Insight into production through smart sensor setup
The students placed two types of sensors on the Melzer punching machine: a rotary encoder to count rotations (for run/downtime and speed) and a distance sensor to measure the diameter of the plastic mother roller (for material consumption).
Using an Arduino and a Raspberry Pi, measurements are collected in real time, filtered locally and automatically transmitted as .csv files to a remote server. Custom 3D designs were realised for the hardware enclosure. Thanks to this setup, there is now a scalable measurement system for production analysis, built entirely with affordable components.
TinyML enables real-time data analysis and predictions ‘on the edge’ (i.e. on the machine) without expensive cloud solutions. Your machines can also work smarter, without heavy IT investments.
what was the result after the deployment of tinyml?
Insight into run/downtime and production speed, real-time monitoring of material consumption and automatic data collection without operator intervention. A good mouthful, right?
First step towards data-driven process optimisationWith the data collected, production efficiency is improved. Material waste is also reduced and faults are spotted earlier. The setup is scalable to other machines and forms the basis for future TinyML applications such as fault detection and predictive maintenance.
more info or register?
Contact the Fieldlab Edge AI for Smart Industry at Andre Gerver and find out what we can do for your production process.
CONTACT
Andre Gerver
📧 a.gerver@techport.nl