Performance Assessment of an Adaptive AUV Survey

SHORT ABSTRACT: Traditional underwater surveys employing side-looking sonars are designed by assuming quality data will be collected out to a fixed range. However, environmental factors at sea introduce variability into the coverage actually achieved. To address this issue, the NATO Undersea Research Centre (NURC) has developed an adaptive survey algorithm for an autonomous underwater vehicle (AUV) equipped with a synthetic aperture sonar (SAS). This algorithm allows the AUV to react to in situ data and adjust its route accordingly to ensure complete coverage of the area of interest is achieved. This adaptive strategy, which assumes no a priori knowledge, is compared to pre-planned approaches in an extensive performance assessment. The study demonstrates the power of the adaptive strategy, which in certain scenarios can match the optimal pre-planned survey that assumes perfect knowledge of sonar performance. Results show that the adaptive survey algorithm outperforms traditional surveys designed with a fixed sonar range, by guaranteeing complete cov-erage while limiting the travel distance (or time) required to do so. The assessment is supported by real data collected at sea