In a near future, social robots will help humans in everyday tasks and they will interact and cooperale with humans in dynamic indoor or ouldoor environments. We will find these lypes of robots in urban areas, building and prívate houses. The main goal of this project is lo advance in the design of robots lhat can interact, learn and cooperale with people for urban tasks, adapting themselves to the environment and changing conditions. Specifically we propose to develop new perception techniques for robots based on learning and adaptation, to incorporate visual-servoing techniques for the interaction between humans-robots, to explore new interaction techniques between mobile-aerial robots with humans and to develop new cooperation lechniques which facilitate the execution of urban tasks in for example, coordinate navigation, environmenl exploration or search for objects or persons in an urban site. We will follow the open research lines of the previous R&D projects RoboTaskCoop (DPl2010-17112) and Consolider Ingenio 2012 MIPRCV (CSD2007-00018) about on line learning and adaptation from the video camera and range laser signals of objects, persons and human activities. We will work in new techniques based on line boosting, SVM and matching techniques. In the humanrobot interaction issue, we will continue working in new models that explore the basic principies of human-robot interaction, for example lhe game search and hide. Johansson y Balkenius said that this is the ideal game to study the robot cognitive functions and the basic cognitive mechanisms for the interaction in mobile robots (the search and hide game requires a great number of cognitive functions which include search, navigation, planning, perception, etc.). We also will continue developing the of-line and on-line POMDPs methods. Moreover we will also incorporate the present results developed in the European project ARCAS (FP?-ICT-2011-7-287617), to design new interaction models between humans and aerial robots based on visual servoing. In the cooperation techniques, we will continue working in cooperative human-robot navigation and in building cooperative semantic maps that will be used for transport, exploration or search tasks. Using lhe previous results developed in other research projects, we will continue designing new navigation techniques between multiple robots and humans in urban areas, taking into account that the environment is dynamic. Finally, we will apply these developments to specific tasks: cooperative transportation, exploration and search of objects or persons in urban areas. We will use our IRI robots, the humanoids Tibi and Dabo, the autonomous vehicle and we will buy a new aerial robot (UAV) to study the interaction between aerial robots and humans. The new techniques will be tested in indoor scenarios and in the outdoor scenario Barcelona Robot Lab (BRL) al the UPC. This Project follows the research developed in CICYT's previous projects, in the Consolider Ingenio MIPRCV, in the European URUS Project and in the European ARCAS project.
Plan Estatal de Investigación Científica y Técnica y de Innovación 2013-2016
Programa Estatal de Fomento de la Investigación Científica y Técnica de Excelencia
Subprograma Estatal de Generación de Conocimiento
Excelencia: Proyectos I+D
Gobierno De España. Ministerio De Economía Y Competitividad, Mineco