Open-source energy monitoring · Dissertation project
A local-first platform for acquiring, processing and visualizing industrial energy data.
ENERVIGIL was developed to provide an accessible alternative to proprietary and cloud-dependent energy monitoring platforms. It integrates industrial meters, stores measurements locally and exposes the data through a secure web interface and MQTT.
Small and medium-sized industrial installations often need practical energy visibility, but commercial platforms can involve high licensing costs, proprietary ecosystems, external cloud infrastructure or difficult integration with existing industrial meters.
ENERVIGIL addresses this by focusing on local deployment, manufacturer independence and extensibility. The system is designed to run on low-cost edge hardware while keeping configuration and measurement data under the organization’s control.
Each meter is represented as a device, and each measured or calculated variable is represented as a node. Communication details are isolated inside protocol adapters, allowing the rest of the application to process values using a uniform internal model.
Real-time energy monitoring from industrial measurement devices
Single-phase and three-phase device modelling
Configurable device and node setup through a web interface
Historical time-series storage using InfluxDB
Local-first deployment with Docker Compose
MQTT publishing for integration with external systems
Secure access with HTTPS, JWT sessions and brute-force protection
Extensible architecture for future protocols and processors
FastAPI service responsible for acquisition cycles, device management, node processing, REST endpoints and authentication.
SvelteKit interface for configuring devices, viewing current values, managing variables and analyzing historical charts.
SQLite stores configuration and user data, while InfluxDB stores timestamped measurements for efficient historical queries.
Nginx provides HTTPS access and Mosquitto publishes device and node state through MQTT for external consumers.
The validation platform reproduced a representative energy monitoring setup using two different communication paths. The results showed stable acquisition, consistent historical logging and low computational resource usage.
The UI lets the administrator configure devices and nodes, inspect live values organized by electrical phase, review alarm states and analyze historical measurements with automatic temporal aggregation.
The project already provides the basic monitoring workflow, but it is still evolving. Future work is focused on improving scalability, data management and operational features.