ALICE
supported by the EU-project DG-XII, F-5 (Teleman)
Photos of ALICE - Technical reports
Contact: Uwe R. Zimmer (former member, uwe.zimmer@gmd.de)
Main tasks and problems discussed in the ALICE project include but are not limited to:
Self-Localization
The problem of localizing a dynamic robot continuously in relation to the current (previously unknown) environment is a elementary task for any mobile robot, with some global memory. In the present project worse conditions, like large drifts-effects, noise, low sensor resolutions, and systematic errors are assumed, in order to emphasis real world relevance of the proposed solutions.
Navigation
Navigation is obviously a basic task for each mobile robot, where a couple of different navigation demands can be considered. "Global navigation" refers to the problem of planning in the complete domain of the explored (and therefore known) environment, whereas "local navigation" handles problems of local manoeuvring without the necessity of global world models. Both problem fields are attacked in this project.
Mobile robot problems regarding target extraction and recognition are attacked in the SPIN project.
ALICE project
Tasks referring to spatial modelling respectively navigation behaviours of a mobile robot, are discussed. The generation of dynamic, self-organizing maps of sensor situations in one exploration phase, is introduced for the first time in this project. Experiments in a real world and hard realtime domain are performed on a platform called "ALICE". Navigation abilities are demonstrated regarding local as well as global navigation. The central idea of the ALICE project is a simple, but adaptive mobile robot of practical relevance, where extremely worse conditions (regarding drift, sensors, systematic errors, and available computer power) are assumed.
Technical reports (check here for further information)
|