IoT Data Analytics Case Study - Packaging Films Manufacturer
- Analytics & Modeling - Predictive Analytics
- Analytics & Modeling - Process Analytics
- Sensors - Pressure Sensors
- Packaging
- Quality Assurance
- Machine Condition Monitoring
- Predictive Quality Analytics
- Root Cause Analysis & Diagnosis
The company manufactures packaging films on made to order or configure to order basis. Every order has a different set of requirements from the product characteristics perspective and hence requires machine’s settings to be adjusted accordingly. If the film quality does not meet the required standards, the degraded quality impacts customer delivery causes customer dissatisfaction and results in lower margins. The biggest challenge was to identify the real root cause and devise a remedy for that.
In the ‘packaging film’ manufacturing process, the films are finally wound downstream on a ‘winder’ machine and slit to order on a ‘slitting’ machine. The solution demanded an ability to process historical machine data and establish requiredcorrelation between machine settings and production output. The deployed solution helped collect historical performance data. It gave the quality team the ability to monitor critical machine settings parameters, ascertain that they are in statistical control, and eventually correlate the process and machine data with different types of quality failures. The solution (Datonis MInt) also helped to identify critical to quality parameters from thirty parameters to two (pressure and tension) using correlation analytics. A prescriptive quality model was built based on the collected data to recommend the machine settings for the product configuration for every new order. This helped reduce quality degradation and achieve predictable quality.