Marine heterogeneous multirobot systems at the great Eastern Japan Tsunami recovery

Abstract This field report describes two deployments of heterogeneous unmanned marine vehicle teams at the 2011 Great Eastern Japan Earthquake response and recovery by the Center for Robot‐Assisted Search and Rescue (USA) in collaboration with the International Rescue System Institute (Japan). Four remotely operated underwater vehicles (ROVs) were fielded in Minamisanriku and Rikuzentakata from April 18 to 24, 2011, for port clearing and victim recovery missions using sonar and video. The ROVs were used for multirobot operations only 46% of the time due to logistics. The teleoperated ROVs functioned as a dependent team 86% of the time to avoid sensor interference or collisions. The deployment successfully reopened the Minamisanriku New Port area and searched areas prohibited to divers in Rikuzentakata. The IRS‐CRASAR team planned to return from October 18 to 28, 2011, with an unmanned aerial vehicle (UAV), an autonomous underwater vehicle (AUV), and an ROV to conduct debris mapping for environmental remediation missions. The intent was to investigate an interdependent strategy by which the UAV and AUV would rapidly conduct low‐resolution scans identifying areas of interest for further investigation by the ROV. The UAV and AUV could not be used; however, the ROV was able to cover 80,000 m 2 in 6 h, finding submerged wreckage and pollutants in areas previously marked clear by divers. The field work (i) showed that the actual and planned multirobot system configurations did not fall neatly into traditional taxonomies, (ii) identified a new measure, namely perceptual confidence, and (iii) posed five open research questions for multirobot systems operating in littoral regions. © 2012 Wiley Periodicals, Inc.