SLAM

Extracting an accurate and complete map of the world is among the most difficult
tasks in mobile robotics. In this regard, enabling the mobile robot to learn the map of the world and choose the optimum destination
to continue the autonomous scanning of the environment, is the topic of this research. For this purpose,
the Rao-Blackwellized particle filter for occupancy grid maps is used to determine the pose of the robot and
create the map of the world. Furthermore, a two layered architecture is used for motion planning, considering
the path planning and obstacle avoidance. The fundamental of exploration problem is determining the optimum
goal among the other potential goals. The main objective of this research is to gain the ability of selecting the
goal without human intervention and creating the map of the world autonomously. Hence in this work, the
frontier-based exploration strategy is applied to the robot for the purpose of selecting the optimum goal. For
practical implementation of the exploration problem, Robot Operating System(ROS) framework is used as the
software platform of the robot. Moreover, the area of the fifth floor of the mechanical engineering department
of the Isfahan university of technology is chosen as the unknown environment to test the robot. As a result, the
frontier-based approach turned to be an effective algorithm for exploring the large and cluttered environments.
 
People
Mr. Majid Fattahian, MSc in Mechanical Engineerig, Isfahan University of Technology
Dr. Mehdi Keshmiri, Professor, Robotics, Isfahan University of Technology
Dr. Hossein Karimpour, Assistant Professor, Robotics and control, University of Isfahan
----