This week I worked on the design for the SLAM subsystem. After doing some research, I determined that a Scan Matching-based 2D Lidar SLAM method would probably be the most effective for our use case. I read the papers A Review of 2D Lidar SLAM Research (Yan et al. 2025) and A Flexible and Scalable System with Full 3D Motion Estimation (Kohlbrecher et al. 2011) as part of this research. The first provided explanations of many SLAM systems which I used to evaluate which would be most effective for our use case. The second went into greater detail on a specific Scan Matching-based algorithm called Hector SLAM, which they also provide in open-source software. I think that this is a good initial design for our SLAM system, and that the open-source software can be used as a starting point for our software. I also worked on the slides for the design review presentation.
My progress right now is on schedule. My goal was to have an initial design for the SLAM algorithm by the end of this week, which I think I have accomplished.
By next week, I hope to have completed an initial version of the software that will demonstrate that the design is feasible, and to begin working on implementing it on the Raspberry Pi. I think that this should be doable by using the open-source software for Hector SLAM as a starting point for our software for the SLAM system.