Anomaly-Based Strategies

Anomaly-based strategies create work orders when device data indicates an abnormal condition — a sudden spike in errors, unusual readings, or threshold violations. This type enables predictive maintenance: addressing problems before they cause downtime.

When to use anomaly-based strategies

When early warning signs are visible in machine data before a device fails:

• Temperature or vibration spikes
• Rising error rates
• Performance dropping below a critical value

Example scenarios:

• Trigger maintenance when the error rate exceeds normal levels
• Schedule inspection after detecting unusual vibration patterns
• Create a work order when device performance drops below a threshold

Configuration

Define the trigger using:

• Metric — Which metric to monitor
• Threshold — The abnormal value that should trigger a work order
• Window size — The time period over which the condition is evaluated

Example: “If the temperature at Motor A exceeds 85 °C twice within 60 minutes → create a work order.”

Note

Anomaly-based strategies require an active machine data integration. Without a data connection, anomalies cannot be detected. See Machine Data.

Practical tips

• Choose the threshold and window so that brief, harmless spikes don’t immediately trigger an order.
• Use the strategy’s assignee so critical anomalies notify the right person immediately.

Related

• Strategy Types overview
• Scheduled Strategies
• Usage-Based Strategies