In industrial manufacturing, a narrow margin often determines profitability and quality. This is particularly true in steel production—for example, in hot-dip galvanizing. In this process, steel strip is passed through a zinc bath to apply a corrosion-protective coating. The challenge: If the zinc layer is too thin, corrosion protection suffers. If it is too thick, material costs and resource consumption rise.

Traditionally, coating thickness is measured only after a delay—once the steel strip has already traveled several meters. Corrections made at this late stage are therefore too late and lead to costly downgrades.

This is where the use of industrial artificial intelligence comes in. A jointly developed solution by the automation company EMG Automation and the industrial AI company NEUROLOGIQ continuously analyzes process data from the plant. A cascade of different AI models predicts the future coating thickness even while the production process is underway.

This allows the plant control system to be adjusted in real time. Instead of reacting to quality deviations after the fact, the process is controlled proactively.

The AI development was initiated, among others, by Simon Sack, founder and CEO of NEUROLOGIQ. Sack is considered a leading expert in industrial AI and focuses on the application of artificial intelligence in complex industrial processes. His approach: to view AI not as an isolated technology, but as an integral part of modern production systems.

The models used analyze large amounts of historical production data, identify patterns, and calculate new setpoints for plant control within milliseconds. This allows parameters such as speed, temperature, or air pressure to be automatically optimized.

The effects are measurable: material usage can be reduced, scrap avoided, and process stability increased. Especially in material- and energy-intensive industries such as steel production, such optimizations can have significant economic impacts.

At the same time, this example illustrates how the role of AI in industry is evolving. While many projects began as pilot applications, productive systems are increasingly being developed that are directly integrated into industrial control architectures and provide automated support for operational decisions.

The use of Industrial AI is thus evolving from an experimental innovation project into a productive tool for process optimization—and into a central building block of the digital transformation of industrial production.