Received: 14 June 2017 Revised: 31 January 2018 Accepted: 11 February 2018 DOI: 10.1002/rob.21784 FIELD REPORT A real-time field experiment on search and rescue operations assisted by unmanned aerial vehicles Tomasz Niedzielski Mirosława Jurecka Bartłomiej Miziński Joanna Remisz Jacek Ślopek Waldemar Spallek Matylda Witek-Kasprzak Łukasz Kasprzak Małgorzata Świerczyńska-Chlaściak Department of Geoinformatics and Cartography, Faculty of Earth Sciences and Environmental Management, University of Wrocław, Wrocław, Poland Correspondence Tomasz Niedzielski, Department of Geoinformatics and Cartography, Faculty of Earth Sciences and Environmental Management, University of Wrocław, pl. Uniwersytecki 1, 50-137 Wrocław, Poland. Email: tomasz.niedzielski@uwr.edu.pl Abstract This paper reports on the performance of a novel system for supporting search and rescue activi- ties, known as SARUAV (search and rescue unmanned aerial vehicle), in a field experiment during which a real-world search scenario was simulated. The experiment took place on March 2–3, 2017, at two sites located in southwestern Poland. Three groups acted in the experiment: (1) SARUAV and unmanned aerial vehicle (UAV) operators, (2) ground searchers, and (3) participants who sim- ulated being lost. In the uncomplicated topography without snow cover, the system identified the lost persons, and ground searchers found them 31 min after the SARUAV report had been dis- seminated. In the mountainous area covered with snow, one person was found within 9 min after searchers received the SARUAV report; however, the other two persons were not identified by SARUAV. The field experiment served as a proof of concept of the SARUAV system, confirmed its potential in person identification studies, and helped to identify numerous scientific and technical problems that need to be solved to develop a mature version of the system. KEYWORDS lost person, nested k-means, ring model, target detection, unmanned aerial vehicle 1 INTRODUCTION 1.1 Background Search and rescue (SAR) activities are nowadays commonly assisted by information technology and various sensors mounted on manned or unmanned platforms (Murphy et al., 2008). Geographic information system (GIS) offers a framework for developing tools to assist SAR procedures. There are several examples for such tools. One of them is “SARPlan”, which is a decision support system designed to increase the probability of locating a lost aircraft and survivors (Abi-Zeid & Frost, 2005). Another example, “MapSAR”, is an ArcGIS solution elaborated to provide ground searchers with maps to plan the search (Durkee & Glynn-Linaris, 2012). There also exists “System GIS w Podhalańskiej Grupie GOPR” that is implemented to produce maps of person's mobility and the probability of area (POA) to allow SAR teams to optimize the search procedures (Chrustek, Zaród, & Filipkowska, 2012). It is based on data from the International Search and Rescue Incident Database and from a seminal book by Koester (2008). It is worth mentioning “CenWits”, which uses radio frequency modules and Global Positioning System (GPS) measurements to determine the most probable person's location in the wilderness (Huang, Amjad, & Mishra, 2005). A particular role in enhancing SAR performance is played by unmanned aerial vehicles (UAVs), which are informally known as drones. They are cheap, operationally flexible, readily available devices, which can be rapidly deployed during a SAR action. There- fore, UAVs may support SAR activities at different phases and kinds of search missions. The ALCEDO system uses drones to carry out res- cue of avalanche victims through autonomous UAV-assisted search for a buried person (www.alcedo.ethz.ch). The SHERPA project aims to develop a system that combines ground and aerial robots to support alpine SAR missions (Marconi et al., 2013). Recently, Sun, Jiang, and Wen (2016) have elaborated a target detection and positioning sys- tem for SAR applications. The availability of UAVs for SAR purposes can be achieved through volunteer networks, such as the Search With Aerial Rc Multi-rotor known as SARDrones (sardrones.org). The net- work associates 1,110 UAV pilots who are able to serve as UAV oper- ators in SAR missions. Some commercial UAV companies often help in SAR activities by offering equipment and personnel at no cost (e.g., FlyTech UAV, Polish drone manufacturer, personal communication, March 13, 2017), which increases chances for an individual to survive. J Field Robotics. 2018;1–15. c  2018 Wiley Periodicals, Inc. 1 wileyonlinelibrary.com/journal/rob