1 hourto a confirmed cause, down from 4
The problem
After a unit trips, an engineer pulls alarms, sequence-of-events records and SCADA data by hand, often in the evening, to work out what happened before the unit can safely return. Every hour of that work is an hour the unit stays off and power is bought to replace it.
What we ran
Faults and protection operations injected into a plant and its interconnection on the model, generating the alarm stream so the true cause of each event is known.
What the agent did
It compresses the alarm flood, builds the sequence of events, proposes the cause, attributes customer minutes correctly, and drafts the event report and breakdown log. The engineer confirms the cause and releases the unit.
Results
| Measure | Today | With Evardi |
|---|---|---|
| Hours to a confirmed cause | 4 | 1 |
| Event report | Written by hand | Drafted for review |
| Cause correct on injected events | Not measured | Reported as a share |
What it is worth
- Faster return: trips a year × hours saved × unit size × replacement-price gap
- Call-back overtime: trips × hours × overtime rate
* Study simulated with partners on public data.
Run it on your network
Alarm and sequence-of-events records from two past trips. We replay them and show the cause the agent reaches, and how fast.
Full study
Request the full study.
The complete write-up of “From a unit trip to a confirmed cause in an hour, with the report drafted”: method, data, results and assumptions, as a PDF. Leave your work email and we'll send it.
