Sensor drivers, state estimation, trajectory generation/publishing, Text User Interface, and acoustic image processing pipeline for navigational purposes utilizing ROS2 Galactic. Master thesis is publicly available here.

Generated Trajectory	Executed Trajectory

Consult separate README's for the usage of respective packages:

• brov2_dvl: Driver for A50 DVL
• brov2_sonar: Driver for OSM Ethernet Sonar System
• brov2_barometer: Driver for Bar30 Pressure Sensor
• brov2_gps: Driver for Ultimate GPS
• brov2_trajectory: Trajectory Generator and Trajectory Publisher
• brov2_qekf: State Estimation using Quaternion based Error-State Kalman Filter
• brov2_sonar_processing: Image Construction Pipeline for Side-Scan Sonar Data
• brov2_tui: WIP

The packages/scripts rely on the following list of libraries and modules to function properly.

Onboard:

• Nlohman Json - JSON for Modern C++
• BlueRobotics MS5837 Python Library
• System Management Bus (SMBus)
• Adafruit_CircuitPython_GPS module
• BNO055 IMU sensor ROS2 library

Locally:

• Ncurses
• Sensor Fusion and Tracking Toolbox

Both:

• GStreamer
• NumPy
• SciPy
• Pandas
• progressbar2

In order to set up 30fps live stream from camera to PC; run the shell scripts start_video_pc.sh and start_video_pi.sh on PC and PI respectively.

The following hardware was used in this project:

• ROV: Blue Robotics BlueROV2 (heavy configuration)
• Embedded Computer: Raspberry Pi 4 Model B
• DVL: Waterlinked A50
• IMU: Adafruit BNO055
• Barometer: Blue Robotics BAR30
• Side-scan sonar: Deepvision OSMEthernet Sonar System, 680 kHz transducers

MIT Apache 2.0