Node.js for Industrial Backend APIs: Architecture Patterns That Scale

Why Node.js for Industrial APIs

Industrial IoT systems have a specific performance profile: many concurrent connections (one per machine or sensor), high throughput of small messages, and mixed workloads (real-time streaming + batch queries). Node.js’s event-driven, non-blocking I/O is a natural fit. A single Node.js process can handle 10,000+ concurrent WebSocket connections — easily covering a 500-machine factory.

The Three-Layer Architecture

Layer 1: Transport — WebSocket server (for real-time machine data), REST API (for web dashboard CRUD), and MQTT subscriber (for sensor data). Layer 2: Business logic — data validation, engineering unit conversion, OEE calculation, alarm threshold checking. Layer 3: Storage — TimescaleDB for time-series machine data, MySQL/PostgreSQL for relational production data, Redis for real-time state cache.

Handling High-Frequency Machine Data

A PLC sending OEE data at 1Hz generates 86,400 records/day per machine. With 50 machines, that is 4.3 million records/day. Use time-series aggregation at write time — store raw data for 7 days, hourly aggregates for 90 days, daily aggregates permanently.

Error Handling and Resilience

Factory APIs must be resilient. Use a message queue between your MQTT receiver and your database writer — if the database is temporarily unavailable, messages queue rather than being lost. Implement circuit breakers for external API calls. Write all errors to a structured log with machine ID, timestamp, and error context.