About Me

I’m Liam Carlin, an engineer driven by curiosity and a love for innovation. I thrive on tackling complex challenges, turning ideas into impactful solutions, and pushing the boundaries of what’s possible. Whether designing systems, building robots, or modeling dynamic behaviors, I bring energy and creativity to everything I do. Let’s create something extraordinary together!

One of my lifelong dreams has always been to feel what a bird feels and be able to soar the skies. There were not many options available that would allow this until I came across a sport called "FPV Drone Racing". The FPV stands for "First Person View", meaning that the pilot of the drone wears a set of goggles that immerse them inside the drone while it's flying. I was awed by this and realized my dream again. I set out to completely design the frame of the drone from scratch in Onshape (CAD platform). This took many attempts and never seemed to reach my level of expectations, I would 3D print it and then it would not fit together or I would stress test it and it would be weak. Though after what felt like years of designing I finally came across a design that worked. I called it MK3 (after the Iron Man Suits). As soon as I finished the files were sent off to a CNC cutting mill to be made out of carbon fiber. Upon its arrival, everything fit and I soldered up and finished the code for the rest of the drone, and finally I took to the skies accomplishing my dreams of flight.

Over the course of the past three years, I have been rigorously working on a self-stabilizing rocket. The purpose of this project was to see if a solid fuel-propelled rocket was able to stay upright without the use of fins on the model rocket scale. During the brainstorming phase there were many possible solutions to this problem though upon watching a SpaceX livestream I figured it out, Thrust Vector Control (TVC). TVC is where the mount for the motor gimbals in order to keep the rocket on a path. Despite this being used in pretty much all large-scale rockets, not many people had tried to alter it for the model scale using Onshape (CAD platform). Currently, the main hardware for the rocket is done including the entire gimbaling assembly, though it is still awaiting its first launch. This is due to some minor software problems including data logging and accurate position changes.

The Digital Poker Assistant (DPA) is a cutting-edge app designed to enhance the social and strategic elements of in-person poker. It features AI-driven chip counting from photos, tracks individual and group stats for friendly competition, offers a virtual chip stack for games without physical chips, and provides card detection with winning probability calculations. DPA seamlessly blends technology with traditional poker, enriching the experience while preserving its social essence.

The Barbell Clip is an innovative solution designed to tackle a common and potentially hazardous problem in CrossFit and other high-intensity workouts. Traditional clips often fail to keep weights securely in place after the barbell is dropped, as the impact causes plates to shift and push against the clip, resulting in sliding weights and an off-balance bar. This new barbell clip is engineered to withstand these forces, ensuring plates stay firmly in position even after repeated drops. Once the design and patent are finalized, the clip will be injection molded for durability and performance, enhancing workout safety and stability while protecting both athletes and their surroundings.

The Baja Wheel Hub is a robust and lightweight component designed for off-road racing vehicles. Engineered to withstand the harsh conditions of Baja racing, this wheel hub ensures durability and performance. The design incorporates advanced materials and manufacturing techniques to provide a balance of strength and weight reduction, essential for competitive racing. The hub's innovative design also allows for easy maintenance and quick replacements, minimizing downtime during races.

The stock analysis employs two predictive methods to gain insights into price movements. The first approach uses candlestick pattern classification with principal component analysis (PCA), achieving 27% accuracy, significantly outperforming random guessing. Despite limitations like grayscale conversion and image resizing, it effectively identifies bullish and bearish trends . The second method applies the Geometric Brownian Motion (GBM) model to Amazon stock data, achieving ~95% directional accuracy by balancing long-term trends (drift) with short-term volatility. While limited by constant volatility assumptions, the GBM model demonstrates strong predictive power, offering valuable forecasts for investment planning. Together, these methods provide actionable insights into market behavior.

Cyclone, a 3-pound featherweight-class BattleBot competing in the NATIONAL HAVOC ROBOT LEAGUE (NHRL), is a formidable contender. Its offensive strategy centers on a high-performance vertical spinner powered by a brushless motor, designed to incapacitate opponents with precision and power. Complementing its aggressive approach, Cyclone’s robust defense features strategically integrated titanium, providing exceptional resistance to heavy impacts and ensuring durability in intense battles.

With the creation of many of my projects comes sections where it is necessary to use software in order to run the hardware. This drove me to start to learn different coding languages to improve the outcomes of my projects. The coding languages that I have learned are C++, CSS, Java, JavaScript, HTML, and Python. In particular, the code that is shown below is written in a variation of C++. This program is for my self stabilizing rocket and runs the majority of the calculations done to filter out vibrations picked up by the gyroscope. As well as converting these filtered values into values that can be transmuted to the servos to "steer" the rocket.