Contattaci

Siamo qui per aiutarti

Case study

Securing OT networks and boosting visual inspection accuracy with edge AI

Industrial

A global manufacturer in the building materials sector set out to strengthen its digital operations by modernizing plant-floor infrastructure. The goal: improve operational technology (OT) security and deploy real-time AI for quality control, specifically, for detecting paint defects during production.

To achieve this, the company partnered with Eurotech to implement a scalable, edge-native architecture.

Annual revenue

~$810 million

Production facilities

~200 across North America

Employees

~1,800

Challenge

 

The customer faced three critical hurdles. First, OT data had to be shared with corporate IT and select vendors without compromising safety or breaching compliance boundaries. Second, paint anomalies on high-value components were causing costly rework and shipment delays. Finally, a modern SCADA system with real-time capabilities was needed: minimizing cloud dependency and enabling fast, local decision-making.

Solution

 

The company deployed a two-tier edge architecture based on Eurotech’s ReliaGATE 10-14 and ReliaCOR 54-13, enabling both secure connectivity and edge intelligence.

 

ReliaGATE 10-14 for OT data segmentation and secure remote access

 

The ReliaGATE 10-14, positioned at the boundary between IT and OT networks, enabled strict segmentation and secure data flow. With dual Ethernet ports and IEC 62443-grade security features, it established a robust cyber-hardening layer.
Built-in VPN tunneling allowed approved vendors to remotely access plant systems without directly exposing sensitive assets. Additionally, support for industrial protocols via the Everyware Software Framework (ESF) ensured seamless integration with legacy equipment and simplified device onboarding. By pre-processing OT data before transmitting, it to IT or cloud systems, the gateway contributed to a more secure and streamlined data architecture.

 

ReliaCOR 54-13 for edge SCADA and AI-powered visual inspection

 

At the plant floor level, the ReliaCOR 54-13 edge server hosted both SCADA and AI applications. Running a full Ignition SCADA environment, it provided real-time HMI, alarm handling, and historical data logging. Simultaneously, it supported a containerized AI inference engine trained to detect paint defects on components as they moved through the line. Thanks to multi-core Intel Xeon processors, GPU compatibility, and rugged fanless construction, the system ensured resilient, high-performance computing in harsh environments. Its native support for Docker and Kubernetes allowed scalable deployment and orchestration of workloads, making it easy to extend the system over time.

Results

 

With Eurotech’s edge solution in place, the customer securely exposed OT data to IT and vendor systems without compromising plant integrity. Remote access became controlled and compliant, streamlining maintenance while maintaining security. On the production line, real-time AI inspection drastically reduced paint defects, cutting rework and improving delivery timelines. Operators benefited from faster, on-site feedback thanks to local SCADA processing, eliminating delays tied to cloud reliance. The result is a more secure, efficient, and scalable edge infrastructure: ready for future AI and analytics deployments.