Autonomous Underwater Vehicle (AUV)

This autonomous underwater vehicle localization project centers on a custom landmark-based FastSLAM implementation developed in C++ and integrated with ROS, Gazebo, and RViz for real-time sensing and simulation capabilities. The system underwent comprehensive algorithm evaluation through analysis of RatSLAM, BioSLAM, and GraphSLAM approaches, examining their biologically inspired mechanisms, graph-based optimization techniques, and topological mapping strategies to determine optimal solutions for robust underwater navigation and mapping in challenging low-visibility, sensor-noisy environments. The SLAM navigation algorithms were seamlessly integrated with the robot’s software stack and rigorously tested using the multi-sensor Caves dataset to validate performance in real-world underwater scenarios. Additionally, the system employs an Extended Kalman Filter (EKF) for sophisticated multi-sensor fusion, combining camera feeds, Doppler Velocity Log (DVL), and Inertial Measurement Unit (IMU) data to achieve precise underwater navigation and localization.
Repository