Chase Lee was born and raised in Honolulu, HI, currently attending the University of Hawai’i at Manoa. Growing up, he particularly enjoyed technology and video games so he is now working towards a Bachelor’s degree in Computer Science. In his spare time, he likes to take part in many clubs such as the International Student Association, Game Dev at UHM, and UH Esports. One day Chase hopes to create his own indie game or be a part of a well-respected company, but for now, he is enjoying his educational journey in computer science.
Home Island: O’ahu
High School: Wallace Rider Farrington High School, HI
Institution when accepted: University of Hawaii at Manoa
Akamai Project: Simulating Drone Flight, Camera Aperture, and Environment for Machine Learning Using Unity
Project Site: KBR – Kihei, Maui
Mentor: Jeremy Case
Drone detecting devices are inadequate to where they lose track of drones that go below the horizon. To improve drone detection, images of flying drones need to be supplied. To do this, Unity, a platform for creating real-time 3D content, can be used. Unity can simulate physics and high-definition graphics of drones that can be utilized to improve machine learning (ML) models. Drones can be observed under realistic standards by simulating cameras using Unity. These simulations will fulfill the niche of drone image data needed to improve drone ML models. By using Unity, Proportional Integral Derivative (PID) controllers were added to virtual drone models which imitated realistic flight. These drones were put under different environmental conditions such as day or night. Cameras were also simulated to further improve this technology so that both ideal conditions and non-ideal conditions can be recorded. The simulation output shows drones affected by wind, and under different appearances, whether it be under different lighting or different colors and textures. A large quantity of drone imagery will be produced and will go on to improve ML technology so that the program can not only detect drones beneath the horizon but also detect drones under different circumstances.