Journal article Open Access

A reconfigurable framework to turn a MAV into an effective tool for vessel inspection

Francisco Bonnin-Pascual; Alberto Ortiz; Emilio Garcia-Fidalgo; Joan P. Company-Corcoles

Vessels constitute one of the most cost effective ways of transporting goods around the world. Despite the efforts, maritime accidents still occur, with catastrophic consequences. For this reason, vessels are submitted to periodical inspections for the early detection of cracks and corrosion. These inspections are nowadays carried out at a great cost. In order to contribute to make ship inspections safer and more cost-efficiently, this paper presents a novel framework to turn a Micro-Aerial Vehicle (MAV) into a flying camera that can virtually teleport the human surveyor through the different structures of the vessel hull. The system architecture has been developed to be reconfigurable so that it can fit different sensor suites able to supply a proper state estimation, being at the same time compatible with the payload capacity of the aerial platform and the operational conditions. The control software has been designed following the Supervised Autonomy paradigm, so that it is in charge of safety related issues such as collision avoidance, While the surveyor, within the main control loop, is supposed to supply motion commands while he/she is concentrated on the inspection at hand. In this paper, we report on an extensive evaluation of the platform capabilities and usability, both under laboratory conditions and on board a real vessel, during a field inspection campaign.

This is a preprint version of publication with DOI: https://doi.org/10.1016/j.rcim.2018.09.009. This work is partially supported by FEDER funding, by the European Social Fund through Ph.D. scholarship FPI10-43175042V (Conselleria d'Educacio, Cultura i Universitats, Govern de les Illes Balears), by project nr. AAEE50/2015 (Direccio General d'Innovacio i Recerca, Govern de les Illes Balears).
Files (22.1 MB)
Name Size
FBP_RCIM2019_preprint.pdf
md5:183a78343d30fda84723f356cdcbb6dc
22.1 MB Download
18
36
views
downloads
All versions This version
Views 1818
Downloads 3636
Data volume 795.6 MB795.6 MB
Unique views 1818
Unique downloads 3333

Share

Cite as