KiwisIoT Live Anomaly Detection
Collect, train, and detect live sensor anomalies
Idle
Workflow
-
Select channels Pick one or more sensor channels to monitor
-
Collect baseline Gather the required baseline readings (200)
-
Build model Train an ML model (Isolation Forest) that learns each channel's normal range
-
Monitor live Values outside the model's learned range are flagged as anomalies
Collecting baseline readings… Latest value: —
Total readings
0
Normal
0
Anomalies
0
Anomaly rate
0.0%
Live sensor stream
Normal readings
Anomaly markers
Anomaly score history
Distance from normal range (negative = outside min–max)