Autonomous mobile robots (on ground, undersea or in the air) must localize themselves in order to navigate. They also need to build a map of their environments in order to plan their motion and accomplish their missions efficiently.

These localization and mapping problems are actually interdependant and have to be solved concurrently, hence the question of Simultaneous Localisation and Mapping (SLAM). This is a fundamental scientific problem and a necessary feature towards autonomy in mobile robots. This is also a mandatory capacity in many applications where external localization means are unavailable, such as planetary exploration.

This summerschool is organized by Raja Chatila, Michel Devy (LAAS-CNRS) and Juan Domingo Tardos (U. Zaragoza). It is the second of its kind after a first event in Stockhom in August 2002 (http://www.cas.kth.se/SLAM/). It will provide a comprehensive coverage of the key topics, state of the art in SLAM and application cases. The summer school will be a combination of lectures by well known experts in the field and computer exercises (MATLAB, C++/C) involving groups of students.

Topics include:

• Fundamentals of Filtering techniques
• Definition of the SLAM problem
• Representational issues (points, lines, grids, etc)
• Bayesian interpretation of SLAM
• Basic SLAM algorithm using EKF
• Data association
• Mapping of large environments
• Dynamic environments
• SLAM with vision
• Multirobot SLAM
• Case studies: indoor, land, sub-sea and air vehicles

• Wolfram Burgard, Univ. of Freiburg
• Raja Chatila, LAAS, Toulouse
• Henrik I. Christensen, KTH, Stockholm
• Andrew Davison, Univ. of Oxford
• Michel Devy, LAAS, Toulouse
• Hugh Durrant-Whyte, ACFR, Sydney
• Kurt Konolige, SRI
• John Leonard, MIT, Cambridge
• Andre Monin, LAAS, Toulouse
• Eduardo Nebot, ACFR, Sydney
• José Neira, Univ. of Zaragoza
• Paul Newman, Univ. of Oxford
• Salah Sukkarieh, ACFR, Sydney
• Juan D. Tardós, Univ. of Zaragoza
• Sebastian Thrun, Stanford Univ.