SensorSleuth

SensorSleuth is a web app (with optional mobile companion) for anomaly detection in industrial sensor streams—vibration, temperature, current draw, pressure, and PLC tags. It connects to common plant data sources (OPC UA, MQTT brokers, and historians), learns a baseline per asset, and flags deviations with clear, operator-friendly explanations (what changed, when, and which sensors drove the alert). The MVP focuses on fast onboarding: a guided connector setup, automatic data quality checks, and out-of-the-box models that work with limited labels. Alerts route to email/SMS/Teams and include a “next best checks” checklist to reduce alert fatigue. This is an AI + traditional monitoring app: AI for detection and ranking, traditional rules for safety thresholds and hard limits. It’s realistic for small-to-mid factories that can’t afford a full reliability team or a heavyweight enterprise suite.