BooleX.ai DOCS Book a demo
Documentation

Get started

Set up

Get answers

Quick reference

Manual

IT documentation

Guides

Reference

Docs / Manual

Sequence of events

The timeline around a stop, the candidate initiating event, floods, chattering alarms and good-run comparison.

Open Sequence of events from the evidence card or an answer. BooleX rebuilds the timeline from the alarm history and the recorded tag changes, read through the historian; nothing is taken from the HMI screen.

How a stop is found

A stop is the sequence state leaving running for held or aborting; a trip is an interlock acting. The window runs from 10 minutes before the event (up to 24 hours when asked) to 2 minutes after it, ordered by source timestamp. Events with the same timestamp are marked =.

The candidate initiating event

BooleX marks the Candidate initiating event and up to two more, ranked:

  1. the earliest event on a cause chain from the holding or tripping interlock's cause tags in the PLC code;
  2. then the earliest event on the same asset as those cause tags;
  3. then the earliest other abnormal event.

Only events the program does not cause itself can be candidates: changes of field inputs, HMI and maintenance settings, and alarm activations. A routine command change at a step change is never marked. These are filtered out as routine:

  • a device's feedback within 10 seconds of its command;
  • a tag or alarm that changed more than 20 times in the 10 minutes before the event (only its last change counts, once it has stopped);
  • changes in the first 20 seconds after the plant starts running (start-up transients).

A site can change these four values in site.yaml.

The timeline shows the stop as a dashed line and the candidate initiating event as a band. Around the stop and Whole window change the time axis; trends of the cause tags are drawn on the same axis.

Floods, chattering and good runs

  • Alarm floods: more than 10 alarm activations in any 10-minute window (EEMUA 191), grouped by the shared root inputs of their cause chains, with the first alarm of each group.
  • Chattering alarms: 3 activations within 60 seconds.
  • Compared with good runs: for the sequence step that was running, each tag's average against the last good runs of the same step, listing those more than 3 standard deviations away.
  • Drive and robot faults imported from their own logs appear as Device fault events.

Every row cites its alm:// or ts:// source.