Engineered for Primacy Industries by SunBio IT Solutions. Vision cameras capture live conveyor streams processed via Raspberry Pi edge devices and YOLO AI models, streaming telemetry over MQTT brokers to databases and automated daily/weekly production planning dashboards.
Manual candle counting and weight-based estimation are slow, labor-intensive, and prone to human error. SunBio IT Solutions replaces the walk-the-line inspector with rugged edge vision nodes running custom-trained YOLO detection models that inspect every tray instantly.
Operating locally on Raspberry Pi edge hardware with zero cloud round-trip required, captured frames are analyzed in milliseconds. Telemetry data is pushed via lightweight MQTT messaging directly into SQL databases, powering live conveyor monitoring and dynamic daily/weekly manufacturing schedules.
A robust 6-stage industrial IoT pipeline connecting line-side vision capture to enterprise production scheduling.
Industrial cameras continuously record live conveyor belt streams as candle trays pass through the inspection zone.
Compact Raspberry Pi edge devices execute optimized YOLO detection models locally for instantaneous counting checks.
Processed count payloads are published securely over lightweight MQTT messaging protocols across the factory network.
Backend database services ingest incoming MQTT telemetry, instantly recording batch counts, timestamps, and defect logs into SQL tables.
Operators view real-time counts, shift performance metrics, and active product runs on responsive web-based plant dashboards.
System automatically reconciles live output against daily and weekly targets, optimizing machine allocation and shift planning.
Knowing exactly what is running on each conveyor line in real time empowers plant managers to bridge the gap between shop-floor execution and executive planning.
Explore the specialized software and hardware layers powering SunBio's industrial vision ecosystem.
High-resolution industrial cameras equipped with specialized illumination filters capture glare-free imagery across high-speed conveyor belts.
Custom-trained neural networks optimized for edge execution accurately isolate candle bodies and detect surface anomalies.
Lightweight publish-subscribe architecture ensures reliable, low-bandwidth telemetry transmission from factory floor nodes to central servers.
Robust relational database schemas store historical shift data, defect logs, and real-time inventory counts for deep enterprise analytics.
Intelligent scheduling modules integrate live telemetry with daily and weekly work orders to streamline plant capacity planning.
Direct integration with factory programmable logic controllers enables automated pneumatic rejection of defective trays.
Robust modular features designed to scale across diverse manufacturing and packaging environments.
Advanced object detection replaces manual tallying and weight-based estimation with exact unit counts.
Compact edge devices execute local YOLO inference without relying on heavy cloud infrastructure.
Reliable messaging protocol transmits real-time telemetry from edge vision nodes to backend servers.
Comprehensive logging provides supervisors with granular visibility into line productivity and downtime.
Sync live counts with daily and weekly plant schedules to optimize machine allocation and shift planning.
Local edge processing ensures secure, instantaneous decision-making without external internet dependency.
Partner with SunBio IT Solutions to deploy bespoke industrial computer vision systems tailored to your factory floor.
Book a Free Automation Demo