Edge Computing Architectures Are Fragmenting — And That's a Problem for Interoperability

The industrial edge computing market is booming. Spending on edge infrastructure in manufacturing is expected to hit $32 billion in 2026, according to IDC. But the rapid growth is creating a fragmentation problem that could haunt the industry for a decade.

The issue: every major platform vendor — AWS, Microsoft, Google, Siemens, Rockwell, Schneider Electric — has its own edge architecture with its own APIs, data models, and orchestration tools. A quality inspection model trained on AWS Panorama can't run on a Siemens Industrial Edge device without significant rework. Data collected by a Rockwell FactoryTalk Edge gateway doesn't natively flow into an Azure IoT Hub.

"We're building the industrial equivalent of the browser wars," said Jason Shepherd, VP of ecosystem at ZEDEDA, an edge orchestration startup. "Except instead of websites that don't render correctly, you have production lines that can't share data."

Standardization efforts exist but are moving slowly. The Open Manufacturing Platform, backed by BMW and Microsoft, has published reference architectures. The Eclipse Foundation's Sparkplug specification is gaining traction for MQTT-based data normalization. And the Linux Foundation's LF Edge project is trying to create a common runtime.

But adoption remains patchy. Most manufacturers are locked into single-vendor stacks because the integration cost of mixing platforms exceeds the benefit — at least in the short term. The result is edge infrastructure that works well within silos but can't support the cross-plant, cross-vendor analytics that Industry 4.0 was supposed to enable.