📖 About This Blog

Blogging my way through a TinyML Swarm Intelligence journey. I’m exploring how resource-constrained embedded systems (like ESP32 microcontrollers) can coordinate through swarm intelligence algorithms to perform distributed machine learning tasks.

Background

I hold a Master’s in Information Systems with a thesis on machine learning for educational analytics. My technical interests span:

  • Embedded systems (Arduino, Raspberry Pi, ESP32)
  • Vintage computing (6502 assembly on VIC-20)
  • Mesh networking (Meshtastic)
  • Linux systems programming
  • Search algorithms and search-based agents
  • Databases and data science
  • Web-based development
  • Electronics, solar power, and the maker/repair community

This blog bridges my existing skills with cutting-edge research in edge AI and bio-inspired computing.

🎯 Research Focus

Core Question: How can we build robust, scalable machine learning systems using swarms of resource-constrained devices?

Key Areas:

  • TinyML model deployment on microcontrollers
  • Federated learning for distributed systems
  • Bio-inspired swarm algorithms (ant colonies, bee democracy, flocking)
  • Consensus mechanisms in mesh networks
  • Energy-efficient edge computing

Target Application: Wildlife monitoring using autonomous sensor networks

📚 Learning Path

Phase 1: Foundations (Months 1-3)

  • Mathematical foundations (linear algebra, calculus, optimization)
  • Swarm intelligence theory and biology
  • TinyML fundamentals (Harvard certification)
  • Distributed systems concepts

Phase 2: Implementation (Months 4-6)

  • ESP32 mesh networking
  • Audio classification with swarm consensus
  • Federated learning prototypes
  • Real-world deployment testing

Phase 3: Advanced Topics (Months 7-9)

  • Convergence analysis and proofs
  • Byzantine fault tolerance
  • Energy optimization strategies
  • Graph theory for network topology

Phase 4: Integration (Months 10-12)

  • Complete system integration
  • Technical report writing
  • Research proposal development

Warning: Dates/timing may change without warning

📝 Blog Content

Post Categories:

  • Weekly Updates: Progress reports and reflections
  • Technical Deep Dives: Implementation details and code walkthroughs
  • Math Explorations: Derivations, proofs, and intuition-building
  • Paper Summaries: Key research paper breakdowns
  • Project Logs: Hardware builds and experiments

🔨 Projects

All project code is open source and available in my main research repository.

Planned Projects:

  • Boids Flocking Simulation (Python)
  • Ant Colony Optimization for TSP
  • Audio Event Detection on ESP32
  • 4-Node ESP32 Mesh Network
  • Distributed Audio Classification Swarm
  • Federated Learning Implementation
  • Wildlife Monitoring Prototype
  • Complete Integration System

🛠️ Tech Stack

Hardware:

  • ESP32 development boards
  • Various sensors (audio, environmental)
  • Raspberry Pi (for edge gateway testing)

Software:

  • Python (NumPy, TensorFlow, PyTorch)
  • C/C++ (Arduino framework, ESP-IDF)
  • TensorFlow Lite for Microcontrollers
  • Edge Impulse platform

Tools:

  • Jekyll + GitHub Pages (this blog)
  • Zotero (paper management)
  • reMarkable (handwritten notes)
  • Google Sheets (progress tracking)

📊 Progress Tracking

Week-by-week updates: Main Research Repository

🎓 Target Focus

Research Areas:

  • Embedded ML / TinyML
  • Distributed Systems
  • Swarm Intelligence / Multi-Agent Systems
  • Edge Computing

📬 Contact

Interested in collaborating, have questions about the projects, or want to discuss swarm intelligence?

  • GitHub Issues: Best for technical questions about code
  • Email: steve[at]employinginnovation.com
  • Edge Foundation Discord: S. Harris –> @sjarn

📄 License

  • Blog content: CC BY 4.0
  • Code: MIT License (see individual project repositories)

🙏 Acknowledgments

Special thanks to:

  • Edge Foundation community for resources and support
  • Harvard’s TinyML team for excellent course materials
  • 3Blue1Brown for making math visual and intuitive
  • The open-source embedded systems community

Last Updated: October 24, 2025

This is a living document. The research direction may evolve as I learn and explore.