Vision based autonomous docking for work class ROVs

Petar Trslic*, Matija Rossi, Luke Robinson, Cathal O'Donnel, Anthony Weir, Joseph Coleman, James Riordan, Edin Omerdic, Gerard Dooly, Daniel Toal

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

35 Citations (Scopus)
299 Downloads (Pure)

Abstract

This paper presents autonomous docking of an industry standard work-class ROV to both static and dynamic docking station (Tether Management System — TMS) using visual based pose estimation techniques. This is the first time autonomous docking to a dynamic docking station has been presented. Furthermore, the presented system does not require a specially designed docking station but uses a conventional cage type TMS. The paper presents and discusses real-world environmental tests successfully completed during January 2019 in the North Atlantic Ocean. To validate the performance of the system, a commercial state of the art underwater navigation system has been used. The results demonstrate a significant advancement in resident ROV automation and capabilities, and represents a system which can be retrofitted to the current ROV fleet.
Original languageEnglish
Article number106840
Number of pages16
JournalOcean Engineering
Volume196
Early online date13 Dec 2019
Publication statusPublished - 15 Jan 2020

Keywords

  • Resident ROV
  • Autonomous docking
  • Computer vision

Fingerprint

Dive into the research topics of 'Vision based autonomous docking for work class ROVs'. Together they form a unique fingerprint.

Cite this