During my time at Gannon University, I worked with a faculty advisor on an artificial intelligence research project called AI-Sandbox. The project simulated evolving agents in a 3D environment to study reinforcement learning and behavior adaptation under environmental constraints. Agents were divided into two groups—seekers and runners—and competed in a tag-based survival scenario. Survivors propagated their neural network parameters and genomes to the next generation, allowing us to analyze behavioral evolution over time. The goal of the project was to study how learning strategies and behaviors emerge through reinforcement learning and neuroevolution.
The research results were accepted for publication and presentation at an engineering education conference, where I presented the system design, experimental results, and lessons learned from developing large-scale AI simulations. I was later invited to present the project at a second research conference, where I discussed the implementation details, training methodology, and observed behavioral patterns in the agents. These presentations provided experience communicating complex technical systems to academic and professional audiences.
Technical Implementation
- Developed simulation environment using Unreal Engine and C++.
- Implemented reinforcement learning and neuroevolution pipelines.
- Simulated multi-agent environments with procedural world generation.
- Analyzed agent behavior across generations to evaluate learning patterns.
- TypeResearch
- ContentFull-Stack Development, C++, Unreal Engine, Cuda, CMake, Blueprints, Artificial Intelligence, LaTeX
- LinkAI Entity, World Generator, ASEE Research Website, Sigma XI Conference, Tag AI Sandbox, LaTeX Code of Research
