Blaze Client

Blaze Client is a Minecraft modpack designed to make it easier to play the game without it being considered cheats. It contains mostly visual mods such as armor HUD, which constantly displays your armor's durability on your screen and TNT counter which renders a countdown on top of TNT, making it easier to time jumps. It also contains a few non-visual mods as well such as toggle spring and toggle sneak, making it easier to play certain minigames in Minecraft servers along with controller support allowing players to fully play the game with compatible PlayStation, XBox, or like controllers if they choose to do so.

Blaze Client is also complete with a cosmetic system and API allowing players to equip capes, helmets, and other cosmetics that aren't usually in the game to be seen by other players playing with the modpack.

QuickCheck

QuickCheck was a product I developed freshman year as an entry in a competition where students of my High School competed to develop a new attendance system for the school. Along with a monetary prize I also won the opportunity to test it with teachers and classes in the school.

QuickCheck was a web app that utilized a WaveID RFID reader to read students ID cards to retrieve their student IDs then it used the student IDs to retrieve their picture, name, grade, and current teacher from a database. Then it would utilize the Google Sheets API and Google App Scripts to put the attendance info (sorted and color coded) into separate spreadsheets shared with teachers and the district.

FTC #6200

FTC team #6200 was my FTC team for the majority of highschool. We ended up winning 1st place Control Award (best code award) twice and made it to state.

My role freshman year on FTC #6200 was programmer and after that I became the lead programmer on team #6198 and later rejoined team #6200. We made significant improvements to our robot over the years. For instance, my freshman year we utilized solely the encoders in the drive wheels to track our position and my junior year we utilized both deadwheels and a camera to both track our position more accurately and to confirm it. Computer vision use also improved over the years evolving to the aformentioned position confirmation with april tags and recognizing the game pieces from just using it to recognize an april tag ID at the beginning of the match.

FRC #8029

FRC team #8029 was my FRC team for the majority of highschool. I started as a programmer as a sophomore and became the lead programmer junior year mentoring 4 other programmers on the team going on to be 2025 Midwest Champions.

My sophomore year we mainly utilized a gyroscope and swerve encoder odometry to track the robot's position, but my junior year we added a camera to confirm the position of the robot using AprilTags, as well as ultrasonic sensors to accurately line the robot up with the game elements autonomously. For both years, our drivetrain consisted of 4 swerve modules, each with a Neo Spark MAX along with a CTRE CANCoder tracking and controlling each wheel's rotation and a CTRE TalonFX to spin the wheels to drive the robot.

Blinker Proxy

Blinker proxy is a unique internet privacy tool that allows users to set up a proxy "server" on any device on any network without the need of port forwarding.

Blinker proxy utilizes a central proxy routing server. Via a GUI in a desktop application, the user is free to connect to the routing server and connect to a proxy server they setup. After that, all internet traffic from the client is directed to the routing server, to the proxy server the client selected, and then out to the target host on the internet. The data the target host sends back, is sent back the same way.

Blinker proxy was designed so that users could set up a proxy on a network with an ISP that either disallows port forwarding completely or uses Network Access Translation (NAT). This project was never distributed or made public as my goal for it was to learn more about how networking works and due the fact that it could be used for unintended purposes.

Garden Protector

Garden protector is a project I plan to complete the summer between my Junior and Senior years of highschool. It's intended to scare away animals that try to eat food from our garden with either a laser or water stream that won't harm the animals.

I intend to utilize the Nvidia Jetson Nano's superior AI vision capabilities compared to it's competitors to recognize animals and their locations in the camera's viewport. I then intent to convert this information into 2 servo positions (one for yaw one for pitch) and to send it to an arduino to aim the servos at the target. The reasoning for the arduino as an intermediate between the Jetson and the servos is due to the Jetson's inability to produce a usable PWM signal.

3D Rendering Engine

While there's a saying that says "don't reinvent the wheel" the goal of this project is to do just that. I wanted to make the most optimized ray tracing rendering engine I could while not sacrificing aesthetics.

Developed in C++ using only the SDL2 library, this rendering engine offloads significant processing to a server. Unlike most rendering engines—such as Minecraft's—which maintain numerous invisible, unrendered objects in the client’s scene, slowing down each frame, this engine optimizes performance by ensuring the client only processes objects that are currently visible or likely to become visible in upcoming frames. After objects sure to be invisible are filtered from the client, the client then casts rays from each pixel of the viewport checking if it intersects the objects that haven't been filtered out and might be visible.

In the near future, I plan to integrate this engine with Virtual Reality using a Raspbery Pi.

Outlet Monitor

The outlet monitor was designed to measure the voltage and current being used by a singular or multiple devices. This information, along with the watts calculated using it, is then plotted on a web dashboard allowing the viewer to compare it with other outlet monitors and to itself over time. This allows you to see how much power certain devices are using and how much it costs to keep them on.

It measures voltage using a ZMPT101B AC voltage module and current with a ACS712 AC and DC current module. These modules are connected to a WiFi-enabled Arduino R4, allowing it to communicate and send this data with a centralized server. It was designed with ease of assembly prioritized over elegance and compactness, which I hope to change if I come back to this project in the future.

Autonomous Drone

While not autonomous yet, this large, fully 3D-printed drone, is equipped with 4 10" propellors, 4 920kv brushless motors, 4 30A ESCs, an Arduino, a gyroscope, and a 2250 mAH battery. It also features a removable top that slides off using a set of rails to access the internal components and change the battery.

The reason for using an Arduino instead of a preprogrammed flight controller was the freedom of being able to control the drone without the controller. Most flight controllers are designed to only control the drone from the controller and using an Arduino gives me the freedom of being able to use computer vision and control the drone autonomously in the future.

NASA ADC

NASA App Development Challenge is a team coding challenge. The year I participated, we were tasked with developing an application to visualize the Artemis II mission flight path as well as the path's communication with satellites.

Our solution was a Unity game programmed in C# that consisted of 2 paths from the ship to the investigation site. One was the shortest path that was relatively smooth, the other was shorter and allowed the path to be a little bit steeper. Our project also contained 6 different views to choose between, each showing a different bit of information about the area of the moon investigated. It also contained a minimap to see these from a birds-eye view and the player's location. To make these views, we used the slope and height data spreadsheets NASA provided for us. We processed this data in C++ to make a rough black and white texture for each view, and then later processed it and turned it into color with Python.