Interrelated functional topics concerning autonomy related issues in the context of autonomous inspection of underwater structures

The architecture of an inspection AUV is much more complex than the one of a survey type AUV since it requires high maneuverability and hovering capacity. It needs a specific Autonomy Management System (AIMS) and navigation software as well as dedicated sensors to allow the vehicle to navigate relatively to the structure to be inspected instead of the type of "absolute" navigation, usually used for survey AUVs and which purpose is only to a given accurate localization of the vehicle. The interrelation between different autonomy related issues due to this complexity will be demonstrated with the help of the findings during development and testing of ALISTAR 3000, Autonomous Underwater Vehicle is designed to answer the complex requirements cited above in enabling the inspection of underwater structures such as pipes, wellheads, mooring lines through embarked intelligence. The paper will focus on the functional description of the AMS in displaying different dual concepts that help to structure general functions linked to autonomy: supervisor and vehicle, perception and decision, symbolical and numerical descriptions, mission and environment description, external and internal worlds, etc. This dualist approach leads to an AMS mainly structured by one main supervisor, two databases and four autonomy functions. The paper will also present the results of a series of sea trials, which have been carried out end of 2004 and first half of 2005.