Sam Lonneman
Sam Lonneman
My Work
Rubik's Cube Solving Robot
June 2021 - August 2022
I recently completed a 12+ month personal project to design, fabricate, and program a robot to solve a Rubik’s Cube using autonomous camera input. My motivation for this project started from my long-time passion for embedded systems and physical computing, along with a fascination with the Rubik’s Cube and the mathematics behind it. The work consisted of writing a program to interpret camera input to determine the state of the scrambled cube, then apply complex algorithms to generate an optimized sequence of turns which the physical robot uses to solve the cube. Apart from the programming, I also used SolidWorks (CAD software) to design several original mechanical parts to be 3D printed and assembled. One of the most challenging parts of this project was programming the camera to self-calibrate to the cube’s colors at the start of each run so that it works consistently under differing light levels. Another challenge during this project was teaching myself how to handle the wiring and power system, along with writing the code to precisely drive the motors, both tasks I was previously inexperienced in. Now that it has recently been completed, it solves the cube consistently in around 35-45 seconds, has safety mechanisms including manual motor power cutoff, and runs completely stand-alone on its attached raspberry pi without being plugged in to an external computer. Along the way, I kept a journal of progress updates and objectives to document my progress, along with a small blog to share the progress and results with my peers.
GatorGuesser
3rd Place Overall | SwampHacks IX | January 2023
During the 36 hour hackathon SwampHacks IX, my team created our own take on the popular game GeoGuessr, in which the player is dropped into a random place on Google Street View and has to guess where they are located in the world. In our version, the player must guess their location on the UF campus from pictures that we collected during the hackathon. The game is programmed entirely in C++, with the front end utilizing SFML for graphics. In order to include nearly 150 images in our game, we had to come up with a scalable solution for coordinating location data with each image. To do so, we enabled location services on our smartphone cameras when taking the pictures, and later extracted the longitude and latitude from the metadata in bulk. Out of the 90 teams that submitted projects, we were excited to win 3rd place overall.
Graph Algorithm Visualizer
November 2021 - December 2021
Last year, two teammates and I worked several weeks to submit this c++ desktop application (with SFML) as part of our data structures and algorithms course. Our app allows the user to randomly generate mazes and visualize different graph algorithms to find the shortest path through the maze. In addition to the visual examples, we also included a large scale timing comparison in which the user can create an arbitrarily large maze and compare stats about each algorithm's performance. My contribution to the project included but was not limited to handling the random maze generation, designing the data structure which relates a graph to a 2d maze, implementing Depth first search, and researching and implementing A* search. This project was a delight to work on with team members who were focused on keeping clean code and adhering to our own agreed progress deadlines.