Pertamina PPE Compliance Monitoring

Industrial facilities demand strict personal protective equipment compliance, yet manually watching dozens of CCTV feeds for missing helmets and vests is both impractical and error-prone. During a Telkom Indonesia internship I built a system that automates exactly that: it ingests Pertamina's existing CCTV/NVR feeds, runs a YOLOv11-based PPE detector, and surfaces violations in real time on a Next.js dashboard — and I owned the full lifecycle in-house, from dataset curation and annotation through training, deployment, and monitoring.

The hardest part was never the model but the integration with the existing NVR network. I normalized heterogeneous RTSP streams, built reconnection logic for flaky cameras, and decoupled ingestion from inference so a single slow feed could never stall the whole pipeline, packaging everything into Docker containers for a clean operations handoff. The result shipped to production at Pertamina Patra Niaga, where it now provides continuous, real-time compliance monitoring across the facility's camera network.

That experience drove home that production computer vision is roughly 80% data and infrastructure engineering. Robust stream handling and an operations-friendly deployment story — not a marginally better mAP — were what separated a polished demo from a system the plant could actually rely on day after day.