A manufacturing operation is full of specialized services. From low-level machine data to massive data warehouses, the digital side of an operation requires integration with many different applications.
As a manufacturing data hub, one of Rhize’s key responsibilities is to coordinate, transform, and store data as it passes between systems. To meet this responsibility, Rhize has a number of built-in integrations with the most common apps and protocols used by manufacturing. While we’ve already written about the data that goes into a data hub, this article highlights some important built-in integrations.
Event data
The Rhize agent has built-in connectors to send and receive messages over various protocols:
-
- OPCUA
-
- MQTT
-
- Kafka
-
- Azure Event Hub and Azure Service Bus
All these data sources can also be connected to your plant equipment and equipment classes to trigger rules. This scopes the event data to a particular context. You can also use the data source read and write tasks to incorporate them into your BPMN workflows.
OPC UA
OPC UA is a well-known protocol to exchange data from sensor applications.
Rhize’s OPC UA integration binds an OPC UA source to the modeled equipment, and it has additional configuration for OPC UA methods.
MQTT
MQTT is a popular alternative to OPC UA, known for its flexibility and simple protocol.
In addition to configuring an MQTT data source, you use MQTT to publish and subscribe to messages from Rhize’s broker. This video shows an example of Rhize handling MQTT messages to send alerts.
Kafka
Kafka is a well-known open-source platform and protocol for event streaming. While it’s widely used across a number of fields, Kafka has become a popular in the manufacturing space for its IoT streaming capabilities.
In addition to connected data sources, Rhize’s BPMN workflow editor has a Kafka service task to handle Kafka messages.
Rhize also connects with the Kafka-compatible Red Panda.
Azure Event Hub and Service Bus
Azure provides different services to deliver events and messages. The Event Hub is for big data and streaming, and the service bus is for high-value transactional messages.
Rhize can configure both as a data source. If you use Azure, Rhize can become a layer to pass your application stream and transactional data through a standard layer.
HTTP services
HTTP is the lifeblood of the internet. In manufacturing operations, a variety of level-3 and level-4 applications, like ERPs and WMS systems, often expose their interfaces through REST APIs.
Rhize can interact with any application that you have behind a REST API or GraphQL API. In your workflows, the REST API service task can use all the standard verbs to read and write data from an HTTP API.
This video provides an example of using Rhize to integrate with event data from an external API and transform the response into a standard ISA-95 representation.
PostgreSQL
PostgreSQL is the juggernaut of open-source relational databases.
The Rhize PostgresQL service task can send queries to a PostgresQL database.
Analytics and downstream
In addition to processing, storing, and passing data to and from external systems, Rhize is also a data source for downstream monitoring and analytics applications.
Here are some services that Rhize integrates with.
Google BigQuery
Google BigQuery is Google’s managed data warehouse platform. In contrast to the row-oriented database of PostgresQL, optimized for large amounts of writes, BigQuery is a column-oriented database, designed for frequent reads and aggregations.
One significant advantage of Rhize is that it provides an upstream semantic layer to clean and transform messages from different manufacturing applications. As we wrote in The Data Scientist’s View, since this data is well-modeled operational data, Rhize makes an ideal data source for your manufacturing data-science and machine-learning teams.
The Rhize BigQuery service tasks provides a way to make ELT processes to BigQuery a part of your automated workflow.
Seeq
Seeq is an analytics and data visualization platform for real-time and time-series data. With powerful means to run statistical calculations on a slice of time-series data, Seeq is often the first choice for manufacturers for tools like statistical-process controls.
The Rhize Seeq Connector provides a way for you to query the data in your knowledge graph directly from Seeq.
Since Rhize models data with standard semantics, you can use different Rhize entities to group and aggregate data, including scheduling and response data at various granularities.
Grafana and other observability tools
Rhize runs in containers and so its applications benefit from observability. We use Grafana as our default dashboard and trace, but Rhize also has integrated with to DataDog. For debugging and visibility, Rhize also has a logging task that logs messages to the pod where a workflow is running.
The hub and the semantics
As you’ve seen from this tour of integrations, Rhize handles data from many sources, and is a data source itself for analytics and observability.
What we haven’t emphasized is that at the core of Rhize is a powerful knowledge graph to represent every event, definition, resource, and schedule in a manufacturing operation. Rhize represents data at all granularities, from individual measurements to high-level schedules.
By connecting your data with Rhize, you thus link your applications in one knowledge graph. Think of Rhize as the central nervous system of your manufacturing operation, passing signals along, and storing the important stuff in a powerful centralized intelligence.