Technology & research

What we mean by Hybrid Intelligence

An operator can explain why a pump runs at a certain time. Historical readings show what happened under that schedule. A model helps the team explore what could happen if the schedule changes. Each tells you something different about the same system.

How experience, data and models work together

Operating experience helps us identify which constraints the model needs to reflect and which scenarios are worth exploring.

Operating data shows how the system has behaved under conditions the team has already encountered.

Our own mathematical models describe how parts of the system interact, such as how pressure and flow change when pumps run at different times. We use machine learning to identify patterns in operating data that can inform those models.

A digital model of your operation

In Calejo Control, these inputs come together in a digital model of a specific operation. We check how well the model reflects the system using operating records and the team’s experience. The team can then compare scenarios and use the results to decide what to test in practice.

Modelling when data is incomplete

Operating records can contain gaps, and sensors only measure part of what happens in a system. Our research explores how to combine what we know about the physics of a system with what we can learn from the available data. Working with operators helps us see where a model reflects their experience and where it needs further work.