A robot localization system solely based on distance sensors

A robust localization system that uses only distances to a small set of non-synchronized radio beacons at very small update rates. This innovative location system has been developed by Professor Fernando Caballero (Universidad de Sevilla) and Professor Luis Merino (Universidad Pablo de Olavide) with the collaboration of PAL Robotics, in a successful example of the benefits of University/Industry cooperation. We had the pleasure to receive both Professors at PAL Robotics’ offices, where they gave us a presentation on their achievements.

The localization system’s code is open-source and ROS-based, it is publicly available at GitHub. During the visit Prof. Caballero, Prof. Merino and our team successfully integrated the location system with StockBot, using PAL Robotics’ radio beacons from Idolink.

Even without synchronization of the radio beacons, and at a low measure frequency of 0.1Hz, the localization system proved to be very robust and efficient, successfully replacing AMCL. The video shows Stockbot using the range_only_localization system for navigation (no AMCL) and how this localization system can recover the robot from a wrong initial position.

Possible applications of this system are countless, as localizing the robot in very dynamic and cluttered environments or as a recovery tool for AMCL, for example. We are confident that this is an important and very useful contribution to the robotics community.

Prof. Fernando Caballero and Luis Merino with StockBot team for localization system integration ROS

Prof. Fernando Caballero (University of Sevilla) and Luis Merino (University Pablo Olavide) with StockBot team after the integration of the localization system to StockBot.

This entry was posted in Articles and tagged , , , , , , , , , , , , , , , , , , , , . Bookmark the permalink.