Welcome to Team-Arbitrary's Arbit Project, an AI-powered system for transformer thermal image management and analysis.
Our goal is to improve power system reliability by detecting anomalies and monitoring transformer health through thermal imaging and machine learning.
Spring Boot–based backend that provides APIs for:
- Managing transformer thermal images
- Running AI-powered anomaly detection
- Delivering transformer health scores and reports
React-based frontend for:
- Uploading and visualizing thermal images
- Displaying anomaly detection results
- Monitoring transformer health dashboards
FastAPI-based AI codebase as the API for:
- ML analysis: an autoencoder-based anomaly detector (red boxes)
- Thermal analysis: hotspot detection using the red channel (yellow boxes)
- Backend: Java 21, Spring Boot, PostgreSQL, Maven
- Frontend: React, Node.js
- AI Models: Custom PyTorch Deep learning models for anomaly detection
Arbit is designed to support smart grid monitoring by combining AI with intuitive web tools, making transformer maintenance more efficient and reliable.
