Emerging Skills Program — no certification exam
Artificial Intelligence & Machine Learning for Oil & Gas
What these tools can and cannot be trusted with in inspection and integrity work, and where the engineer has to stay in the loop.
Why this matters
The problem this solves
The useful question is not whether these tools work. It is which parts of an integrity engineer’s week they can be trusted with, and which parts they will confidently get wrong.
Report drafting, standard lookup, data tidying and first-pass summarisation are genuinely faster with a language model. Calculation, code interpretation and anything that ends in a signature are not — and the failure mode is a fluent, plausible, wrong answer, which is considerably worse than an obvious one.
This programme is about the boundary, and about the discipline that makes the boundary safe: engineer-in-the-loop, verifiable outputs, and a clear rule on what confidential data may go where. The hands-on treatment with the tools themselves is the live course.
Before you enrol
Is this a certification?
The short answer is no, and it is worth being explicit about it rather than leaving it to be discovered later.
What this programme gives you
- A clear boundary between the tasks these tools help with and the tasks they endanger
- A workable engineer-in-the-loop review pattern
- A data-handling rule you can put in front of an IT or compliance function
- An EIM certificate of completion
What it does not
- Not preparation for a certification exam — no certifying body examines this subject
- Not a data science, programming or machine learning engineering qualification
- Not a vendor tool certification
- Not a licence to let a model perform an engineering calculation unchecked
If you need a recognised credential, start at certification exam preparation.
Who it's for
Roles this is built around
Inspection and corrosion engineers
Any level
Producing a lot of documentation, and wondering how much of it is safely automatable.
Asset integrity leads
Team level
Being asked what the team’s AI policy is, and needing an answer that survives an audit.
Data and reliability analysts
Technical
Sitting between the historian and the engineering decision.
What it covers
The outline
An outline, not a weighted Body of Knowledge — nobody publishes weights for a subject with no exam behind it.
What these models actually do
Enough of the mechanism to predict where the failures will be, without the mathematics.
Where they help today
Report drafting, summarisation, standard navigation and data cleanup — the defensible wins.
Where they must not be trusted
Calculation, code interpretation and anything carrying a signature. Why fluency is not accuracy.
Engineer in the loop
Making the output verifiable, and what review has to look like for it to be real.
Confidentiality and data handling
What may be pasted where, and the questions to ask before a tool is approved.
Predictive analytics, honestly
What condition data can genuinely support, and why most integrity datasets are too thin for it.
Outcomes
What you will be able to do
- Judge whether a given integrity task is a good or a bad fit for a language model
- Design a review step that would actually catch a confident wrong answer
- Set a defensible rule for what plant and client data may be entered into which tool
- Recognise a predictive-maintenance claim that the underlying data cannot support
- Brief a team on where these tools help without overselling what they can do
How it runs
Ways to take it
Formats offered — not scheduled dates. Live cohort dates live on the course page, so there is only ever one copy of a date.
| Format | Duration | Group size | Notes |
|---|---|---|---|
| Live online | 1 day | Up to 15 | The explainer. The hands-on tool work is the live course |
| In-house | By arrangement | 6–20 | Delivered against your own equipment, your own findings and your own procedures |
Also available live
Take this as a scheduled course
FAQs
Questions people ask first
No programming is required. The programme assumes an inspection, corrosion or integrity background and explains the tools from there, rather than assuming a software background and explaining the industry.
The live course is hands-on with the tools across inspection reporting, RBI and FFS support, NDT, shutdown planning and SAP PM. This page sets the boundary — what to trust, what not to, and how to review — and points at the course for the practice.
The tool list will. The boundary will not: the distinction between tasks where a plausible wrong answer is cheap and tasks where it is expensive is a property of the work, not of this year’s models.
Want the full outline?
Tell us your role and what you are trying to fix, and we will send the programme outline and what it would take to run it for your team.