Senior Engineering Design Capstone Project
Project Date: August 2019 – Present
Collaborators: Prince Steven Annor, Sampanna Bhattarai
Faculty Mentors: Dr. Anthony Tzes, Dr. Jeremy Teo
Description of Project:
Many industries, such as agriculture, search and rescue, environmental management, and defense, rely on autonomous systems to navigate and map unfamiliar environments. Historically, drone swarms, which consist of two or more Unmanned Aerial Vehicles (UAVs) working collaboratively, have been used to perform tasks in these sectors. These drones need to map the target area, avoid obstacles, and estimate the relative pose and distance of other drones in the swarm. Current navigation and localization systems are often inaccurate due to sensor noise and do not perform robustly across various environmental conditions.
One cause of sensor noise is the oscillation of the drone as it flies. This unsteadiness makes it difficult to obtain quality data from sensors or cameras mounted on the drone. GPS localization, one of the most commonly used methods,does not work well at close range and cannot be used indoors. This project seeks to address these challenges and implement an optimal control design for environmental localization and mapping using drone swarms. After assessing the design constraints, an integrated approach using both 3-dimensional laser scanning (Lidar) and spherical camera imaging with computer vision analysis will be used to localize and control the UAV swarm for optimal area coverage.
Technologies Used: SLAM navigation, Lidar, Robot Operating System (ROS), Python, Turtlebot3, Gapter Drone
Presentation Delivered December 8, 2019: