Mapping, Localization, and Trajectory Estimation with Mobile Robots Using Long-Range Passive RFID

Beschreibung

This thesis addresses the use of radio-frequency identification (RFID) for three fundamental tasks in mobile robotics: mapping, self-localization, and trajectory estimation. These topics have widely been studied in recent years, because they are regarded as key ingredients to autonomous robots. In particular, we investigate the application of long-range passive RFID in the work at hand. This RFID standard is expected to play a major role in the automation of identification processes in industry and commerce. We present solutions to the automated inventory of RFID-tagged goods, which we interpret as mapping the locations of RFID transponders. In order to quickly provide the required model of the detection behavior of the RFID reader on board the mobile robot, we devise a semi-autonomous learning method for probabilistic sensor models. Moreover, we compare different techniques to self-localize in an RFID-tagged environment. In the last part, we show how the trajectory of the robot can be estimated in previously unknown surroundings by tailoring approaches from the field of simultaneous localization and mapping (SLAM).
This thesis addresses the use of radio-frequency identification (RFID) for three fundamental tasks in mobile robotics: mapping, self-localization, and trajectory estimation. These topics have widely been studied in recent years, because they are regarded as key ingredients to autonomous robots. In particular, we investigate the application of long-range passive RFID in the work at hand. This RFID standard is expected to play a major role in the automation of identification processes in industry and commerce. We present solutions to the automated inventory of RFID-tagged goods, which we interpret as mapping the locations of RFID transponders. In order to quickly provide the required model of the detection behavior of the RFID reader on board the mobile robot, we devise a semi-autonomous learning method for probabilistic sensor models. Moreover, we compare different techniques to self-localize in an RFID-tagged environment. In the last part, we show how the trajectory of the robot can be estimated in previously unknown surroundings by tailoring approaches from the field of simultaneous localization and mapping (SLAM).

Autor*in

Philipp Vorst

Stimmen zu »Mapping, Localization, and Trajectory Estimation with Mobile Robots Using Long-Range Passive RFID«

Details

Link teilen