Dry Etcher Sensor Data with Sync Ratio Anomaly Detection Method

A case study presents sensor data from a semiconductor dry etcher to demonstrate a novel anomaly detection method called the Sync Ratio. The method groups sensor data based on the synchronization of rare data points in time to reduce false alarms. The dataset and accompanying code were published by Mahya Qorbani in April 2026.

Use Cases

• Evaluating the Sync Ratio method's performance against benchmark algorithms for anomaly detection on sensor time-series data.
• Analyzing the synchronization patterns of rare data points across multiple sensors from a dry etcher.
• Training machine learning models to distinguish true alarms from false alarms using the proposed sensor grouping features.

Strengths

• Includes full implementation code in an IPYNB notebook for method replication.
• Dataset is shared under a permissive CC-BY-4.0 license allowing reuse and modification.

Limitations

• The dataset scale is 42.1 MB, indicating a relatively small sample size for training complex models.
• The authors note limitations in handling temporal delays and highly controlled sensors.
• Specific row counts and column details for the sensor data are not provided.

Provenance

Source

figshare, authored by Mahya Qorbani.

Collection Method

Sensor data collected from a semiconductor dry etcher during manufacturing.

Time Range

null

Freshness

Last updated in April 2026.

Geography

null