Skip to content
Proxus 2.0.0 Proxus 2.0: the largest release in Proxus history Release notes
Try free Book a demo

Integrations · Apache Kafka

Plant events on your Kafka topics.

For teams whose data platform is already event-driven: Proxus publishes plant readings, and the alarms and computed values your pipelines send, to a Kafka topic, so industrial data arrives on the same backbone as everything else in the business.

Book a demo All integrations Setup guide
Equipment to Apache Kafka Apache Kafka
Model once, deliver to your systems Equipment sends data through an edge gateway to Unified Namespace, then to Apache Kafka. Controllers S7 / CIP / MELSEC Machines & robots FOCAS / MTConnect Meters & SCADA Modbus / OPC UA Modelled signals 02 / EDGE Proxus Edge Collect / evaluate / buffer 03 / PLATFORM Unified Namespace Modelled industrial signals Apache Kafka Destination 01 / PLANT FLOOR04 / CONSUMERS Model once, deliver to your systems Read from bottom to top. Field equipment is at the bottom and data travels upward. Equipment sends data through an edge gateway to Unified Namespace, then to Apache Kafka. 01 Plant floor PLCs / CNCs / robots 02 Proxus Edge Collect / evaluate /buffer 03 Unified Namespace Modelled industrialsignals 04 Apache Kafka Destination
Publishes
Readings, plus whatever your pipelines and rules send (alarms, computed values)
Volume control
On Change sends only values that changed; Pipelines Only sends only what a pipeline selects
Reliability
Buffered at the edge when brokers are unreachable
Message
Key = tag name, value = reading; device name, device ID and read time in headers

Data path

What actually moves, and where it stops.

Published on change, or on every read, per target

Model once, deliver to your systems Equipment sends data through an edge gateway to Unified Namespace, then to Apache Kafka. Controllers S7 / CIP / MELSEC Machines & robots FOCAS / MTConnect Meters & SCADA Modbus / OPC UA Modelled signals 02 / EDGE Proxus Edge Collect / evaluate / buffer 03 / PLATFORM Unified Namespace Modelled industrial signals Apache Kafka Destination 01 / PLANT FLOOR04 / CONSUMERS Model once, deliver to your systems Read from bottom to top. Field equipment is at the bottom and data travels upward. Equipment sends data through an edge gateway to Unified Namespace, then to Apache Kafka. 01 Plant floor PLCs / CNCs / robots 02 Proxus Edge Collect / evaluate /buffer 03 Unified Namespace Modelled industrialsignals 04 Apache Kafka Destination

Field

  • PLCs, CNCs, robots
  • OPC UA servers
  • Meters and SCADA

Proxus

  • Collect at the line
  • Model into the namespace
  • Buffer and forward

Apache Kafka

Apache Kafka
  • topic and partition
  • consumer groups
  • stream processors

How it works

From the machine to Apache Kafka, in four steps.

  1. 01

    Collect and name

    Signals are read from equipment, scaled and given tag names.

  2. 02

    Decide what is an event

    Rules and anomaly detection at the edge turn raw changes into events, and a pipeline sends them to the target.

  3. 03

    Publish to a topic

    An outbound target writes to the topic you configure, with the strategy you choose.

  4. 04

    Consume anywhere

    Existing Kafka consumers read industrial data the same way they read everything else.

What you get

Why teams route Apache Kafka through Proxus.

  • One backbone

    Industrial data stops being the exception that needs its own transport.

  • Keyed by tag name

    Each message’s key is the tag name, so partitioning and log compaction work per tag, and consumers filter by tag without parsing the value.

  • Read time preserved

    Buffered values keep the time they were read, in the Timestamp header, so late messages still line up.

  • Topic checked first

    Test Connection asks the cluster for its brokers and whether the topic exists, before anything is published.

Where it is used

Typical deployments.

  • Event-driven MES and scheduling

    React to machine events as they happen rather than polling a database.

  • Streaming analytics

    Feed stream processors with industrial readings and events.

  • Fan-out to many consumers

    One publication, many independent readers, without extra load on the equipment.

Targets
Apache Kafka topics; RabbitMQ, MQTT and Sparkplug B are also supported
Payload
One message per value: key = tag name, value = the reading as text; headers DeviceName, DeviceId and Timestamp (the read time)
Strategies
Pass Through, On Change, or Pipelines Only (only what a pipeline sends)
Security
SASL and TLS (Ssl or SaslSsl); the broker’s host name is verified
You configure
BootstrapServers, Topic, ClientId; SecurityProtocol, SaslMechanism, SaslUsername, SaslPassword. Test Connection checks the address and credentials from the hub or from a gateway you choose, without writing data.
Setup guide All target parameters

Proven in production

Running in production at these plants.

All deployments

Questions

Before you wire it up.

Does Proxus replace our broker?
No. Proxus publishes into your existing Kafka cluster as a target; broker operations stay with your platform team.
What is in a message?
One value per message: the tag name as the key and the reading as the value, with the device name, device ID and the time it was read in the headers DeviceName, DeviceId and Timestamp.
What happens if the cluster is unavailable?
The gateway buffers locally and publishes when brokers are reachable again.
Can we publish events rather than every reading?
Yes. Rules and anomaly detection at the edge turn raw changes into events; set the target to Pipelines Only and only what your pipelines send is published. On Change avoids republishing unchanged values.

Other integrations

Next step

See it delivering into your own Apache Kafka.

Bring one machine and a Kafka topic. We connect the machine, publish its readings to the topic in front of you, and show what the gateway does while the cluster is down.

Book a demo Talk to an engineer
  1. 01

    Scoping call

    Equipment, protocols, sites and the systems the data has to reach.

  2. 02

    Technical walkthrough

    A working demo against a scenario that resembles your plant, not a generic tour.

  3. 03

    Pilot plan

    A defined first line or site, with the deployment and security questions answered up front.

  4. 04

    Roll-out

    The same templates bring the next lines and sites online.