Apache StreamPipes

The typical open-source IIoT stack — Mosquitto + InfluxDB + Node-RED + Grafana (the "MING stack") — works great as a prototype. But scaling it to production means managing four separate services, wiring up SSO, coordinating upgrades, and hoping your Node-RED flows don't become unmaintainable spaghetti. StreamPipes is the answer to "what if all of that was one product, and my plant operators could use it without calling IT?"

It's the only open-source IIoT analytics platform specifically designed for self-service use by non-technical manufacturing staff. Plant operators, process engineers, and quality managers drag and drop data pipelines instead of writing code.

What it does

Connect — StreamPipes Connect provides adapters for 20+ industrial protocols out of the box: OPC-UA, S7 (via Apache PLC4X), MQTT, Modbus TCP, REST, ROS, plus message brokers like Apache Kafka and Apache Pulsar. New adapters can be added at runtime without restarting the system.

Analyze — 100+ built-in processing algorithms: trend detection, peak detection, anomaly detection, numerical filters, sequence detection, frequency calculation, and pre-trained neural networks. The Python client integrates with OnlineML/River for custom ML models.

Explore — Live dashboards, time-series data explorer with heatmaps, shopfloor monitoring views, and asset organization. Data sinks to PostgreSQL, Elasticsearch, Apache IoTDB, Apache CouchDB, plus Kafka/Pulsar for downstream consumers.

Deployment

Pipeline elements are standalone microservices. Run processing centrally on a server or at the edge close to sensors — or mix both.

Developer SDKs

• Java SDK — Build custom adapters, processors, and sinks
• Python Client — Programmatic pipeline interaction, data extraction to Pandas DataFrames, OnlineML/River ML integration
• TypeScript Library — Custom UI components via microfrontend framework

Technical specs

• Backend: Java 17 / Frontend: Angular
• Build: Maven 3.8+, Node.js v12+
• License: Apache License 2.0
• Current version: 0.98.0 (not yet 1.0 — keep this in mind for production bets)
• GitHub: 716 stars, 226 forks, 114 contributors
• Backed by: FZI Research Center for Information Technology (Karlsruhe, Germany), graduated to Apache Top-Level Project

Limitations

• Pre-1.0 — API and data model may still change between releases. Migration tooling exists but isn't guaranteed smooth.
• Smaller community than Node-RED — 716 stars vs Node-RED's 22K+. Fewer third-party resources, tutorials, and Stack Overflow answers.
• Limited non-IIoT use — Purpose-built for manufacturing. If you need general-purpose data integration, NiFi or Node-RED is a better fit.
• Java-heavy — Backend is Java 17. If your ops team is Python/Go-native, expect a learning curve for custom extensions.
• No commercial support — Pure open source with no vendor offering enterprise support contracts (unlike Node-RED with FlowFuse, or NiFi with Cloudera).