` Intelligent and Mobile Robotics

Large Maps Framework (LaMa)

The Large Maps framework - LaMa - is intended to acquire, store and handle spatial knowledge about large diverse environments. The framework focuses on providing information suitable for navigation, localization, spatial reasoning, planning, human-machine and machine-machine interaction. Framework stores information necessary to know how to distinguish one place from another and to traverse between them. Modular architecture allows easy incorporation and cooperation of methods, sensors and behaviors in this framework.

The representation of spatial knowledge is based on the cognitive theories of the human spatial knowledge. The main structure is a topological map, a directed graph which consists of vertices and edges, likewise existing robotic implementation of topological maps. The vertices stand for places in the environment whereas the edges represents navigation paths. All the knowledge can be stored with the measure representing the uncertainty of the knowledge. The subjective logic is used for uncertainty representation.

This classical representation is extended by explicit association of the procedural knowledge. The procedural knowledge represent the knowledge how to do something in contrast to declarative knowledge describing what it is. It is the information which is necessary to know how to get from one place to another or how to distinguish one place from another. The algorithmic part of the procedural knowledge is denoted as Jockey, as, in certain sense, are “riding” the robot. It means that Jockeys have direct access to the robot hardware and are able (and allowed) to control the robot actuators, to read the robot sensors and to process raw data gathered from the robot’s sensors. Jockeys provide the unified way to handle different procedural knowledge.

Downloads

https://github.com/lama-imr

ROS LaMa packages

People involved:

Karel Košnar, Gaël Ecorchard, Miroslav Kulich

Contact:

Karel Košnar

Demos:

see more demos ...

Selected publications:

  • Krajník, T. - Faigl, J. - Vonásek, V. - Košnar, K. - Kulich, M. - et al. Simple, Yet Stable Bearing-Only Navigation In: Journal of Field Robotics. 2010, vol. 27, no. 5, p. 511-533. ISSN 1556-4959.
  • Přeučil, L. - Štěpán, P. - Krajník, T. - Košnar, K. - Rossum, A.V.R. - et al. Cognitive World Modeling In: Symbiotic Multi-Robot Organisms. Berlin: Springer, 2010, p. 165-183. ISBN 978-3-642-11691-9.
  • Košnar, K. - Krajník, T. - Vonásek, V. - Přeučil, L. LaMa - Large Maps Framework In: Proceedings of Workshop on Field Robotics, Civilian-European Robot Trial 2009 [CD-ROM]. Oulu: University of Oulu, 2009, p. 9-16. ISBN 978-951-42-9176-0.
  • Košnar, K. - Krajník, T. - Přeučil, L. Visual Topological Mapping In: European Robotics Symposium 2008. Heidelberg: Springer, 2008, p. 333-342. ISBN 978-3-540-78315-2.
  • Košnar, K. - Přeučil, L. - Štěpán, P. Topological Multi-Robot Exploration In: Proceedings of the IEEE Systems, Man and Cybernetics Society United Kingdom & Republic of Ireland Chapter 5th Conference on Advances in Cybernetic System. New York: IEEE - Systems, Man, and Cybernetics Society, 2006, p. 137-141. ISSN 1744-9170.

Projects and funding:

PostDoc grant no. 13-30155P of the Czech Science Foundation
CAK Project No. 1M0567 National Research Programme (2005-2011) of the Ministry of Education of the Czech Republic