Summer Project Robotic mapping

Robotic mapping is a problem of building models (maps) of environment from sensor measurements by mobile robots. Models of environment are the basic knowledge an autonomous robot must possess in order to carry out any higher level tasks like planning. Therefore robotic mapping is one of the most intensively studied problems of mobile robotics in recent years. The mapping problem is also known as simultaneous localisation and mapping problem (SLAM ), since imperfect odometry sensors create a problem of determining robot‘s pose related to the map. The main subject of the report is a real-time algorithm addressing the problem of building maps from geometric primitives like lines or surfaces. The underlying idea is very simple the EM algorithm estimating the parameters of suitably constructed Gaussians which represent map components lines (2D) or surfaces (3D). Object mapping undoubtedly brings some advantages like reduction of map complexity (lossy map compression) or simplification of high-level postprocessing based on a small number of geometric shapes rather then a huge number of objects like grid cells. Unfortunately, it also poses many challenges, where the most prominent is probably the map consistency problem in large cyclic environments. The mapping algorithm was designed as a part of a larger client-server software platform for mobile robots written in the last few months at the University of Birmingham. The platform is a kind of framework for developing miscellaneous algorithms for “intelligent” robots; it is an extension to the popular Saphira/Aria robot control system.