Towards detached communication for robot cooperation
2005 (English)Licentiate thesis, monograph (Other academic)
This licentiate thesis deals with communication among cooperating mobile robots. Up until recently, most robotics research has focused on developing single robots that should accomplish a certain task. Now, it looks like we have come to a point where the need for multiple, cooperating robots is increasing since there are certain things that simply are not possible to do with a single robot. The major reasons, as described in this thesis, that make the use of multiple robots particularly interesting are distribution (it may be impossible to be in two places at the same time), parallelism (major speed improvements can be achieved by using many robots simultaneously), and simplicity (several, individually simpler, robots might be more feasible than a single, more complex robot). The field of cooperative robotics is multi-faceted, integrating a number of distinct fields such as social sciences, life sciences, and engineering. As a consequence of this, there are several sub-areas within cooperative robotics that can be identified and these are subsequently described here as well. To achieve coordinated behaviour within a multi-robot team communication can be used to ease the necessity of individual sensing (because of, for instance, calculation complexity), and with respect to this two different explicit approaches have been identified. As the survey presented here shows, the first of these approaches has already been extensively investigated, whereas the amount of research covering the second approach within the domain of adaptive multi-robot systems has been very limited. This second path is chosen and preliminary experiments are presented that indicate the usefulness of more complex representations to accomplish cooperation. More specifically, this licentiate thesis presents initial experiments that will serve as a starting point where the role and relevance of the ability to communicate using detached representations in planning and communication about future actions and events will be studied. Here, an unsupervised classifier is found to have the basic characteristics needed to initiate future investigations. Furthermore, two projects are presented that in particular serve to support future research; a robot simulator and an extension turret for remote control and monitoring of a physical, mobile robot. Detailed descriptions of planned future investigations are also discussed for the subsequent PhD work.
Place, publisher, year, edition, pages
Linköping: Linköpings universitet , 2005. , 112 p.
Linköping Studies in Science and Technology. Thesis, ISSN 0280-7971 ; 1192
IdentifiersURN: urn:nbn:se:liu:diva-33295Local ID: 19297ISBN: 91-85457-22-1OAI: oai:DiVA.org:liu-33295DiVA: diva2:254118
2005-09-30, G110, Hus G, Högskolan i Skövde, Skövde, 14:15 (Swedish)