Three underwater remotely operated vehicles (ROVs) were used over a five day period to inspect critical infrastructure and to assist with victim search and recovery at six sites in the Iwate Prefecture following the Tohoku Earthquake and Tsunami. Unmanned marine vehicles have been used since 2004 for disaster recovery operations but in limited applications, in single areas, and in short deployment durations. The joint IRS-CRASAR deployment matched robots for missions specified by civilian municipalities and the Japanese Coast Guard. The ROVs successfully allowed a fishing port to be re-opened and searched for victims trapped underwater in five different locations in varying areas (marinas, bridge debris, and waterfront residential areas) that could not be searched by manual divers. From a scientific perspective, the deployment provides a corpus of 15 hours of data of how rescue robots can be used. It illustrates that rescue robots are i) valuable for both economic and victim recovery, not just response, ii) that disaster robots need to be optimized for the unique missions and stakeholder needs, and iii) that human-robot interaction remains a challenge. This paper also identifies new areas for research: computer vision and cognitive engineering, cyber-physical systems, heterogeneous multirobot coordination, human-robot interaction, simulation and geographical information systems.

No 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 m2 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.

The Gulf of Mexico Coastal Hypoxia Glider Experiment was designed to assess the feasibility of using ocean glider technology in the coastal hypoxic zone of the northern Gulf of Mexico in Summer/Fall 2014. The objectives were (1) to coordinate and operate multiple autonomous buoyancy ocean gliders in depths less than 50 m and (2) to determine how close to the bottom gliders can reliably reach without making contact. Strong vertical and horizontal stratification gradients, strong coastal currents, and the low-oxygen conditions that occur within the lower water column characterize the coastal area of the northern Gulf of Mexico. These environmental conditions combine with the presence of more than 5,000 surface piercing oil/gas structures to make piloting and navigation in the region challenging. We quantify glider performance to assess the usefulness of buoyancy gliders to address the National Oceanic and Atmospheric Administration Action Plan goal to monitor the spatial extent, duration, and severity of the Gulf hypoxic zone. We find that the gliders, despite the operational challenges, were consistently able to travel from the surface to the oxygen-depleted depths of subpycnocline waters, that is, within 2 m of the ocean bottom. Our assessment is that gliders are able to provide real-time observations suitable to monitor coastal hypoxia.