How Batch AI improves concrete quality while reducing cement use and emissions
Batch AI – an AI-powered tool for optimising ingredient mixes in concrete production – helps Heidelberg Materials produce concrete more consistently while using less cement. It makes batching more precise, so plants can meet requirements without relying on extra safety margins. It is also part of Heidelberg Materials’ broader strategy to use digital and AI solutions in core operations to improve performance and sustainability.
We spoke with Nicolas Gunkel about how Batch AI works in practice, where it is already delivering results, and why precision in concrete production matters for both operational performance and emissions reduction.
Can you explain to someone who has never worked in a concrete plant: What is batching and why is it so critical to concrete production?
Batching means mixing different ingredients – cement, aggregates, water, and chemical admixtures – according to a specific recipe to produce concrete. Each recipe defines how the concrete should behave, for example how strong it becomes, how quickly it sets, or how it performs under certain weather conditions.
This step matters because even small deviations affect quality. Concrete must meet strict regulatory and customer-defined standards. If a batch fails to meet them, customers can reject it on site, which leads to financial losses and operational disruptions. Precise and consistent batching therefore directly determines product quality, cost efficiency, and customer satisfaction.
And what is Batch AI? How would you explain it to someone outside operations or technology?
Batch AI is a software from our partner Command Alkon that helps concrete mixes match their recipes as closely as possible. It uses machine learning to learn from previous production cycles and continuously improves how plants weigh and dose materials.
In everyday terms, Batch AI fine-tunes the batching process. It adjusts valves, gates, and material flows more precisely than manual intervention alone. The goal is not to change the recipe, but to execute it more accurately and consistently, with less waste.
Where exactly does Batch AI operate in the production process? At what point does it support the batching operation?
Batch AI operates directly during weighing, dosing, and mixing, which are the core steps of batching. The batch computer first translates incoming orders into a concrete recipe.
Once Batch AI is active, it runs alongside the batch computer. Based on what it learned from earlier batches, it intervenes at specific moments during weighing and dosing. These targeted real-time adjustments help keep each batch as close as possible to the intended recipe while staying within quality tolerances.
How does Batch AI work together with existing equipment and control systems? Did introducing it require major changes to plants or workflows?
Batch AI integrates with existing batch computers and control systems as a software layer. Plants do not need to redesign workflows or replace equipment.
Operators continue to work with familiar systems. Batch AI optimises how these systems interact by adjusting control signals in the background. This approach allows plants with very different technical setups to introduce the solution without major disruption.
How does Batch AI support plant operators in their daily work? What types of tasks does it reduce or simplify?
Batching demands constant focus. Operators monitor camera feeds, scales, and material flows, often making rapid decisions throughout the day. Timing errors – such as closing a valve slightly too late – can quickly lead to overuse of materials.
Batch AI reduces this pressure. It automatically intervenes at critical moments and corrects small deviations. As a result, operators spend less time on manual fine-tuning and more time supervising the overall process.
How did you address topics such as trust in the system, accountability, and the role of human expertise?
We built trust by clearly positioning Batch AI as a support tool that enhances, rather than replaces, human expertise. Operators remain accountable for quality and production decisions.
The combination of human judgment and machine precision delivers more stable and consistent results than either could achieve alone.
What measurable improvements have you observed so far in terms of accuracy, material usage, or consistency?
In plants around Edmonton in Canada – where the rollout is ongoing across North America – batch accuracy increased by up to 30 percentage points after introducing Batch AI. Another high-performing plant in the same region still improved accuracy by around 10 to 15 percentage points.
These results show that Batch AI improves consistency not only where performance lagged, but also where plants already operated at a high level.
How does Batch AI support Heidelberg Materials’ sustainability goals without compromising quality?
Cement production is energy-intensive and a major source of CO₂ emissions. It is also the most expensive component in concrete.
Batch AI improves precision, allowing plants to meet quality requirements without relying on excessive safety margins. This reduces both emissions and costs while maintaining performance.
What are the next steps for the project, and how will employees be involved as it evolves?
The next phase focuses on rollout. After North America, Batch AI continues to expand in the UK and Australia, followed by several European countries such as France and Germany. Asia, including Malaysia, also sits on the roadmap.
Employees remain closely involved. Feedback from plant teams helps validate results in different environments and guides further improvements.
- Proven impact in Canada
- Accuracy increased by up to 30 percentage points
- High-performing plants still improved by 10-15 percentage points
- Why it matters: Higher accuracy directly reduces cement overuse, which lowers both production costs and CO₂ emissions without compromising quality.
- Command Alkon – one of Heidelberg Materials’ key tech partners – is specialised in creating digital solutions for the heavy building materials industry. Its digital platforms support end-to-end operations such as batching, dispatch, logistics, and production management, helping producers improve efficiency and reduce emissions.