Case Study


The ARRIA NLG Engine is monitoring large-scale, mission-critical machinery located on oil and gas production platforms in deepwater Gulf of Mexico

A typical deepwater oil platform is a highly complex, integrated, industrial production facility operating 24/7/365. Its production capabilities depend on a number of large scale rotating machines such as turbines, compressors, pumps, generators and engines. An oil platform operates in one of the most challenging production environments in the world. The production output as well as the safety of the oil platform both depend on the reliable performance of its rotating equipment. When equipment begins to perform outside acceptable operating envelopes, embedded sensors detect the issue and cause alerts to be generated. In addition to rotating equipment, alerts cover a wide range of sources including wells, facilities, and sub-surface equipment. These alerts are monitored by automated systems and highly skilled engineers.

An Example of the Current Monitoring Situation in Complex, Multi-Asset, Surveillance Centres

  1. Automated systems monitor data streams and raise alerts.
  2. Expert analysts assess whether alerts are valid and need to be further investigated.
  3. For valid alerts, the experts write a situational analysis summation which includes context and background.
  4. Subject matter experts then further analyse the situation, and create a service recommendation for the equipment.

The problem is that while current monitoring systems generate alerts and notifications, they cannot articulate what has gone wrong. So expert engineers with decades of experience are required to analyse the alerts. But there is such a vast array of equipment generating so many thousands of alerts, it is impossible for human experts to analyse them all. Some control centre operators have likened this data overload experience to “drinking from a fire hose”. Further, the level of expertise required is so high, there is an acute shortage of human experts available to accurately analyse all of the alerts. As a result, alerts are missed and production suffers.

The Arria NLG Engine Effectively Solves This Problem

The Arria NLG Engine effectively becomes the expert. It instantly, automatically, continuously, accurately and effortlessly:

  • Reviews every alert even in the absence of the expert.
  • Performs analysis that determines what the alerts mean.
  • Provides a plain English report identifying which alerts need the expert’s attention and provides the situational awareness that allows the engineer to immediately formulate an action plan to correct the cause of the alerts.

The Arria NLG Engine determines and reports the problem and solution faster than the engineer ever could. It saves each engineer hours of analysis and report writing per day. It works around the clock—never sleeping.1


The Arria NLG Engine Instantly
Analyses Multiple Complex Inputs
From Large Scale Rotating Equipment


The Arria NLG Engine can take inputs from multiple, complex systems and disparate technology platforms, from any point in the service delivery stack, and from such inputs, generate engineering analysis and situational awareness reports.


The Arria NLG Engine effectively becomes the expert, doing the work of any number of surveillance engineers—24/7/365.



“It is a compelling summary that you would believe was produced by a human”


The Arria NLG reports are written as if by the most expert engineers in the field. Here’s how our existing oil and gas client described them: “We’re not talking about simple statements around readings of levels of curves. We’re talking about how many times a certain threshold was exceeded over a certain period of time. We’re talking about how many times people acted in the past. It is a compelling summary you would believe was produced by a human.”





2 Source: Dr. Anthony Edwards, CEO, Stepchange Global Limited.