Skip to content
Close up of the meshed teeth of two large steel machine gears

Why Unmined

Built by engineers who have spent their careers inside oil and gas systems.

This is not a team learning what a plant is. The people here have led the delivery of the engineering, document and control systems that oil and gas operators and major capital projects run on, at BP, at Chevron and at Autonomy. That is why the product starts with your documents. It is why it does not need a capital request. And it is why every answer can be traced back to the machine it came from.

Unmined Limited · Dublin, Ireland

85 yrs

Combined experience across the four people below

Oil and gas

Where most of that experience was earned

Real time

Process control and time series data, on systems not allowed to stop

Audited

Built for regulators, courts and capital projects that check the working

The team

Four people who have done this before.

Every one of them has delivered systems into working industrial sites, under someone else's rules and someone else's deadline.

Portrait of Steve Loughran

Steve Loughran

Director

Thirty years building enterprise systems, and delivery teams led at Autonomy, BP and Chevron. He has built the engineering systems that oil and gas operators and major capital projects run on.

Portrait of Artem Demidovich

Artem Demidovich

Technology and innovation

Twenty years taking products from a strategy document to a live launch, in energy, finance, government and industry, including national platforms carrying millions of transactions.

Portrait of Tatyana Kan

Tatyana Kan

Enterprise architect

Ten years designing enterprise systems for oil and gas. Her work is the awkward part: real time process control, time series data, and systems not permitted to go down.

Portrait of Cory Brightman

Cory Brightman

Project and product delivery

Twenty five years delivering information and workflow systems in oil and gas, banking and insurance. Global rollouts, in regulated environments, across several companies at once.

What it changed

Four decisions that came out of that experience.

Each one is a mistake we have watched happen from the inside.

01 · Documents first

We start where the errors start

Plant modelling projects fail on the equipment list, not on the modelling. So the first thing we do is read your drawings and datasheets and build a list you can search. Nothing gets modelled on top of a list nobody trusts.

02 · Operating budget

Nothing here needs a capital request

We have sat on the capital projects where sensors lose the argument to a crusher rebuild, and they should. So the cost sits in the operating budget and is measured against a production number your site already reports.

03 · Traceable

Every answer names its machine

Building evidence systems for regulators teaches you that an answer nobody can trace is worthless. Every reading carries the machine it came from, the date, and how confident we are. It stands up when somebody checks.

04 · Whole signal

We keep the whole sound of the machine

Real time process data is where most of this work is lost. One average reading tells you something changed and nothing about what broke, so we send the whole sound of the machine, wirelessly, and keep the detail that gives a fault away.

The gap

What that experience keeps running into.

Four published industry estimates, not our own numbers. They describe the job we started the company to do.

5%

Of machines get a sensor that hears everything

95%

Of machines send one average reading, or nothing at all

80%

Of unplanned downtime comes from those unmeasured machines

75%

Of plant modelling projects fail on the data going in

Which machines get measured has little to do with how important they are. It is decided by the cost of cable. A main compressor is worth wiring up, so a conveyor drive at the far end of the plant gets a cheap sensor or gets nothing, and then stops the plant. Reaching those machines without a cable is the problem we set out to solve.

Somebody else paid for the hardest engineering. Aerospace and heavy engineering customers ask for the difficult things and fund that work directly. Two of our hardest problems were solved for them: sending everything a machine says through rock and water with no cable, and taking a reading straight off a spinning shaft, which a university research group tested and confirmed. We then strip the cost out of the design and keep the engineering. See how we do it

Questions

What buyers ask before the first call.

Six questions come up in nearly every first conversation. The answers are the same whether you ask them now or on the call.

We already run condition monitoring. Why add this?
Because your existing system almost certainly covers your handful of big machines properly and reports a rough average on everything else. We are not asking you to replace it. Our readings go into the systems you already own, and the corrected equipment list is written back into them too.
How long until we see something useful?
An equipment list for one unit comes back in weeks, from your documents alone. A sensor reports from the day it goes on, and we tune it to your equipment during the first project, before you commit to anything bigger.
Do we need to hire a specialist to read it?
No, and most sites could not hire one anyway. The reading of the data is done for you. What you get is the failing part named, an estimated date, how hard the machine is working, and an honest admission when we are not sure.
Our drawings and equipment lists are a mess.
That is the normal starting point, and it is why the Asset Register division exists. Anything the software is unsure about goes to a person, and where your systems disagree we show you the disagreement and let you settle it.
Will the hardware survive our plant?
It is sealed against dust, mud and high pressure hosing. A hundred of them have run for more than two years in a heavy processing plant, and other units have run three years without a break. Versions for trains meet the rail safety standards.
What does it cost?
It depends what you want covered, so there is no price list. The structure is always the same: a fixed price for the first project, then a running cost from the operating budget, measured against output, uptime, yield or energy. Nobody needs to raise a capital request to start.
“We did not have a data shortage. We had thousands of numbers and nothing we could model against.”
Head of processing, precious metals producer.

Next step

Test the argument on your plant.

Name the machines you have never been able to justify wiring up. We will tell you three things: what we have already proved on that kind of machine, what a first job would measure, and what it would cost to find out.