home



















IRMA

Autonomous Mobile Robot Research


Members (2005-2010)

Tomás Arredondo V., Wolfgang Freund, Cesar Muñoz, Nicolas Navarro, Fernando Quirós


In this project we study and develop evolutionary multi-objective algorithms for autonomous mobile robots.  That is the study and development of algorithms based on evolution to improve the capabilities of current mobile robots.  The principal methods for the developmente of evolutionary algorithms known currently include genetic algorithms and genetic programing.  Both are based in an artificial model of evolution in order to enable a global search of posible solutions to complex problems such as robotic navegation and environmental mapping.  It is well known that these tipes of methods are capable of finding solutions that many times  have lower cost and complexity to problems that escape solutions based on local search techniques.

In practical terms for industry and engineering this is a problem of interest because it is desired to utilize mobile robots in tasks that many times are boring (ex: cleaning floors or windows), dangerous (ex: removing mines or bombs), unclean (ex: inspecting sewage lines), impossible (ex: studying deep underwater trenches) or simply mundane (ex: serving drinks in parties or directing groups in museums).  Each of these aplications is composed of many objectives (multi-objective) that have to be completed in order to complete the aplication of interest.  Some of these objectives are: navegation to an objective, modeling and environment, object detection and moving objects to points of interest.

These posible applications are limited because the costs of implementation of a mobile robot for these aplications if prohibitive.  That is why it is important to be able to build a mobile robot that can efectively accomplish these objectives with low cost sensors and actuators.

With this in mind we will first study the state of art results in evolutionary algoritms por individual objectives such as navegation and environmental modeling of mobile robots.  This first stage is to determine with certainty the capabilities and limitations of current algorithms.  In this stage we will utilize simulators of mobile robots in order to make this process more efficient.  In a second stage we will develop new algorithms that will be capable of completing tasks that are composed of combinations of objectives not previously investigated.  This stage will be studied with simulators and with mobile robots that will include sensors and actuators.

ruta reciente de YAKS modificado (Fuzzy YAKS)

A recent route taken by a modified version of YAKS (Fuzzy-YAKS)


References

Arredondo, T.: "Fuzzy Motivations in Behavior Based Agents", book chapter in Smart Information and Knowledge Management: Advances, Challenges, and Critical Issues, Series: Studies in Computational Intelligence, Vol. 260, Szczerbicki, Edward; Nguyen, Ngoc Thanh (Eds.), Springer, ISBN: 978-3-642-04583-7, Berlin,  2010. [Springer]

Arredondo, T.. Freund, W.: ''Motivation and Local Image Entropy Based Measures in Evolutionary Mobile Robot Navigation'', book chapter in Mobile Robots Navigation, Barrera, A. (editor), In­tech, ISBN: 978-953-307-076-6, Vienna, Austria, 2010. [Intech]

Arredondo, T., Freund,W., and Muñoz, C.: "Entropy Based Diversity Measures in Evolutionary Mobile Robot Navigation". Lecture Notes in Computer Science, Vol. 5027. Springer, Berlin (2008) p. 129-138. [pdf-draft]  [Springer]

Freund,W., Arredondo, T., and Muñoz, C., "Applying Real-Time Survivability Considerations in Evolutionary Behavior Learning by a Mobile Robot", book chapter in “Frontiers in Evolutionary Robotics”, Lazinica, A. (editor), Intech, ISBN 978-3-902613-19-6, Vienna, Austria, 2008. [Intech]

Arredondo, T., Vásquez, F., Candel, D., Dombrovskaia, L., Agulló, L., Córdova, M., Latorre-Reyes, V., Calderón, F., Seeger, M.: "Dynamic Penalty Based GA for Inducing Fuzzy Inference Systems". Lecture Notes in Computer Science, Vol. 4756. Springer-Verlag, Berlin (2008) p. 957-966. [pdf-draft]  [Springer]

Arredondo, T., Freund,W., Muñoz, C., and Quirós, F.: "Learning Performance in Evolutionary Behavior Based Mobile Robot Navigation". Lecture Notes in Artificial Intelligence, Vol. 4827. Springer-Verlag, Berlin (2007) p. 811-820.  [pdf-draft]  [Springer]

Freund,W., Arredondo, T., Muñoz, C., Navarro, N., Quirós, F.:"Real-Time Adaptive Fuzzy Motivations for Evolutionary Behavior Learning by a Mobile Robot".   Lecture Notes in Computer Science, Vol. 4293. Springer, Berlin / Heidelberg (2006) p.101-111. [pdf-draft]  [Springer]

Arredondo, T., Freund,W., Muñoz, C., Navarro, N., and Quirós, F.: "Fuzzy Motivations for Evolutionary Behavior Learning by a Mobile Robot". In: Ali, M., Dapoigny, R.(eds): Innovations in Applied Artificial Intelligence. Lecture Notes in Artificial Intelligence, Vol. 4031. Springer-Verlag, Berlin (2006) p. 462-471. [pdf-draft]   [Springer]


This work is ongoing as part of the IRMA-II project.... IRMA-II