On November 16, 2018, NECTEC selected the NU Maker Club as a finalist in the Maker Startup 2018 “Smart Factory” Contest. The premises of this event are to encourage innovation, foster IoT deployment, and increase collaboration between developers or makers and factories through the NETPIE platform.
Our group first proposed to use anomaly detection (a machine learning technique) to flag production line bottlenecks. But after a conversation with the representatives from Wuttisak Aesthetic Care CO., LTD., we saw how we could improve their QC system with our expertise in image processing.
During the normal QC process, the workers on the factory floor have to log parameters displayed on a mixing machine. These records are tracked and maintained manually on a piece of paper.
The logs are kept and then verified by the R&D team. The company would like to minimize the time spent during this process.
This is where we came in…
The mixing machine shows the current status on a panel of several seven-segment displays. A worker reads the displays and writes the values onto a process tracking paper. To automatize and digitize the company’s QC process, we propose an IoT camera with onboard image processing to read these measurements from seven-segment displays and organizing the data using the NETPIE IoT platform.
Why don’t we just plug in some sort of cables to read the data through serial connection?
Well, we do not have that option in this factory. And more importantly, we want to be unobtrusive and undisruptive to their normal workflow. Additionally, our proposed setup is easily scalable. We can deploy the computer vision unit anywhere as needed.
Business Related Details:
- Legacy data collection technique is slow/expensive/limited
- Manually/visually/discretely aggregate data (for R&D and QC)
- Manually transcribe data into electronic format
- Reduced human workload (long term cost)
- Remove human error source
- Expands scope (frequency) of data collection
- Initial steps to real-time process control
- Supports future development of real-time process control
- Thailand 4.0 compliant
We believe computer vision and IoT together can solve industrial problems.
Read more about it in our next blog!