AWS Architecture Blog

Field Notes: Connecting Industrial Assets and Machines to the AWS Cloud

One of the challenges faced by manufacturers who are building a smart factory, is how to securely connect to and ingest data from operational data sources. These include machines and industrial assets connecting into their industrial data platform.

The ‘industrial edge’ is a foundational architectural building block that addresses this and enables OT/IT integration (operational technology/information technology) by extending Cloud services into the four walls of a factory. These services provide secure industrial connectivity, data ingestion and transformation, and the platform to deploy and run Industry 4.0 microservice applications and machine learning models in factories for low latency and production critical use cases.

This post shows how to connect to and extract data from industrial assets and machines via industrial protocols such as OPC UA using AWS IoT Greengrass connectors into AWS IoT SiteWise or an Industrial IoT (IIoT) application running in the AWS Cloud.

Architecture Overview

Industrial Edge

AWS IoT Greengrass is a software-based edge runtime and cloud service that is installed on an industrial PC or an Edge gateway and is connected to the manufacturing or industrial network.  It provides the platform to securely build, deploy, and manage industrial applications for local data processing and messaging.

Data can be ingested directly from industrial devices into the AWS Cloud by deploying the required industrial protocol connector to the AWS IoT Greengrass group as shown in the following diagram.

IoT Sitewise Architecture diagram

The AWS IoT SiteWise connector provides ingestion of data from OPC UA sources, and in December 2020 AWS announced the launch of the IoT Ethernet IP Protocol Adapter and Modbus-TCP Protocol Adapter connectors to ingest data from EtherNet/IP and Modbus-TCP data sources respectively.

Data ingestion

To ingest data into AWS IoT SiteWise, the AWS IoT Greengrass group must be configured to include the AWS IoT SiteWise gateway connector. Configuration of an AWS IoT SiteWise gateway can be done using the AWS Console, the AWS Command Line Interface or using AWS CloudFormation, which now supports AWS IoT SiteWise resources.

The following code is an example of a AWS IoT SiteWise gateway resource definition for an AWS CloudFormation template, which creates an AWS IoT SiteWise gateway to ingest data via OPC UA from a PLC. The example uses the following AWS CloudFormation parameters to define the connection:

Parameter examples

    Type: AWS::IoTSiteWise::Gateway
      GatewayName: Edge-GW-01
          GroupArn: !Sub "arn:aws:greengrass:${AWS::Region}:${AWS::AccountId}:/greengrass/groups/${GreengrassGroupId}"
        - CapabilityConfiguration: !Sub >-
              "sources": [
                  "name": "${PlcName}",
                  "endpoint": {
                    "certificateTrust": {
                      "type": "TrustAny"
                    "endpointUri": "opc.tcp://${ PlcIp}:${PlcPort}",
                    "securityPolicy": "NONE",
                    "messageSecurityMode": "NONE",
                    "identityProvider": {
                      "type": "Anonymous"
                    "nodeFilterRules": [
                        "action": "INCLUDE",
                        "definition": {
                            "type": "OpcUaRootPath",
                            "rootPath": ""
                  "measurementDataStreamPrefix": "/${PlcPrefix}"
          CapabilityNamespace: iotsitewise:opcuacollector:1

Alternately from the AWS console you can navigate to AWS IoT SiteWise > Gateways > Create Gateway to configure and deploy the AWS IoT SiteWise connector to the required AWS IoT Greengrass group device.

In addition to ingesting data directly into AWS IoT SiteWise, the AWS IoT Greengrass industrial protocol connectors can be configured to send the data to AWS IoT Analytics, Amazon S3 and Kinesis Data Streams using Stream Manager.

For example, data from industrial data sources using Stream Manager can be transferred easily and reliably to an IIoT application running in the AWS cloud. This is done by configuring the data streams from Kinesis Data Streams to an Amazon TimeStream database using Amazon Kinesis Data Analytics for Apache Flink.


By providing industrial connectivity, data ingestion and storing of data from industrial data sources such as PLCs in AWS IoT SiteWise, Amazon S3, AWS IoT Analytics, or Amazon Timestream the challenge of getting access to isolated data as part of architecting a smart factory is made simpler.

From here, the data is analyzed to optimize production assets or to identify anomalies using AWS for Industrial service and Amazon Lookout for Equipment as part of a predictive maintenance application.

Stay tuned for more blogs in this space, also, don’t forget to check out the Architecture Monthly magazine on Manufacturing.

Field Notes provides hands-on technical guidance from AWS Solutions Architects, consultants, and technical account managers, based on their experiences in the field solving real-world business problems for customers.





Steve Blackwell

Steve Blackwell

Steve Blackwell is the Worldwide Tech Leader for Manufacturing at Amazon Web Services with over 20 years of global experience in the industry, with roles across manufacturing in operations and IT. In his role at AWS he leads the Manufacturing Technical Community and defines the technical strategy and solutions for manufacturing working with customers, partners and the AWS service teams. Steve has worked in the Aerospace, Automotive, CPG, High-Tech, Pharmaceutical and Industrial segments, a trained 6-Sigma and Lean practitioner.