Development of Autonomous Strategies for High Resolution Data Acquisition For sUAS-Based Atmospheric Research

This study presents the development and implementation of an autonomous atmospheric sensing platform using a Hexsoon EDU-450 quadcopter, a CubePilot Cube Orange running ArduPilot, a Raspberry Pi companion computer, and an Alphasense OPC-N3 particulate sensor. The goal of this work was to create a fully programmable and synchronized flight pipeline capable of executing structured autonomous missions while collecting high-resolution particulate measurements in real time. Hardware integration included custom sensor mounting, accounting for Electromagnetic interference, and a dedicated UART link between the Raspberry Pi and Cube Orange for MAVLink communication. ArduPilot was selected over PX4 due to its stability during low-speed and hover-dominant operations, long with its stronger companion-computer interface. DroneKit was chosen as the primary autonomy library after demonstrating more reliable timing, parameter access, and mission execution than MAVSDK. Four autonomous flight patterns were evaluated: a boustrophedon and zigzag path for horizontal coverage, fixed-position hovers for sensor stability analysis, anda box trajectory for navigation precision. To synchronize autonomous navigation and sensing, the OPC-N3 sampling routine was executed as a Python subprocess embedded within the main DroneKit script. This allowed both processes to run concurrently while using MAVLink timestamps to fuse particle data with GPS, altitude, velocity, and attitude information. Autonomous data collection and path planning were successfully achieved. Results indicate that the EDU-450 effectively served as an autonomous atmospheric research testbed, successfully executing all mission types despite its small frame