vLock AI Vision

Industrial machines rely on protective devices such as light curtains, safety interlocks, and presence sensors. However, these devices do not always provide the full picture. A person may remain inside a restricted area after passing through a light curtain, or a sensor may not capture everything happening around a machine. vLock-AI Vision adds intelligent visual monitoring alongside existing safeguards, giving industrial control systems greater awareness of what is happening around equipment. Built for SCADA (Supervisory Control and Data Acquisition) developers, who build software used to monitor and control industrial equipment, vLock makes AI vision easier to configure and integrate, reducing the need for specialist computer vision development or custom model training. Developers describe what they want to monitor in plain English. Meta Muse Spark assists with configuration, a human reviews and approves the setup, and SAM 3.1 segments camera images to identify objects and their positions. Deterministic rules evaluate the observations and publish three clear states: Active, Clear, or Unknown. These states integrate with existing SCADA systems through OPC UA, MQTT, or REST, enabling supervisory workflows without building a custom vision pipeline for every machine. The result is additional scene awareness, faster integration, and a reusable AI vision workflow for industrial automation. vLock AI Vision provides supplementary supervisory monitoring. It is not a safety-rated system and does not replace certified protective devices, safe restart procedures, or lockout/tagout. vLock AI Vision tells SCADA what it sees. SCADA decides what the machine does. (Development Note: vLock AI Vision and the Ignition 3D Printer Farm demo were developed during the Meta Global AI Developer Hackathon 2026. My earlier vLock Digital LOTO project is separate and is not part of this submission)