Robot navigation and localization is a fundamental problem in robotics. This thesis studies the problem for robots in environments that have many regular structures such as typical office environments in office buildings. While there have been many algorithms for robot navigation and localization, they are complicated and need a lot of computational power, none of them seem to work well in these environments. However in a regular and static environment, the navigation and localization problem becomes easier with the help of landmarks. Therefore, in this thesis I propose a simple and intuitive algorithm for robots with wheels and two types of sensors: infrared sensors and a low resolution stereo camera. This algorithm is implemented and field-tested on 914 PC-BOT, a commercial robot from...[ Read more ]

Robot navigation and localization is a fundamental problem in robotics. This thesis studies the problem for robots in environments that have many regular structures such as typical office environments in office buildings. While there have been many algorithms for robot navigation and localization, they are complicated and need a lot of computational power, none of them seem to work well in these environments. However in a regular and static environment, the navigation and localization problem becomes easier with the help of landmarks. Therefore, in this thesis I propose a simple and intuitive algorithm for robots with wheels and two types of sensors: infrared sensors and a low resolution stereo camera. This algorithm is implemented and field-tested on 914 PC-BOT, a commercial robot from WhiteBox Robotics.