LoopGuard

LoopGuard is a web + desktop app for industrial monitoring & control teams that continuously checks the health of feedback loops (PID and similar) using existing historian/SCADA data. It flags sensor drift, valve stiction, oscillations, saturation, and tuning degradation, then recommends concrete actions (e.g., “check transmitter calibration,” “inspect valve positioner,” “reduce integral gain”). The MVP focuses on being a practical “early warning system” rather than a full control platform replacement. It connects read-only to common data sources, learns normal loop behavior per tag, and produces a prioritized daily list of loops that are costing throughput, energy, or stability. Reports are designed for maintenance and controls engineers to act on quickly, with evidence plots and a simple ROI estimate. This is a combination traditional + AI app: traditional analytics for known failure modes plus ML anomaly detection for drift and regime changes.