3D sensors implementation for mobile robots and autonomous vehicles navigation

Abstract

Autonomous vehicles navigation, as well as mapping areas have endless commercial and
industrial applications. This article explains the development of an application that allows a
mobile robot to navigate autonomously, avoiding obstacles, until it reaches a specific coordinate
point defined by the user, using low cost 3D sensors. For the design and implementation of this
application, it is necessary to keep in mind that navigation and obstacle detection are based on
the use of 3D sensors. For this purpose, the Kinect was selected. The simulation and testing of
the project was done in a special simulation environment for robotic applications, called V-REP.
The navigation algorithm for obstacle avoidance was implemented through the potential fields
theory. Other sensors, as GPS and IMU, were used to determine the robot’s spatial position and
orientation, with the objective of defining a trajectory to the goal coordinate point, thus doing a
more efficient navigation. Finally, real data from the physical sensors was used to corroborate
the functioning of the application, with the purpose of comparing it with the previously obtained
results from the simulated part.