Edge Industrial Vision & Telemetry
Production-line monitoring that combines computer vision, edge inference and telemetry, with industrial control over Modbus RTU / RS485.
Overview
A pipeline concept and prototype for factory monitoring: cameras and edge inference detect production events, telemetry is streamed, and results are integrated with enterprise software. Related hardware work covers PLC-style control over RS485.
Problem
Manual counting and inspection on a production line is slow and error-prone, and the resulting data rarely reaches enterprise systems in a usable form.
Solution
Run vision models at the edge to detect and count events, publish telemetry over lightweight protocols, and integrate with backend systems. Control hardware is driven through Modbus RTU over RS485.
Architecture
- 01
Capture
- Cameras
- Production line sensors
- 02
Edge inference
- Computer vision
- Counting / defect detection
- 03
Telemetry
- MQTT
- Industrial telemetry
- 04
Control
- ESP32 / M5Stack / StampPLC
- RS485 + Modbus RTU
- Relay module
- 05
Integration
- Backend APIs
- SQL database
- Enterprise systems
- Target applications: cement bag counting, defect detection, production monitoring and industrial event detection.
- Edge-first inference keeps latency low and reduces the need to stream raw video.
- Control path: microcontroller, RS485, Modbus RTU, R421C32 32-channel relay.
Key features
- Vision-based counting and defect detection concepts
- MQTT telemetry from the line to enterprise systems
- Modbus RTU relay control over RS485
Engineering challenges
- High-speed counting where objects move quickly past the camera.
- Running reliably on constrained edge hardware.
- Bridging industrial protocols and modern enterprise APIs.
Results
No measured outcomes are published for this project. I only list results that have been verified.