ECE Ph.D. Student
About the project
The goal of my thesis is to create a robust platform and collection of algorithms that will simultaneously locate a group of robots inside a building and cooperatively map the interior of a building. The result of the mapping should be a fully rendered 2.5D model of the interior of a building. Additionally, a relative coordinate system of the entire building will be created such that the sensor platforms and any anomalies will have coordinates that are relevant to the robot group. The goal is to fuse various forms of sensor readings, including stereo vision readings, magnetic readings, acoustic readings, and inertial measurements. Furthermore, the entire system will be generalized to a group of robots, which can map and explore faster than a single robot. The result of 2.5D Cooperative Simultaneous Localization and Mapping is to have a group of robots perform the task of SLAM better by virtue of using more mapping sensors and more cooperatively controlled platforms.
A key constraint in the project is the cost of the individual platforms. Our goal is to make them “expendable.” Typically, most robots use a LIDAR sensor costing on the order of $5,000. The goal of my thesis is to create an entire robot for less than the cost of a single LIDAR sensor (under $5,000). Cheap, off-the-shelf components will be used to demonstrate the fact that the robots are expendable and easy to build. The indoor mapping will provide both humans and machines a common map and coordinate system that can be used for other algorithms and in human search-and-rescue scenarios.
Professor Gilmer Blankenship, Director
J. Karvounis, J. Patel, M. Stanley graduate assistants
Office: 1427 A.V. Williams Building