In the realm of underwater robotics and energy exploration, a team of researchers from the University of Technology Sydney, led by Hao Wei, has developed a novel approach to improve the navigation and mapping capabilities of underwater robots. Their work, titled “FAR-AVIO: Fast and Robust Schur-Complement Based Acoustic-Visual-Inertial Fusion Odometry with Sensor Calibration,” addresses the significant challenges posed by underwater environments to visual-inertial odometry systems. This research was published in the IEEE International Conference on Robotics and Automation.
Underwater environments present unique challenges for navigation systems, including strong light attenuation, marine snow, and turbidity, which can degrade the performance of visual and inertial sensors. These issues often lead to frequent tracking failures and reduced accuracy in state estimation. To overcome these challenges, the researchers developed FAR-AVIO, a tightly coupled acoustic-visual-inertial odometry framework specifically designed for underwater robots. This framework integrates measurements from an acoustic Doppler Velocity Log (DVL) with visual and inertial sensors to provide accurate state estimation.
The key innovation in FAR-AVIO is the use of a Schur complement formulation within an Extended Kalman Filter (EKF). This approach enables joint pose-landmark optimization for accuracy while maintaining constant-time updates by efficiently marginalizing landmark states. The researchers also introduced an online sensor health module called Adaptive Weight Adjustment and Reliability Evaluation (AWARE), which continuously assesses the reliability of visual, inertial, and DVL measurements and adaptively regulates their sigma weights. Additionally, they developed an efficient online calibration scheme that jointly estimates DVL-IMU extrinsics without the need for dedicated calibration maneuvers.
The practical applications of this research for the energy sector are significant. Underwater robots equipped with FAR-AVIO can perform more accurate and reliable inspections of offshore oil and gas infrastructure, such as pipelines and subsea installations. This improved navigation capability can enhance the efficiency and safety of underwater maintenance and repair operations. Furthermore, the ability to operate on low-power embedded platforms makes FAR-AVIO particularly suitable for long-term underwater monitoring tasks, which are crucial for environmental impact assessments and ensuring the integrity of offshore energy installations.
Numerical simulations and real-world underwater experiments consistently demonstrated that FAR-AVIO outperforms state-of-the-art underwater SLAM baselines in both localization accuracy and computational efficiency. The researchers have made their implementation available as open-source software, facilitating further development and adoption of this technology within the energy industry. This work represents a significant advancement in the field of underwater robotics and has the potential to enhance the safety, efficiency, and reliability of underwater operations in the energy sector.
This article is based on research available at arXiv.