Examiner: Dr. Per Lindström.

The use of Laser Range Finder or ladar in localization of mobile robots is increasing since last decade. One of the reasons of its popularity is that the ladar is faster and performs simpler modeling of surrounding environment than its counter part video camera. The aim of this thesis work is to test different 2D ladar based methodologies in order to find the best applicable methodology for 2D ladar based localization in unknown environments. In the beginning of thesis work, experimental evaluations of some ladar characteristics were performed. Later on, feature and other ladar based localization techniques were tested. We found that the pixel based localization methodology proposed by Jorma Selkäinaho in his PhD. desertion is simpler and faster. We further tested this methodology in different domains and discussed its drawbacks. A solution to minimize these drawbacks has been proposed. Experimental results show that the proposed methodology is faster and more reliable than the original one. 3