BIND-USBL: Bounding IMU Navigation Drift using USBL in Heterogeneous ASV-AUV Teams

A team from the Indian Institute of Science, led by Pranav Kedia, has published a breakthrough paper titled 'BIND-USBL: Bounding IMU Navigation Drift using USBL in Heterogeneous ASV-AUV Teams.' The research tackles a fundamental problem in marine robotics: the accurate, continuous localization of Autonomous Underwater Vehicles (AUVs) in GPS-denied environments like the deep ocean. Currently, AUVs rely on inertial dead-reckoning, which accumulates unbounded positional drift—often exceeding 10% of the distance traveled—due to sensor noise and bias. Without external corrections, this makes long-duration missions like seabed mapping or pipeline inspection unreliable.

The BIND-USBL framework proposes a cooperative solution using a heterogeneous team of robots. It deploys Autonomous Surface Vessels (ASVs) as mobile acoustic anchors, each equipped with an Ultra-Short Baseline (USBL) positioning system. These ASVs provide intermittent acoustic fixes to bound the AUV's growing navigation error. The key innovation is not just using USBL, but optimizing the entire system. The framework includes a model linking survey scale and ASV placement to acoustic coverage, and a novel TDMA (Time-Division Multiple Access) uplink scheduler based on conflict graphs. This scheduler allows multiple AUVs to share the acoustic channel efficiently, improving fix delivery rates without signal collisions.

Evaluated in the HoloOcean simulator, the system demonstrated that localization performance is a complex interplay of survey scale, acoustic coverage, team composition, and ASV formation geometry. The spatial-reuse scheduler successfully maintained low end-to-end fix latency while maximizing updates. This approach shifts the focus from the precision of individual USBL measurements to ensuring their reliable geometric and temporal availability, creating a robust navigation backbone for underwater fleets.