Production Engineering Simplified with AI-based Advisory Systems

Advisory systems represent a specific and valuable category of AI application in production engineering — systems that improve decision-making without removing human judgement from the loop.

What an advisory system actually is

An advisory system is a tool that presents ranked recommendations to a human practitioner, who then reviews, modifies, and applies them. The system does not execute actions autonomously. The engineer remains in control.

This distinction matters commercially and practically. An advisory system:

Requires no changes to control system access or security architecture
Can be deployed without changes to operational procedures around autonomous action
Builds trust incrementally, as engineers validate recommendations over time
Generates a track record of predicted vs. actual outcomes

For most production teams, advisory mode is the right starting point — regardless of how confident they eventually become in the underlying models.

Where AI adds value in advisory mode

In production engineering, AI-based advisory systems add the most value in three areas:

Systematic screening — evaluating every well every day is not practical for a human team working with raw data. An advisory system can do this automatically, surfacing the wells that warrant attention and the specific actions that are likely to improve performance.

Impact ranking — not all optimisation opportunities are equal. An advisory system can estimate the production impact of each recommended action and rank them, so engineering time goes where it will have the most effect.

Supporting context — presenting a recommendation without context puts the entire evaluation burden on the engineer. A well-designed advisory system provides the trends, calculations, and reasoning that help an engineer quickly assess whether a recommendation makes sense.

The path to supervised and autonomous optimisation

Advisory mode isn't a compromise position — it's the foundation for more advanced capabilities. Once a team has validated that an advisory system is producing reliable recommendations, they can extend it:

Supervised optimisation — the system presents recommendations and applies them automatically after a defined approval window
Autonomous optimisation — approved action classes are executed automatically within operator-defined constraints

Nexgineer™ supports all three modes, with teams configuring which wells and action types fall into which category.

Learn how Nexgineer™ supports advisory, supervised, and autonomous optimisation workflows. Book a demo.