Decide on Localization Algorithm

Background: given that we are using a fairly sophisticated system, it would be a shame to hard code everything when we are capable of more. The main options are a Histogram localization (fairly basic), Adaptive Monte Carlo Localization (AMCL), Hector SLAM, Gmapping, Kalman Filter, a hybrid of these, or potentially others.

Gmapping, AMCL and Hector SLAM are very common with ROS applications.

Decision: how will sensor data be fused? Is ultrasound a backup for lidar, or complimentary? Is IR our main localization strategy, or a combination of the field viewed by Lidar? Does Hector SLAM require a premade map? Do we want to include odometry in the ROS environment, or keep encoder data internal to the Arduino? Also, could IR data be used to add weight to probabilistic nature of localization?

Useful links: https://adityakamath.github.io/2021-08-31-first-steps-with-nav-stack/

https://www.researchgate.net/publication/352675079_Comparison_of_Two_SLAM_Algorithms_Provided_by_ROS_Robot_Operating_System