Overview / Usage

In this project we implement state-of-the-art navigation, perception, and decision making algorithms to allow efficient autonomous navigation and mapping capabilities for ground and aerial robots (drones). The robots are equipped with various sensors to allow perception of the environment, such as depth cameras and lidars (light-radar).

Methodology / Approach

We use the active-SLAM (Simultaneous Localization And Mapping) technique, in which the robots fuses information from its sensors to build a map of the unknown environment, locate itself in it, and optimally plan a safe course of action.

Technologies Used

The algorithms are implemented using C++ and MATLAB, and run under the industry-standard “Robot Operating System” (ROS) framework. Initial testing is done in simulation using the highly-realistic "Gazebo" engine. Our aerial robots are printed in-lab, and are embedded with a Single Board Computer (SBC), which runs the code.

Repository

http://www.khen.io