Orvex
Orvex is an AI-assisted inspection copilot for solar operations and maintenance teams. It helps turn photovoltaic panel images and bounded video frame evidence into structured, human-reviewable triage results. The system analyzes visual inputs, identifies possible defects such as hotspots, soiling, shadowing, cracking, vegetation, diode issues, or offline modules, then returns prioritized findings with confidence, risk score, visual evidence, and recommended next actions. The product is designed as a decision-support workflow, not an autonomous diagnostic system. Every result is validated against a strict schema, stored in an inspection job, and routed for human review. Orvex combines a FastAPI backend, a Next.js review interface, supervised solar anomaly baselines, local VLM inference, and an experimental worker queue for bounded video processing. Its goal is to make solar inspection review faster, more traceable, and easier to standardize.
Category tags:
For Innovators & Creators