The Connected Factory: Real-Time Intelligence from Every Workstation

Overview

A global manufacturer partnered with Postindustria to modernize production oversight – connecting the precision of machines with the flexibility of human craftsmanship.
The result: a unified operational data platform that automatically collects, analyzes, and links data from video streams, sensors, IoT devices, and digital work orders into one integrated system.

By merging camera-based motion analysis, IoT data ingestion, and automated content generation, the company gained real-time visibility across its manufacturing stages – from initial assembly to final packaging.
The platform not only streamlines quality control but also transforms raw operational data into a foundation for continuous process improvement.

Key Highlights

Unified operational data platform combining video streams, sensor data, and order metadata

120+ cameras and IoT devices continuously collecting high-resolution production data

QR-based order recognition connecting each video segment to a specific work order

Computer vision and AI analysis detecting motion, pauses, and stage transitions

Automated video clipping for active work periods, trimming idle time to seconds

Automated media generation for reporting and operational summaries

Scalable, modular architecture integrating Apache NiFi, Kafka, and AWS S3

The Challenge

Before implementation, the company’s production data lived in silos:

Sensors and IoT devices provided readings but weren’t linked to specific work orders.
Video streams from cameras existed only as raw footage – unstructured and unsearchable.
Manual assembly stages lacked digital traceability or real-time monitoring.
Reporting and content generation required extensive manual work.

The goal was to create a single data environment where information from devices, workstations, and operators could flow seamlessly – providing traceable insights into every step of the production process while preserving existing workflows.

The Journey

Postindustria developed a data-driven architecture capable of handling large-scale video ingestion, motion detection, and multi-device synchronization.

Data Ingestion Layer

Apache NiFi ingested RTSP streams from cameras and readings from IoT devices, including precision scales, and packaged them into structured events.

Video Analysis

Python-based motion detection (OpenCV) identified when work began and ended at each workstation, trimming pauses longer than 10 seconds.

Order Recognition

Each workstation displayed a QR-coded worksheet, allowing the system to associate activity with the correct order and stage.

Video Clipping & Metadata

Active work intervals were automatically clipped, enriched with metadata (order number, stage, duration, idle time, etc.), and stored in PostgreSQL.

Automated Media Generation

Selected production stages were automatically compiled into summary videos and uploaded to AWS S3 for reporting and review.

This architecture turned video, IoT, and order data into a single source of operational truth – a foundation for analytics, quality assurance, and future automation.

Technology in Action

Data Ingestion & Orchestration: Apache NiFi
Stream Processing: Faust (Python)
Event Storage: Apache Kafka
Computer Vision & AI: Python, OpenCV, custom motion detection, and event tagging
Storage & Infrastructure: PostgreSQL, AWS S3, and local file system for midterm storage
Video Tools: FFmpeg and GStreamer for stream capture and clipping
IoT Integration: Precision scales and other factory sensors feed real-time measurements into the unified data pipeline for synchronization with video and workflow data.
Monitoring & Visualization: Grafana and Metabase dashboards for production analytics

The Results

Unified Visibility

Video, IoT, and process data consolidated into one platform

Operational Efficiency

Automated monitoring eliminated manual video editing and data entry

Traceable Production

Each stage now produces verifiable, timestamped evidence

Quality Assurance

Motion-based analytics highlight idle time and performance patterns

Process Transparency

Real-time metrics improved collaboration across technical and managerial teams

Scalability

The platform supports new devices, data sources, and analytical modules with minimal setup

Conclusion

By integrating video analytics, IoT data, and workflow automation, the manufacturer achieved full visibility across its operations – from manual workstations to automated processes.

Postindustria’s unified data platform transformed production oversight into a connected, measurable, and continuously improving system – setting a foundation for the next generation of data-driven manufacturing intelligence.