SBIR/STTR Award attributes
The U.S. Navy uses Unmanned Underwater Vehicles (UUVs) with different configurations and capabilities for critical missions. To ensure successful mission completion, these UUVs must always know their precise geo-location. UUVs typically have inertial measurement units and a doppler velocity log (DVL) onboard for measuring their speed with respect to the ground, and using these data, position updates are calculated via dead reckoning algorithms. Because these dead reckoning calculations accumulate position error over time, UUVs reduce this error by surfacing for GPS fixes or by using acoustic transponders (beacons) with known positions. However, transponders and GPS fixing methods are not practical in contested and GPS-denied environments. Although available terrain-aided navigation (TAN) and landmark-based simultaneous localization and mapping (SLAM) techniques can perform localization, they are not suitable for unstructured environments, because TAN depends on high-resolution bathymetric maps, and current SLAM methods only work in highly structured environments. To address UUV position fixing in unstructured environments, we are proposing Underwater SLAM (U-SLAM), an innovative SLAM algorithm that combines Charles River’s expertise in underwater sonar image processing and SLAM-based navigation solutions for a variety of unmanned vehicles. In Phase I, we plan to demonstrate U-SLAM through simulations.
