A.E.G.I.S.-GRID: Energy Monitoring & Homelab Resilience

Python ESP32 Home Assistant Web Scraping Modbus

AEGIS_GRID: ACTIVE // SYSTEM_STATUS: STABLE

Ecuador suffers from recurring grid blackouts and failures in the national power system. Achieving technological resilience requires proactive monitoring and automated responses. A.E.G.I.S.-GRID is the orchestration layer that ensures the survival of our On-Premise infrastructure.

Components and Workflow

1. Scraping and National Grid OSINT

A dedicated scraper agent continuously monitors public APIs and announcements published by local power distribution companies to predict scheduled blackouts in our specific sectors (Pallatanga / local nodes).

2. Solar Telemetry and Field Integration

The system reads the State of Charge (SoC) from the LiFePO4 batteries (Pylontech) and the production of our Must/Growatt hybrid inverter in real-time via Modbus/RS485 using ESP32 chips. It also integrates our Pallatanga weather station to predict sudden solar radiation drops caused by heavy cloud cover or rain.

3. Intelligent Island-Mode Switching

When the scraper detects an imminent blackout or sensors log a voltage drop on the public grid:

  • Load Shedding: Selective, automated shutdown of non-critical LXC containers and VMs in Proxmox (local-pve).
  • Island Mode: Physical electrical isolation, routing critical consumption directly to the off-grid solar battery bank.
  • Watchdog: Active temperature and current spike monitoring to prevent thermal or physical damage to the hardware.

Real-World Benefits

The homelab has not experienced a single forced blackout or data loss since the system was deployed. All automation and decision-making run entirely On-Premise without external third-party cloud dependencies.

TUXBOT@SYSTEM:~$ ./chat
> SYSTEM INITIALIZED. FLEET STATUS: ACTIVE.
🐧🤖 [Tuxbot]: Hello, I am Tuxbot. Your Ghost in the Shell for dragont.ec. What do you want to query today?
>