Cement & Concrete

Clinkum

Autonomous AI for cement and concrete production — burning-zone control, clinker-quality sensing, energy and CO2 optimisation.

Seed Physical AI $2M raised 2026 clinkum.com
Overview

What Clinkum does.

Clinkum runs the cement and concrete process: burning-zone and kiln control, clinker-quality sensing, grinding, mixing, energy and CO2 optimisation, and the plant digital twin.

Every kiln deployment writes labelled pyroprocess data back into the model, so the system generalises across fuels, raw mixes and plant configurations.

At a glance
DetailValue
StageSeed
CategoryPhysical AI
Year2026
Websiteclinkum.com
HeadquartersNot disclosed
Raised$2M
Investor10xC
ModelPer-kiln software subscription with an energy and yield outcome component.
Measured byClinker quality, fuel and power per tonne, kiln uptime, plants expanded per customer.
Engineers reviewing kiln process data on monitors
The kiln Burning-zone control that holds clinker on quality
Products

One loop, module by module.

Every deployment on clinkum.com writes labelled traces back into the model, so the system generalises across plants, alloys, grades and geometries.

01

Pyron

Burning-zone and kiln control that holds temperature and chemistry on profile.

02

Clinvue

Clinker-quality sensing from the cooler and the lab.

03

Cemtwin

Plant digital twin for what-if, ramp and energy planning.

04

Carbex

Energy and CO2 optimisation across kiln, mill and power.

05

Refron

Refractory and kiln-health intelligence.

06

Pourix

Concrete and ready-mix batch control.

Why we invested

The thesis behind Clinkum.

Cement is a high-tonnage, low-margin business where fuel and power dominate cost and kiln uptime dominates output. The kiln is still run on the experience of a few operators who retire faster than they are replaced.

Clinkum encodes that experience. The first deployment is one kiln, entered through a paid pilot, and the same model generalises to the next plant with privacy-preserving priors. Energy saved and clinker quality held are the two numbers the buyer already reports to the board.

Who is building it

Marcus Whitfield

Chief Executive Officer

Marcus owns plant relationships, the pilot programme and the commercial model.

Sarah Kowalski

Chief Technology Officer

Sarah owns the kiln models, the edge runtime and the write-back into pyroprocess control.

A production chart on a monitor
The output Tonnes on quality, fuel down

Building something similar?

Building in Physical AI? We want to hear about it.

← Back to portfolio