At CSULA, the RoboSub club is developing the Lanturn autonomous submarine for the 18th annual RoboSub competition, in collaboration with senior design teams from Mechanical, Electrical, and Computer Science (MEEECS). The project is organized into three sub-teams—Controls, Autonomy, and Computer Vision—each tackling key areas for competition readiness.

The Controls team is researching MicroROS communication protocols, integrating microcontrollers like the RP2040 using the Arduino framework within PlatformIO, and implementing PID control loops for precise maneuvering. The Autonomy team is reviewing the updated 2024 codebase on GitHub, completing tutorials for the Groot2 behavior tree builder, and exploring MicroROS integration with the Jetson Orin Nano for onboard decision-making. The Computer Vision team is working with ROS2 Iron, training models using RoboFlow and Google Colab, and deciding whether to upgrade from YOLOv7 to YOLOv8 for better object detection accuracy.

Additionally, the team is addressing challenges in localization and mapping due to limited camera field of view, aiming to improve spatial awareness and navigation in the pool environment during competition.