Turning industrial complexity into predictable, data-driven operations

As one of the world’s largest manufacturers of copper and copper‑alloy products, KME supplies sectors ranging from electronics and energy to industrial engineering and infrastructure, producing everything from rolled copper to sheets, strips, bars and pipes. At KME’s Fornaci di Barga plant in Italy, production takes place in a highly interconnected industrial environment. Rolling lines, heat-treatment processes and material-handling systems operate continuously, with each phase influencing the next. Even minor deviations in temperature, pressure or speed can have a direct impact on product quality, process stability and overall efficiency.

For years, Fornaci di Barga generated large volumes of technical data reflecting these variations. However, the data came from heterogeneous machines and systems, each with its own format, frequency and level of granularity. Operators and engineers lacked a comprehensive, repeatable view of how the production process behaves under different operating conditions. As a result, they often relied on experience and one-off analyses to manage daily operations, and a consistent, end-to-end understanding of the full production process remained out of reach.

The challenge was not the lack of data. It was the absence of a unified, reliable foundation capable of supporting systematic analysis and timely decision-making.

From fragmented data to a shared analytical view

To address this gap, KME sought to strengthen its ability to monitor critical process parameters, reduce uncertainty around complex behaviors and anticipate conditions that could lead to inefficiencies, scrap or unplanned downtime. Achieving these goals required both technological integration and analytical capability.

KME collaborated with SAS Partner Alleantia to implement an industrial OT/IT integration layer capable of collecting, standardizing and continuously streaming data from diverse production assets. This infrastructure ensured stable and consistent data flows, creating the conditions needed for advanced analytics.

Building on this foundation, KME introduced SAS Viya as its central data and AI platform. The goal was to move beyond isolated observations and develop analytical models that could describe, correlate and explain process dynamics that are not immediately observable in daily operations.

“We worked to create an infrastructure capable of unifying heterogeneous data streams, ensuring stability in data collection and preparing the foundation for reliable analytical use on the SAS Viya platform,” says Antonio Conati Barbaro, co‑founder and COO of Alleantia. “That technical base became the prerequisite for transforming shop‑floor observations into a structured advanced analytics initiative.”

At the Fornaci di Barga plant, that foundation had to support one of the most complex metallurgical production environments in Europe.

“This is a metallurgical plant the size of 70 football fields, with a long and complex production process built on huge, traditional machines that must be interconnected and made to communicate through a more modern language enabled by machine learning – driving greater efficiency, continuous improvement and sustainability, says Manuele Fanucci, Plant Director at Fornaci di Barga.

Advanced analytics to support operational decisions

For KME, the project represented a fundamental shift in how production processes are understood and managed, moving from experience-driven tuning of machines to insight-driven optimization based on governed data. As Francesco Bucci, Engineering and Technology Manager at Fornaci di Barga, says, “It marked a transition from machine engineering to data engineering.”

Using SAS Viya, KME’s engineering team now manages the full analytics life cycle within a single environment – from data preparation and cleansing to model development, validation and comparison. This unified approach reduces development time and increases transparency in analytical methods, allowing engineers to test different hypotheses and configurations with confidence.

“After the project was fully operational, we had a more robust information base,” Bucci says. “Beyond standardized data collection, we developed new capabilities to correlate variables, introduce predictive indicators to anticipate risk conditions, simulate alternative scenarios and validate analytical models over time.”

Three core capabilities of SAS Viya proved decisive in supporting this transformation.

First, OT/IT connectivity provided by Alleantia made it possible to reliably collect and contextualize high‑frequency data from machines, sensors and control systems across the production environment, creating a unified analytical foundation for advanced shop‑floor analytics at scale.

Second, the platform’s ability to support both descriptive and predictive analytics allowed KME to build models that identify critical thresholds, detect anomalous conditions and reveal non‑linear dynamics embedded in connected manufacturing processes. Engineers could experiment with different analytical configurations, compare models and develop synthetic performance indicators that translate large volumes of industrial data into actionable operational insight.

Finally, the scalability of SAS Viya allows these analytics to be operationalized and extended to additional processes or plants without changing the overall project structure – supporting a consistent approach to industrial analytics as KME expands its connected manufacturing and data-driven operations.

As a result, KME has gained greater visibility into the behavior of its production processes and a more reliable basis for intervention, improving the stability of daily operations.

Tangible operational and organizational benefits

By identifying deviations earlier, KME can intervene before issues escalate. Continuous monitoring and modeling have also helped reduce inefficiencies and scrap, while faster analysis cycles have improved the speed and quality of decision-making.

Beyond immediate operational gains, the project has helped consolidate technical knowledge about equipment behavior and process dynamics. This deeper understanding supports not only production optimization but also maintenance planning and the prevention of unplanned downtime.

Extending analytics to product-level environmental traceability

The same data-driven approach has enabled KME to extend analytics beyond operational performance to environmental accountability. By linking detailed process data with energy consumption and production parameters, the company can calculate the carbon footprint of each individual product based on real manufacturing data.

“It was possible to create digital twins of machines – and even of each meter of metal moving through the production lines,” says Stefano Linari, President and CEO of Alleantia. “Thanks to SAS Viya, KME can now evaluate the real environmental impact of every coil, based on the energy actually used for each individual operation.”

This capability shifts sustainability measurement from high-level approximations to precise, verifiable analysis embedded in industrial operations. Product-level environmental traceability becomes a direct outcome of measured and governed processes, rather than a downstream reporting exercise.

A scalable model for future growth

The project at Fornaci di Barga has established a scalable model for KME. The integrated OT/IT architecture and analytics framework can be replicated across other plants and processes, supporting a more consistent and comparable approach to data-driven manufacturing.

“Collecting and interpreting more than 40,000 industrial data points – and turning them into practical analyses and machine‑learning models for optimizing metallurgical processes – was a major undertaking,” says Conati Barbaro. “It enabled a significant step forward in digital growth for both KME and Alleantia.”

Looking ahead, KME plans to expand the range of monitored variables, refine predictive models and further integrate analytics into daily operations – strengthening its ability to observe, understand and govern complex manufacturing systems through data.