INTELLIGENT MOBILE MULTI-ROBOTIC SYSTEMS: SOME CHALLENGES AND POSSIBLE SOLUTIONS Fl´ avio S. Corrˆ ea da Silva, Renata Wassermann, Ana Cristina V. Melo, Leliane N. Barros, Marcelo Finger Dept. of Computer Science, University of S˜ ao Paulo Rua do Mat ˜ ao, 1010, S ˜ ao Paulo, BRAZIL Keywords: Distributed multi-agent robotic systems, software for robotic agents, robotic surveillance and security. Abstract: Intelligent mobile multi-robotic systems (IMMRSs) are coordinated systems of autonomous mobile robots endowed with reasoning capabilities. This sort of systems requires the integrated application of a variety of state-of-the-art techniques developed within the realm of Artiﬁcial Intelligence, as well as instigates the further development of different specialisations of Artiﬁcial Intelligence. In the present article we examine some of these techniques and specialisations, discuss some speciﬁc challenges proposed to the ﬁeld of Artiﬁcial Intelligence by IMMRSs, and suggest possible solutions to these challenges. In order to make our presentation more concrete, we employ throughout the article a speciﬁc example of IMMRS application, namely security surveillance of an empty building by a team of robots. 1 INTRODUCTION Intelligent mobile multi-robotic systems (IMMRSs) are coordinated systems of autonomous mobile robots endowed with reasoning capabilities. These systems require the integrated application of a variety of state- of-the-art techniques developed within the realm of Artiﬁcial Intelligence (AI), as well as instigates the further development of different specialisations of AI. In the present article we examine some of these techniques and specialisations, discuss some speciﬁc challenges proposed to AI by IMMRSs, and sug- gest possible solutions to these challenges. In order to make our presentation more concrete, we employ throughout the article a speciﬁc example of IMMRS application, namely security surveillance of an empty building by a team of robots (Gerkey et al., 2004): Given an environment, modelled as a connected polygonal free space, a ﬁxed number of searchers – autonomous mo- bile robots equipped with cameras, each camera having a ﬁxed angular aperture – and an unknown number of evaders – autonomous entities which can move arbitrarily fast – de- termine trajectories for each of the searchers so that the de- tection of all evaders is guaranteed. This problem has high computational complex- ity with respect to the complexity of the environ- ment (e.g. characterised by the number of edges of the polygon that comprises it) and the number of searchers. Moreover, determining the optimal (i.e. minimum) quantity of searchers given an environment is NP-hard (Gerkey et al., 2004). This problem, however, does not take into account a few parameters that can be important to improve its accuracy for practical applications. In the following sections we analyse this problem under the light of different specialisations of AI, which provide us with conceptual tools to reﬁne its description. The present work can be thus regarded as a reﬁnement of the work presented in (Gerkey et al., 2004), aiming at taking it from foundational research that provides some math- ematically well founded guidelines for IMMRS to a de facto applied work that can effectively be used to build IMMRS. In section 2 we add to the model the inherent un- certainties of the estimates of where a searcher and an evader are at a given moment. In section 3 we con- sider the fact that sometimes the sensed environment and its corresponding model may disagree, requir- ing a revision of the model of the environment. The reﬁnements of multi-robotic models that result from taking into account the features discussed in these two sections can lead to high consumption of computa- tional resources. In section 4 we focus on the efﬁ- ciency of inference systems, to counterbalance this consumption of resources. In section 5 we consider the possibility of the searchers communicating with 479 S. Corrêa da Silva F., Wassermann R., Cristina V. Melo A., N. Barros L. and Finger M. (2005). INTELLIGENT MOBILE MULTI-ROBOTIC SYSTEMS: SOME CHALLENGES AND POSSIBLE SOLUTIONS. In Proceedings of the Second International Conference on Informatics in Control, Automation and Robotics - Robotics and Automation, pages 479-485 DOI: 10.5220/0001190904790485 Copyright c  SciTePress