Anomaly launch with Anomaly Detect Type - Launcher

When Batch Tracking receives a detection signal that is associated with an Anomaly Detect configured as a Launcher, an anomaly will be inserted automatically in the pipeline.

A unique name is generated for the automatically created anomaly using the format: {prefix}_{number}, where the prefix is “ANOMALY” (for example, ANOMALY_123). If the Anomaly Type is configured to have a different prefix, the configured prefix will be used in the name (for example, SCRAPER_123). The other attributes of the anomaly are set to the default values configured for the corresponding Anomaly Type.

To determine where to insert the anomaly, all segments downstream of the Anomaly Detect location will be inspected for flow conditions, and one of the following cases will apply:

1. If there are no segments, the anomaly will not be inserted. A configuration error alarm is raised.

2. If there is exactly one flowing segment, an anomaly is inserted into that segment. The anomaly will then start to move with the flow of the product in the segment adjusted by a slip factor.

3. If there are multiple flowing segments, an anomaly is inserted into the first flowing segment and is considered the original anomaly. For each additional flowing segment, an anomaly is inserted into that segment but is considered a clone of the original anomaly. The original and clone anomalies will then start to move with the flow of the product in their corresponding segments adjusted by a slip factor. The Controller is expected to correct this situation by choosing which anomaly to keep.

4. If there are no flowing segments, an anomaly is inserted into the first non-flowing segment and is considered the original anomaly. For each additional non-flowing segment, an anomaly is inserted into that segment but is considered a clone of the original anomaly. The original and clone anomalies will remain stationary until the segment starts flowing. The Controller is expected to correct this situation by choosing which anomaly to keep.

   Note: A unique name is generated for all automatically created clones using the naming convention explained above. However, to help the Controller track where clones originated from, an “original name” field is provided for cloned anomalies.

Anomaly flow through the pipeline

At each node, Batch Tracking needs to determine where to move the anomaly after it has reached the downstream boundary of a pipeline segment. Batch Tracking uses the same logic described in the previous section to move the original anomaly and, if needed, insert clones into downstream segment(s).