Darryl van der Peijl
What runs where?
The question for architects in manufacturing is not whether you go to the cloud or stay local. It is about which workloads belong where, and how you connect them without losing control.
Azure Local provides a clear separation. The platform runs physically in the factory, connected to Azure. That connection is what sets it apart from traditional on-premises infrastructure: a hybrid architecture that actually scales.
MES systems, PLC communication, real-time quality checks: everything that is latency-critical runs on Azure Local hardware in the factory. These workloads cannot tolerate the delay that cross-region connections introduce. At high latencies and high query volumes, processing time can climb to seconds. For production processes, that is too much.
One of our clients produces animal feed at locations worldwide, including sites where internet connections are unreliable. MES systems run locally, with the latency that production processes require. The hardware is prepared by Splitbrain and is plug-and-play on arrival at the remote location.
Centralized dashboards, AI model training, analytics on aggregated data, software updates. These workloads run in Azure. They benefit from cloud scalability and do not require low latency.
AI for process monitoring is a good example. The model is trained in the cloud on large datasets. Inference runs locally. Predictive maintenance works the same way: sensor data is processed locally, only aggregated insights go to the centralized dashboard.
Azure Arc connects the local infrastructure to the cloud. You manage all locations from a single dashboard in Azure. You see the status of every node, every VM, every workload. Updates are planned and rolled out centrally. Alerts come into one place.
That means automatic firmware and software updates without anyone needing to be physically present. Applications and configurations managed as code through Infrastructure as Code, not manually per location. Azure services such as monitoring, policy management, and security running via the cloud but operating on the local infrastructure. And integration with the hardware vendor for direct troubleshooting.
For organizations operating worldwide, this removes the dependency on local IT expertise that is increasingly hard to find.
Want to know what this architecture looks like for your organization? Get in touch.
This article is part of the longread Factory of the Future.
Darryl van der Peijl
Gerelateerde blogs
