Darryl van der Peijl
Why infrastructure makes the difference
The smart factory market is expected to grow at 9 to 10% per year, reaching a size of more than $300 billion by the end of this decade. Industrial AI, real-time process monitoring, and predictive maintenance are becoming standard in manufacturing. But that growth does not happen by itself. It requires infrastructure that bridges the gap between cloud and factory floor.
AI model training requires large datasets and significant compute power. That belongs in the cloud. But inference, applying the trained model to real-time production data, needs to happen locally. The same latency requirements that apply to MES systems apply to AI inference on the factory floor.
Think of quality control on a conveyor belt moving products at high speed. The AI model needs to detect defects and intervene in milliseconds. If that inference call has to travel to a cloud region and back, the product has already passed. The same applies to predictive maintenance: sensor data needs to be processed locally and in real time to prevent unplanned downtime.
Azure Local makes this pattern possible. Training happens in the cloud on large datasets. Inference runs on local hardware, close to the machines. Azure Arc connects both, so you manage the entire pipeline from a single location.
For process monitoring, sensor data is processed locally. Only aggregated insights go to the centralized dashboard in Azure. This keeps latency low where it matters and uses cloud scalability where it adds value.
One of our clients applies this approach across production sites worldwide. The AI models are trained centrally. Updated models are rolled out to local Azure Local environments in the factory as part of the regular update cycle. No manual intervention at the remote location.
The conversation about AI in manufacturing often starts with the use case: predictive maintenance, quality control, process optimization. But the real question is where the compute power sits. Without local infrastructure that supports AI inference with low latency, most of these applications remain theoretical.
According to IDC, global spending on edge computing will reach approximately $261 billion in 2025, with expected growth to $380 billion by 2028. Manufacturing is one of the largest sectors driving that growth. The reason: real-time monitoring of equipment and processes, reducing downtime, and improving operational efficiency.
Are you working on a hardware refresh or a migration away from VMware? Then this is the moment to also determine which AI workloads you want to run locally. The infrastructure decision you make now determines what is possible two years from now. Get in touch to discuss the options.
This article is part of the longread Factory of the Future.
Darryl van der Peijl
Gerelateerde blogs
