Industrial IoT Data Collection Platform: Neuron v2.0 Architecture, Design, and Case Studies

The presentation introduces EMQ's industrial IoT data collection platform Neuron, originally built on MQTT and a stream‑processing database to provide end‑to‑end connectivity, mobility, processing, storage, and analytics from edge to cloud.

Version 1.x revealed several practical challenges: massive device connection, limited SCADA feature usage, access‑control constraints, high‑throughput requirements, and the need for multi‑instance Docker deployments.

To address these, Neuron v2.0 was redesigned with a lighter focus on IIoT device connectivity, abandoning SCADA‑specific functions, supporting multiple protocols on a single instance, extending north‑bound capabilities (EdgeX Foundry, Sparkplug B, eKuiper), and offering an open‑source core with commercial licensing.

The new architecture adopts the NNG basic message library for high‑efficiency, lock‑free message routing using a pair‑based bus topology, and a plug‑in system where each plug‑in (or adapter) runs in its own thread, enabling dynamic loading of .so modules for protocols, drivers, or custom services.

Data‑tag management groups tags by read interval, allowing north‑bound applications to subscribe to tag groups, while the manager routes messages based on tag configurations, ensuring low‑latency data flow from devices to applications.

Integration with eKuiper provides edge‑side stream processing using SQL, converting Neuron’s unified data type to JSON for downstream consumption, and storing processed data locally before optional persistence to databases.

Two case studies illustrate the platform’s applicability: (1) an oil‑field data‑collection solution that aggregates Modbus‑RTU/TCP data via MQTT to a central platform, enabling real‑time alerts and remote control; (2) a smart‑factory scenario where Neuron collects PLC data at sub‑100 ms intervals, feeds eKuiper for filtering, and pushes results to EMQ X Enterprise for storage and analytics.

The article concludes with a call to follow the public account for free data‑intelligence resources and invites further engagement.