Rockwell Automation’s Singapore factory rolled out a GenAI-Powered Maintenance Copilot in October 2025 to accelerate troubleshooting and cut downtime across hundreds of machines.

The implementation aims to preserve tacit knowledge, standardize troubleshooting, and reduce knowledge loss as experienced workers retire, addressing the tribal knowledge gap.

In June, the facility earned World Economic Forum global lighthouse status for efficiency gains, including a 43% increase in units per person-hour, a 35% drop in defects, and a 67% faster time-to-competency.

The copilot organizes information by symptoms, highlights likely causes, and provides step-by-step, natural-language troubleshooting guides drawn from veteran engineers’ insights and colleagues’ notes.

The AI assistant is trained on Rockwell’s expert knowledge, manufacturing software data, and machine manuals, and runs on Microsoft Azure via Azure OpenAI in Microsoft Foundry with OpenAI’s GPT-5.4-mini.

Since deployment, overall machine downtime has fallen by a third and servicing/spares costs have dropped about a quarter, while onboarding times for new workers shrank from nine months to roughly three.

Beyond the copilot, the Singapore plant utilizes additional AI capabilities, including a multi-agent quality assurance system and predictive maintenance, contributing to broader efficiency gains.