Most industrial IoT monitoring platforms assume reliable cloud connectivity. That’s fine for office environments. It’s not always fine for a factory floor in a remote location, an offshore platform, a water treatment works, or any industrial site where network connectivity is intermittent, restricted, or subject to latency constraints that make cloud round-trips impractical.

AWS IoT SiteWise Edge addresses this by running the data collection and processing layer locally, on-site hardware, syncing to AWS when it can and functioning independently when it can’t. Here’s what that actually means in practice.

What SiteWise Edge Does

At its core, SiteWise Edge is a software component that runs on a gateway — a server or ruggedised industrial PC on your site — and does three things: it collects data from your industrial equipment, it processes and organises that data into an asset model, and it sends the harmonised output to AWS IoT SiteWise in the cloud.

The connectivity options are the key practical feature. SiteWise Edge supports OPC-UA natively, which means it can talk directly to PLCs, SCADA systems, and manufacturing equipment from Siemens, Rockwell, Beckhoff, and most other major industrial vendors. It also handles Modbus TCP for older devices and EtherNet/IP for Allen-Bradley equipment. If you have a mixture of legacy and modern kit on the same plant floor — which is almost universal — SiteWise Edge can pull from all of it through a single gateway configuration.

Asset Models and Local Computation

The asset model is what distinguishes SiteWise from a simple data forwarder. You define a hierarchy — factory, production line, machine, sensor — and assign measurements and calculated metrics to each level. Those calculations run at the edge, not in the cloud.

This matters for two reasons. First, it means KPIs like OEE (Overall Equipment Effectiveness), quality rates, and throughput calculations are available on the local network even when the WAN link is down. Supervisors can see their dashboards without waiting for a cloud round-trip. Second, it reduces data egress costs significantly: instead of sending every raw sensor reading to the cloud, you send aggregated, processed metrics at a lower frequency, with the raw data available locally for deeper analysis when needed.

Data buffering is built in. If connectivity drops, SiteWise Edge holds the data locally and syncs it to the cloud once the connection is restored. You don’t lose historical data because of a network interruption.

Setting It Up: What to Expect

The gateway software runs on AWS-certified hardware or, more interestingly, on Siemens Industrial Edge — which means you can deploy SiteWise Edge on the same platform you’re already using for Siemens edge applications. Dell Edge Gateways and similar industrial PCs are also supported.

Configuration happens in the AWS console: you define the gateway, configure the OPC-UA server connections with the relevant endpoint URLs and security certificates, and map OPC-UA node IDs to SiteWise asset properties. This is the part that takes real engineering time, particularly on older SCADA systems where the OPC-UA configuration hasn’t been documented in years.

Once data is flowing, you can set up anomaly detection and alerting using AWS IoT Events at the edge, create dashboards in AWS IoT SiteWise Monitor, or pipe the data onwards into AWS analytics services for more advanced processing.

Predictive Maintenance Use Cases

The most common deployment pattern is predictive maintenance: collecting vibration, temperature, and current draw from rotating equipment — motors, pumps, compressors — and using that data to identify degradation before it becomes failure.

A practical example: a paper mill running 40 pumps across its site. With SiteWise Edge on each production hall’s local gateway, you collect vibration signatures from all 40 pumps locally, compute rolling averages and standard deviations at the edge, and trigger alerts when a pump’s signature deviates beyond a threshold. The maintenance team gets an alert on their local dashboard before the bearing fails. The detailed historical data is synced to the cloud for trend analysis and to improve the alert thresholds over time.

When SiteWise Edge Makes Sense

SiteWise Edge is worth evaluating if you’re already invested in AWS — you’re using other AWS services for cloud analytics, your IT team is familiar with IAM and CloudFormation, and you want IIoT to integrate naturally with that environment. It’s also a strong choice if you need OPC-UA connectivity to a broad range of equipment and want a managed, supported solution rather than building your own broker.

It’s less obviously the right choice if you’re already using Azure (Azure IoT Operations is the equivalent), if your connectivity is genuinely reliable and low-latency round-trips aren’t a concern, or if your data volumes are small enough that a simpler MQTT-to-cloud approach would do the job without the additional complexity.

The Honest Trade-off

SiteWise Edge involves real configuration work upfront. Mapping OPC-UA node trees, defining asset models, and debugging connectivity to 20-year-old PLCs isn’t glamorous, and the documentation, though thorough, rewards patience. The payoff is a structured, cloud-managed industrial data layer that integrates with the rest of AWS — which, for teams already running their analytics in AWS, is significant.

For new industrial IoT deployments on sites where cloud connectivity can’t be guaranteed, the edge-first architecture makes more sense than it did even two years ago. SiteWise Edge is a mature option in that space.