Autonomous UAV Navigation in an Indoor Environment for Warehouse Inventory Recording

Overview. In this project, I plan to investigate techniques that will enable unmanned aerial vehicles (UAVs) to autonomously explore an indoor environment. While UAVs are currently used extensively in outdoor environments for various applications, including wildfire monitoring [1], agricultural surveying [2], and traffic management [3], their use in indoor applications have been fairly restricted, owing mainly to the difficulty to maneuver them in smaller indoor spaces and the inability to use critical functionalities, such a GPS-based navigation system. Nevertheless, there are many indoor applications where UAVs could provide a safe, reliable and resilient means to perform operations that are dangerous for humans. For example, UAVs could be used for searching for survivors inside a building following a disaster scenario, for reconnaissance operations inside chemical plants to detect toxic chemicals following accidents, for determining the health of engineering structures such as trusses and beams inside a building, and for inventory management within extreme indoor environments such as cold storages. To use UAVs in such indoor environments, there are several challenges that need to be addressed, including localizing the UAV without using a GPS, maneuvering the UAV to avoid collisions with obstacles, especially within tight indoor spaces in the presence of turbulence caused by the UAV rotors, and controlling the UAV to autonomously follow a prescribed flight path to complete its operation. To gain a better understanding of these issues, in this project, I will investigate techniques to address the problem of autonomously maneuvering a UAV to find a collision-free flight path to cover an indoor environment. I plan to use the results of this research toward implementing an indoor warehouse inventory management system using a UAV where the UAV autonomously flies in front of a shelf-like structure constructed in our lab, and uses its camera to systematically capture images at every rack in the shelf to report the inventory data back to a base station. My work in this project will build upon the background research on robotic coverage done in the CMANTIC robotics lab at UNO [5] and my experience on autonomously navigating a UAV called the Parrot AR Drone 2.0 [4] that is available in the lab. The results of this project will help to provide a better understanding of issues in autonomous indoor navigation of UAVs toward deploying them to aid humans in extreme indoor environments.