Why reliability is built long before a PLC leaves the factory

Quality and reliability are related but they are not interchangeable. Quality tells you whether a product meets expectations when it leaves the factory. Reliability tells you how it performs over years of real-world operation. At Hornberg, reliability is the standard we hold ourselves to - and it begins long before a product reaches the production line.

Reliability from prototype to production

The Hornberg site operates as an internal electronics manufacturing service provider for several ABB businesses. That breadth of experience gives our team an unusual vantage point. Manufacturing specialists are involved from the prototype phase, working alongside R&D teams to assess producibility, automation potential and long-term process stability. The goal is not only to confirm that a product works, but that it can be manufactured efficiently and to the highest quality standards across its entire lifecycle.

Production and quality data are reviewed both on a daily and monthly basis, trends are tracked and recurring issues are escalated. When something keeps appearing in the data, manufacturing and development engineers work together to find the root cause and close the loop.

Every failure is a source of learning

The word "failure" carries a negative connotation in most settings. At Hornberg, it is one of the most valuable things that can happen - provided that we respond to it in the right way.

Products move through multiple inspection and testing stages before shipment. When a PLC fails a test, the goal is to understand why the failure occurred in the first place and use that knowledge to prevent similar issues in the future. Removing the unit from the production flow immediately is how that process begins. An operator then scans the barcode and straightaway accesses detailed failure data: measurement results, test logs and a full record of what went wrong. The product is analysed, repaired and retested.

The operator records the root cause and the corrective actions taken. This information is stored in a centralised database and becomes part of a growing knowledge base that is reviewed and analysed every month. Where patterns appear, processes are improved - and where necessary, that learning feeds back to product development.

This approach has delivered measurable results. Over the past several years, first-pass yield on AC500 PLC production has improved from around 90% to almost 98%. That gain sounds modest until you consider what it represents: constant attention to hundreds of variables including component tolerances, supplier variation, environmental conditions on the factory floor and the variation of multi-source material needed to cover the latest geopolitical disruptions. Some effects take days or even weeks to fully understand. Finding the answer, and continuously improving the process, is one of the most motivating parts of the job.

Testing beyond normal operating conditions

A PLC that only performs well under ideal conditions is not reliable - it is merely functional. Real-world applications push equipment to their limits, and our testing reflects that.

At Hornberg, products are routinely tested at the edges of their operating range and beyond: under-voltage and over-voltage scenarios, maximum load testing, and reliability checks inside climate chambers. Selected units from production batches undergo endurance testing at elevated temperatures to help verify long-term performance.

The value of this approach was put to the test not long ago. During production testing of an AC500 IO-Module, we test even below the specified lower limit of supply voltage - 19.2 V DC - rather than testing at a nominal supply voltage of 24 V DC. The test revealed a tolerance variation that caused the product to malfunction, one that would have been completely invisible under standard operating conditions. The product would have passed every standard check and reached the customer, only to potentially fail in the field when facing multiple environmental factors like temperature, undervoltage EMC, etc.

Because we test beyond the limits, we caught it. The circuit design was revised and an alternative component was selected before any products left the factory. The principle is clear: we do not only test for what should happen. We test for what could happen.

Traceability as a risk management tool

Every test result, manufacturing record and component dataset is stored in central databases, allowing us to trace individual components and production batches across the full product lifecycle. If a supplier issue were ever identified after shipment, we could quickly determine exactly which products were affected - supporting targeted corrective action rather than broad recalls, and directly protecting customer operations.

Quality, reliability and sustainability are the same conversation

These three topics are often discussed in isolation, but they shouldn't be. Reliable manufacturing processes generate less waste because fewer failures mean fewer repairs, fewer returns and less unnecessary transportation. Every product repaired rather than scrapped helps to extend the life of valuable resources.

At Hornberg, the default is always to repair. Whether a defect is found during production or after a customer return, the preferred approach is to identify the root cause, replace the defective component and return the product to service wherever possible. The site also supports ABB's broader Mission to Zero ambitions through investments in heat-pump technology and more sustainable packaging.

The future of reliability

Artificial intelligence offers a meaningful opportunity to extend what we already do well. One concept currently being explored is an AI-based repair assistant that analyses historical test data and repair records to help operators identify root causes more quickly. This development could accelerate diagnosis for experienced engineers and give newer team members immediate access to decades of trouble-shooting knowledge. The goal is to make human judgement faster and more consistent, not replace it entirely.

Ultimately, whether through advanced testing, continuous improvement or future AI applications, the objective remains unchanged: zero customer returns. Reliability is not measured by test scores, database entries or yield percentages. It is measured by the confidence customers place in ABB products every day, and by their decision to come back for the next one.