Project Definition

This project aims to perform state estimation on robots with variable amounts of communication between robots to increase the accuracy of robot localization.

The goal of this project is to investigate the issues surrounding cooperative localization, which is the process by which multiple robots work together to determine their locations. Specifically, the project will focus on the application of simultaneous localization and mapping in cooperative localization scenarios, with a focus on swarm robotics. Swarm robotics involves the use of multiple, low-cost robots to accomplish a task, such as distributed sensing, search and rescue, or distributed 3D printing. Cooperative localization is particularly relevant in the context of swarm robotics because it allows for cost-effective localization of a large number of robots. However, the benefits of cooperative localization extend beyond swarm robotics, as it can improve the performance of any fleet of robots.

This project will explore three different scenarios of cooperative localization, where the base algorithm used for simultaneous localization and mapping (SLAM) will be the Extended Kalman Filter SLAM (EKF-SLAM) with the state estimate of 33 items and a covariance estimate of a 33x33 matrix. The correction step is replaced by the Covariance-Intersection algorithm when communication between robots is present due to the assumption of independent uncertainties being violated in this case.