n8n Smart Meter Anomaly Detection Workflow (HTTP API Alerts)

Detect smart meter anomalies and trigger HTTP API alerts automatically

This n8n workflow pulls recent smart meter readings on an hourly (or manual) run, detects abnormal consumption patterns like spikes, drops, leakage-like continuous off-hours flow, and “stuck” meters, then prepares structured alert payloads for downstream dispatch—sending critical alerts through a cooldown step and standard anomalies to logging.

What this workflow does

• Schedules hourly (cron) or supports a manual test run.
• Uses configurable inputs for the metering API URL, meter group ID, lookback window, and anomaly detection thresholds.
• Fetches readings from your smart meter API and waits briefly for aggregation.
• Computes per-meter baselines using mean and standard deviation, then evaluates anomalies such as:
  • Spikes and drops via thresholded z-score checks
  • Off-hours continuous flow (potential leakage pattern)
  • Low-variance “stuck” patterns indicating abnormal meter behavior
• Assigns an anomaly score, filters out meters with no detected issue, and outputs a no-action note when appropriate.
• Classifies severity and category, then builds a structured alert payload.
• Routes critical anomalies through a cooldown wait and marks them ready for critical dispatch; non-critical anomalies are marked ready for standard logging.

Use cases

• Spotting sudden consumption spikes or drops across a meter group for utility or facility operations.
• Detecting potential leakage through continuous off-hours usage patterns.
• Identifying “stuck” meters with low-variance readings to prioritize maintenance.
• Automating anomaly alerting from an HTTP-based metering API without manual review.

Technical details

• Auth: configure an HTTP Request node with an HTTP Header Auth credential to your metering API.
• Nodes/logic: if, set, code, wait, sticky note, http request.
• Core detection tuning: z-score limits, off-hours window, minimum flow, variance threshold, and critical score cutoff—set in the workflow configuration.