Indoor positioning system using optical beacons for robot localization

• Main Author: Sathyamangalam Sanaullah Sajith.
• Other Authors: Seet Gim Lee, Gerald
• Format: Theses and Dissertations
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering::Mechanical engineering::Robots
• Online Access: http://hdl.handle.net/10356/41814
• Institution: Nanyang Technological University

The Robot localization is an imperative parameter in the field of robotics, to make the mobile robot autonomous. A primitive knowledge for a robot is that it should be able to recognize its position in an unknown environment to execute a given task & make it meaningful. In outdoor environment, GPS is used as general technique for robot localization, but when it comes to indoor environment GPS is not a precise solution. GPS doesn't work properly in an indoor environment since their signals are barren by building & other obstacles. It has become a demanding task to provide information to robot about its position and orientation in an unknown indoor environment. In recent years different types of indoor positioning systems were developed & they differ from one another in working principle, accuracy, precision and cost. Every indoor positioning system has its own pros and cons. In this dissertation, a study is made on "Northstar Indoor Localization" system which works on optical beacons projected on indoor environment such as ceiling. The optical beacons become reference for a detector (optical sensor) to find its position and orientation in an indoor environment. This dissertation includes experimentation that is conducted on the localization system to measure its performance. The different variables and parameters that affect the performance of Northstar localization system are studied. Accuracy, repeatability, deviation were calculated with different parameters. The best set of parameters required to obtain good performance on localization is determined. This experimental study will give an idea about the capability of localization system and one can identify the localization difficulties, before integrating it on robots. The experimental results reveal that the Northstar localization system can be used for small indoor environment (4 meter square area) with positional error of 15 cm. Northstar localization system performance results are compared with the other localization system such as the Ubisense and Cricket Motes in terms of accuracy, scale and cost.