Case Study · Lonza · Pharma

OFS Validation and Control-System Resiliency Testing

Lonza · Modicon M580 · Schneider OFS/OPC · AVEVA Application Server / Batch

Context

At Lonza (pharmaceutical manufacturing), the control stack included Modicon M580 PLCs, Schneider OFS OPC connectivity, AVEVA Application Server, AVEVA Batch, virtualized Windows Server, and redundant infrastructure components. The work focused on proving that this stack could take disturbances and recover without turning PLC-to-server failures into lasting manufacturing risk.

Problem

PLC–server failures were not just a blank HMI problem. They could produce bad tags, data loss, HELD/ABORTED batches, and failover gaps — symptoms that look like “display issues” but are really control-path and batch-continuity failures.

Role

I led functional and failure-mode testing for the OFS architecture: reviewing expected behavior, running controlled failover and infrastructure tests, monitoring PLC / OPC / Application Server / batch behavior, assessing tag quality and historian data, distinguishing transients from defects, and documenting results for engineering and validation audiences.

Tests and acceptance

Controlled disturbances included:

OFS/OPC interrupt
Application Server failover while active
DA server failover
VM migration, snapshots, antivirus scan, and server failover scenarios

Acceptance criteria:

Tags return to Good quality
Automatic recovery
Recipes continue
No persistent HELD/ABORTED batch state
Missed reads stay within defined thresholds

Result

The system recovered from disturbances without persistent batch or process failures, producing measurable resiliency evidence — not just HMI display recovery.

Takeaway

Test failure modes, not only happy-path function. Measure recovery. Treat automation and IT as one stack. Use the historian to separate display symptoms from real manufacturing risk.

View resume Get in touch