FAR-AVIO: Fast and Robust Schur-Complement Based Acoustic-Visual-Inertial Fusion Odometry with Sensor Calibration

TL;DR

FAR-AVIO addresses robust underwater localization by fusing acoustic, visual, and inertial data within a Schur-complement based EKF, enabling constant-time updates through landmark marginalization. It adds online DVL-IMU extrinsic calibration and the AWARE sensor health module to adaptively weight sensors under varying conditions, while maintaining accurate state estimation even during poor visual conditions. The approach shows superior localization accuracy and lower computational requirements than state-of-the-art baselines on real underwater data and simulations, and runs in real time on embedded hardware. The authors release the implementation as open source to support reproducibility and further research in underwater navigation.

Abstract

Underwater environments impose severe challenges to visual-inertial odometry systems, as strong light attenuation, marine snow and turbidity, together with weakly exciting motions, degrade inertial observability and cause frequent tracking failures over long-term operation. While tightly coupled acoustic-visual-inertial fusion, typically implemented through an acoustic Doppler Velocity Log (DVL) integrated with visual-inertial measurements, can provide accurate state estimation, the associated graph-based optimization is often computationally prohibitive for real-time deployment on resource-constrained platforms. Here we present FAR-AVIO, a Schur-Complement based, tightly coupled acoustic-visual-inertial odometry framework tailored for underwater robots. FAR-AVIO embeds a Schur complement formulation into an Extended Kalman Filter(EKF), enabling joint pose-landmark optimization for accuracy while maintaining constant-time updates by efficiently marginalizing landmark states. On top of this backbone, we introduce Adaptive Weight Adjustment and Reliability Evaluation(AWARE), an online sensor health module that continuously assesses the reliability of visual, inertial and DVL measurements and adaptively regulates their sigma weights, and we develop an efficient online calibration scheme that jointly estimates DVL-IMU extrinsics, without dedicated calibration manoeuvres. Numerical simulations and real-world underwater experiments consistently show that FAR-AVIO outperforms state-of-the-art underwater SLAM baselines in both localization accuracy and computational efficiency, enabling robust operation on low-power embedded platforms. Our implementation has been released as open source software at https://far-vido.gitbook.io/far-vido-docs.

Figures (7)

• Figure 1: Real-world deployment of the proposed system in underwater environments. (a) illustrates the experimental setup, where an ROV inspects the target underwater structure. (b), (c), and (d) presents typical visual challenges encountered underwater, including motion blur, longtime textureless regions, and marine snow. (e) shows the dense reconstruction result and estimated trajectory produced by FAR-AVIO.
• Figure 2: Proposed system architecture of FAR-AVIO. Input is from an underwater robot equipped with a stereo camera, an IMU, and a DVL.
• Figure 3: DVL transducer measurements are illustrated in both 2-D and 3-D views. The instrument comprises four transducers oriented in different directions, with Transducer 1 shown as a representative example.
• Figure 4: Different baseline methods estimated trajectory comparisons.
• Figure 5: CPU load and memory usage of FAR-AVIO and baselines on Structure--Easy sequence.