Towards fully autonomous underwater vehicles in ASW scenarios: an adaptive data driven AUV mission management layer
Abstract
We present an adaptive, data driven Mission Management Layer (MML) to manage the phases (surveillance, track prosecution and target reacquisition) of the mission of AUVs operating as receiver nodes in a multistatic network for littoral surveillance. Among the tracks produced by the on board tracker, the MML selects those which are likely originated by a target. The MML then issues commands to the vehicle Control Layer to trigger data-driven optimization behaviours to prosecute these candidate tracks. Selecting few candidate tracks is indeed crucial since the addressed scenario usually presents many tracks hence resulting challenging from the tracking point of view. To select the candidate tracks, a metric is introduced to quantify their quality. It is based on the kinematic features of the target, on an acoustic model and on the quality of measurement-to-track associations. By selecting a limited subset of produced tracks, the MML can improve the benefits of using data-driven behaviours activating them only on interesting cues and reducing the time wasted to optimize tracks not related to a target. This aims at achieving a balance between the surveillance of the area of interest and exploitation of target cues (tracks). These features are essential for improving the overall mission performance and for an effective AUV decision making in realistic conditions. We present results from sea trials demonstrating the effectiveness of our approach and how the proposed MML can represent a step forward towards the full autonomy of our system. These results represent one of the first examples of AUVs autonomously taking decisions in a realistic and complex littoral surveillance scenario.
Report Number
CMRE-PR-2019-117Source
In: Proceedings of the OCEANS 2015 MTS/IEEE Conference, 18-21 May 2015, Genoa, Italy, doi: 10.1109/OCEANS-Genova.2015.7271515Date
2019/06Author(s)
Ferri, Gabriele
; Munafò, Andrea
; Goldhahn, Ryan A.
; LePage, Kevin D.