Home Made ROV

Self-Balancing Robot

Facing an employment shortfall during the summer of 2020, thanks to the economic fallout of COVID-19, I decided to take up a self-guided project to grow and further my engineering abilities. To this purpose I designed a self-balancing mobile robotics platform, as seen below. Using an Arduino Mega as its core processing unit, the platform can operate in a simple autonomous mode or can be driven manually using a 2.4GHz RC transmitter. Currently, the Arduino drives the entire system through an extensive series of interrupts and I2C routines, but future iterations of the project will included utilization of the Raspberry Pi Zero W (seen towards the front of the vehicle).

March Madness Algorithm

For the 2019 NCAA March Madness Tournaments, I developed an algorithm using C++ to predict the outcome of individuals games and then generate a tournament bracket using these outcomes. By the time the tournament came to a close, my algorithm had placed in the top 10% of brackets nationally and won my bracket pool!

Over the years the algorithm has undergone several iterations and the current size of the project is roughly 3000 lines. The algorithm includes a number of its own structs/classes and utilizes a custom hash bucket structure to sort teams and game results.

Looking to 2021, I am hopeful the tournament will resume and further evaluation of the program will be possible.

2021 Update: Ahead of the 2021 NCAA Tournament I spent a significant amount of time refining the algorithm to find its optimal fixed-weighting performance. By the time the 2021 Tournament had come to a close, this model landed in the top 2% of all ESPN brackets with a score of 1320/1920 points. Similarly, another upset-oriented bracket generated by the algorithm, which I entered into my bracket pool, finished in the top 4% with a score of 1220/1920 points.

COMAP Mathematical modelling Contest

During the spring semester of my freshman year, 2018, I teamed up with two friends to compete in the COMAP International Mathematical Modelling Contest. The contest, which lasts for exactly one-hundred grueling hours, challenges students to develop different models for problems. In our particular case, we chose to model the distribution of medical supplies using a drone delivery system in Puerto Rico in the aftermath of Hurricane Maria.

Our model consisted of two main elements:

• An algorithm for packing shipping containers with drones and medical supply packages of varying sizes

• An model that weighted regions of interest for damage surveillance and medical package distribution

The images below show two integral portions of our submission, which earned Honorable Mention. The plot to the left shows a sample packing arrangement for a 40 inch, cubic shipping container. The screengrab to the right is comprised of our binned-pixel weighting system overlaid on Google Earth to show the accuracy of our binning algorithm.

sPARKFUN AVC

During the Fall of 2017, I designed and manufactured my own custom beetle-weight (3 lb.) combat robot to compete in a tournament put on by Sparkfun Electronics. After the dust had settled, my combat bot, Pistachio, finished with a record of 1-2. Below is a video of my first match up against Oreo Typhoon! (Pistachio is the one with green wheels...and also the one that loses.)

DIY Quadcopter

In the early days of my high school tenure (2014-2015) I began building a quad-copter from scratch. Over the course of the next several years I worked to get the craft airborne. Eventually, however, the vehicle took a crash from which it could not recover and I scrapped it for parts to begin building a series of other projects.