UROP

Student researcher contributing to a faculty-led research project exploring federated learning on distributed edge devices using NVIDIA Jetson platforms.

Worked on distributed computing experiments involving local model training pipelines and multi-device edge AI systems.

Supported experimentation in edge-device communication, deployment workflows, and distributed system integration.

What I Did

• Set up and networked three physical Jetson Orin Nanos running decentralized federated learning (FedAvg with OCD-FL peer selection) in Docker, controlled over SSH
• Extended the training script with non-IID data splits and a local-only baseline

Worked On

• NVIDIA Jetson platforms
• Federated learning systems
• Distributed edge computing
• Local model training pipelines
• Embedded AI deployment
• Linux-based environments
• System integration
• Experimentation workflows