unthinkable
Case study · Mechademy, oil and gas turbomachinery engineering consultancy

Oil and gas plant teams get an alert about a failing compressor or turbine before it breaks down.

Mechademy wanted the oil and gas plants it advises to catch equipment failures earlier than their protection systems, and in some cases predict them, so production keeps running in harsh offshore and onshore sites.

Supply chain & assets (inventory, assets, procurement)

What changed

100+sensors connected (as reported)
80%less unplanned downtime at clients (as reported)
Up to 5%more yearly production output (as reported)
What stood in the way

What stood in the way.

What stood in the way
  • Equipment in harsh oil and gas sites broke down often, and failures stopped production.
  • Data from tens or hundreds of sensors sat in plant historians without a way to spot anomalies or poor performance in time.
What changed
  • 100+sensors connected (as reported)
  • 80%less unplanned downtime at clients (as reported)
  • Up to 5%more yearly production output (as reported)
What we built

What we built.

  • A connector agent that takes data from 100+ sensors in the plant historian into Mechademy's data lake.
  • ETL pipelines that remove missing values and outliers and check data types before analysis.
  • Performance models and machine learning and deep learning models, trained on historical data, that flag anomalies and send prescriptive alerts.
  • Plant layouts, machine parameters and models that admins configure and deploy from the admin screen, with role-based access.
  • Dashboards each client can shape, from the whole plant down to one data point, on AWS.
  • AWS
  • Apache Airflow
  • AWS EMR
  • PySpark
  • Kubernetes
  • Node.js
  • React
  • MongoDB
What it shows

What this work shows about us.

We build on the data a plant already has

The platform plugs into existing historians and condition monitoring, so plants get alerts without new instruments.

Engineers stay in charge

Mechademy's experts configure the models and layouts, and plant teams decide what to do with each alert.

Want similar machine learning monitoring for industrial equipment? Tell us what you want to change. Our AI assistant, Ask Unthinkable, helps you define your requirements, then connects you with the right team.

Start your plan ↗
If we built it today

If we built it today, with AI.

What we would add now, moment by moment.

Availability

An agent would turn each alert into a draft maintenance job with the readings behind it and the best window to stop the machine. Plant engineers still decide when equipment comes offline.

Ordering

An agent would check spare parts for the likely fault and draft the order before the job starts. Maintenance planners approve every order.

Ask Unthinkable

Want this for your business?

Tell us what you run or what you're building. Our AI assistant, Ask Unthinkable, helps you define your requirements, then connects you with a senior AI engineer.

Send us a note instead

Required

Thanks. A senior AI engineer will reply to you.
u
Ask UnthinkableClarity first, then the right team
Your conversation

An AI agent. This prototype uses prepared answers and sends nothing.