Deployment of Ground and Aerial Robots in Earthquake-Struck Amatrice in Italy (brief report) Ivana Kruijff-Korbayov´ a ⇤ , Luigi Freda † , Mario Gianni † , Valsamis Ntouskos † , V´ aclav Hlav´ aˇ c ‡ , Vladim´ ır Kubelka ‡ , Erik Zimmermann § , Hartmut Surmann ¶ , Kresimir Dulic k , Wolfgang Rottner k and Emanuele Gissi ⇤⇤ ⇤ Multilingual Language Technology Lab, DFKI, Saarbr¨ ucken. Stuhlsatzenhausweg 3 D-66123 Saarbr¨ ucken, Germany † ALCOR Lab, DIAG, Sapienza University of Rome Italy ‡ Czech Technical University in Prague, Czech Republic § Fraunhofer IAIS, Schloss Birlinghoven, 53757 St. Augustin, Germany ¶ Fraunhofer IAIS, University of Applied Science Gelsenkirchen, Neidenburgerstr. 43 45877 Gelsenkirchen, Germany k Ascending Technologies GmbH (A Part of Intel), Konrad-Zuse-Bogen 4, 82152 Krailling, Germany ⇤⇤ Corpo Nazionale dei Vigili del Fuoco, Comando di Genova, Italy Abstract—We provide key facts about the TRADR project deployment of ground and aerial robots in Amatrice, Italy, after the major earthquake in August 2016. The robots were used to collect data for 3D textured models of the interior and exterior of two badly damaged churches of high national heritage value. I. I NTRODUCTION On September 1 2016 a team of the TRADR project 1 de- ployed two ground and three aerial robots in Amatrice, Italy, to assist the response after the 6.2-magnitude earthquake, which hit and devastated the town on August 24 2016, killing 234 people. The Italian firebrigade Vigili del Fuoco (VVF) asked TRADR for a one-day mission to deploy robots in two medieval churches: San Francesco and Sant’Agostino, both severly damaged and too dangerous for humans to enter due to the possibility of further collapse. The goal of the mission was to provide 3D textured models of the interior an exterior of these important national heritage monuments to facilitate precise damage assessment and plan preservation operations. II. SCENARIO DESCRIPTION First to inspect was the San Francesco church 2 (SF, Fig. 1). The only potential UGV ingress point was a side door, surrounded by large rubble (Fig. 2(a)). The only potential UAV ingress point was the hole left after the rose window, obstructed by a metal bar across the middle (Fig. 2(b)). Second was the Sant’Agostino church 3 (SA, Fig. 1) The only potential UGV ingress point was the front door, sur- rounded by large rubble. The only potential UAV ingress point was a narrow hole in the collapsed roof (Fig. 2(c)). 1 TRADR (http://www.tradr-project.eu/) is funded by EU-FP7-ICT grant No. 609763. We wish to thank all TRADR partners for their contributions. 2 http://www.amatriceturismo.it/la-citta-in-virtual-tour/luoghi-di- culto/basilica-di-san-francesco/ 3 http://www.amatriceturismo.it/la-citta-in-virtual-tour/luoghi-di- culto/chiesa-di-sant-agostino/ (a) (b) Fig. 1. (a) San Francesco Church; (b) Sant’Agostino Church (Sep 1 2016). (a) (b) (c) Fig. 2. (a) SF: UGV ingress ; (b) SF: UAV ingress; (c) SA: UAV ingress. III. DEPLOYMENT TRADR received the request 48 hours prior to the deploy- ment start. We promptly organized a team of ten TRADR researchers to travel the next day by cars and plane, bringing the robots and other equipment along. 4 At the site the TRADR team operated under the authority of a senior VVF commander. VVF provided additional logistics assistance, such as equipment transport in the red zone, power generator and tables and benches for the command post. A. Technology We used UGVs based on the BlueBotics Absolem 5 , two UAVs AscTec Falcon 8 6 and a DJI Phantom 4 7 (Fig. 3). 4 The team travelled from Italy (Rome); Czech Republic (Prague) and Germany (Munich, St. Augustin, Saarbr¨ ucken). One UGV was brought from Prague, one UGV and one UAV from Rome, and two UAVs from Munich. 5 http://www.bluebotics.com/mobile-robotics/absolem/ 6 http://www.asctec.de/en/uav-uas-drones-rpas-roav/asctec-falcon-8/ 7 http://www.dji.com/phantom-4 Proceedings of the 2016 IEEE International Symposium on Safety, Security and Rescue Robotics, EPFL, Lausanne, Switzerland, October 23-27, 2016 978-1-5090-4349-1/16/$31.00 ©2016 IEEE 278