Original Projects

rania-fds was a capstone project designed in Fall 2023 and developed in Spring 2024 by myself and two other peers. We had aimed to use 2D (radial) LiDAR sensors statically positioned on the floors of individual rooms for sensing outlines of persons after samples of the geometry of the environment were culled. A trained kNN system would then be applied on a normalization of these outlines, taking a fixed number of means on samples from intervals on candidate outlines, to make a determination of the fall state.

Despite having a fairly comprehensive vision and design, this project was largely unfinished after our team's total proficiency in software development was overestimated as we came to implement our devised architecture. Even with our deficiency, we were able to produce a working program for computing local, cached calibrations for placed sensors to filter the static geometry from any and each LiDAR sensor. We also left behind a framework containing the [untested] pipelines structured for concurrent processing and multithreading which was anticipated to be left for future efforts to work on.