

Decreased network latency between a ground control station to UAVs over a 900Mhz link in order to provide reliable control in a RF saturated environment.Designed and implemented a Reynolds Flocking algorithm to explore collective behaviours in swarms of flying wing UAVs. In addition developed a yaw rate control loop on for the UAVs autopilot in order to increase performance by providing more aggressive lateral movement of aircraft.

Explored and implemented path planning algorithms on ground robotic vehicles. Most notable of which was the Rapidly Exploring Random Tree algorithm that allowed the nonholonomic vehicle combined with LIDAR to navigate through an obstacle rich environment.Experimented with collective behaviors for small groups of quadcopters. This was done by combining the quadcopters with VICON motion capture system for accurate positioning and ROS in order to control movement. Showed the abilitiy to create stable clusters of 3-4 UAVs using artificial potential fields.