This project page is currently a high-level overview. Full technical documentation coming soon.

Overview

LUMNET is a wireless mesh LED display system developed as a third-year group consultancy project with ARM. The system demonstrates how mesh networking can solve connectivity challenges in smart environments, creating a self-healing network of synchronized LED panels with integrated environmental sensors.

Working in a team of six, we delivered a complete solution spanning hardware design, embedded firmware, mesh networking protocols, and a React-based web interface for real-time control.

The System

Hardware:

  • ESP32-C6 nodes with Wi-Fi 6 and IEEE 802.15.4 radio
  • 16×16 addressable WS2812B LED matrices
  • Sensirion SHT41 temperature and humidity sensors (±0.2°C accuracy)
  • Custom 3D-printed PLA enclosures
  • Dual-ESP32 gateway architecture for independent RF paths

Software Architecture:

  • PainlessMesh protocol for self-healing mesh networking
  • React frontend with WebSocket real-time communication
  • Custom animation sketchpad with Base64 PNG encoding
  • JSON-based animation transmission segmented into 1000-byte chunks
  • Real-time sensor monitoring with 3-second polling intervals

Key Features:

  • Synchronized animations across all mesh nodes
  • Live environmental data visualization
  • Custom animation creation and transmission
  • Self-healing network topology
  • Independent operation from home Wi-Fi networks

Technical Highlights

Mesh Network Design: The system implements a true mesh topology where each node can relay data, eliminating single points of failure. When a node disconnects, the network automatically re-routes traffic through alternative paths. A dedicated gateway node with dual ESP32s provides the internet bridge whilst maintaining mesh integrity.

Animation Pipeline: Animations created on the web sketchpad are converted to PNG sequences, Base64 encoded, packaged as JSON with frame rate metadata, transmitted to the border router via HTTP GET, segmented for mesh transmission, and reconstructed at each node for synchronized playback.

Energy Efficiency: Wi-Fi 6 on ESP32-C6 provides low-latency, energy-efficient connectivity. Future firmware updates will enable deep sleep whilst maintaining active Wi-Fi connections, reducing power consumption to hundreds of microwatts.

My Contributions

As part of the hardware team, I focused on:

  • Mesh network protocol evaluation and implementation
  • ESP32 firmware development for mesh communication
  • Sensor integration and data handling
  • Gateway architecture design with dual-ESP32 UART communication
  • Animation decoding and LED matrix control
  • Real-time command processing and node synchronization

Applications

Beyond synchronized LED displays, the architecture demonstrates principles applicable to:

  • Smart lighting control in commercial and residential spaces
  • Environmental monitoring networks
  • Distributed IoT sensor systems
  • Security and presence detection
  • Building automation and energy management

What I Learned

This project strengthened my understanding of:

  • Distributed systems and mesh networking protocols
  • Real-time embedded systems with tight latency constraints
  • Hardware-software co-design and systems integration
  • Client stakeholder management and iterative development
  • Team collaboration on complex technical projects
  • Trade-offs between different wireless protocols (Wi-Fi vs. Zigbee vs. Thread)

The shift from initially planned Zigbee to PainlessMesh taught valuable lessons about adapting to technical constraints whilst maintaining project timelines and client requirements.

Team Members: Aikaterini Dionysiou, Marta Lopez Gallo, Robin A Masih, Sakhawat S Salam, Nithi Sendhil, Janvid Wu
Client: ARM (Yiangos Yiangou)
Supervisor: Yunjie Gu
Technologies: C++, ESP32-C6, PainlessMesh, React, WebSockets, I²C, UART
Hardware Repo: github.com/pb1n/LEDMeshProject
Software Repo: github.com/Shakofalltrades/ARM_LED_APP
Status: Placeholder - full technical writeup in progress
Date: December 2025