Extruder

IoT improves plastic extrusion by recording all process parameters in high resolution and deriving the best settings for each product. At mk Plast, the maintenance team installed the ENLYZE SPARK during operation in around two hours. The production history yields recommendations that lead to stable processes regardless of individual operators' experience. In the first month, output rose by 4.8 percent, equivalent to a time saving of around 25 hours.

The melt temperature can be approximated from external measurements using a machine learning model instead of placing a sensor in the melt. At the end of the twin screw extruder, the melt flows through a channel 10 centimetres in diameter, and temperature is only measured at the wall. In the centre it can be up to 50 kelvin hotter. SKZ's model runs on Cybus Connectware and gives recommendations on feed temperature via dashboards.

Real-time data help detect problems in blown film extrusion immediately, even when nobody is standing at the line. Storopack produces around the clock, with its extrusion lines spread across two halls several hundred metres apart. In the past, the night shift lacked the data needed to get support. With ENLYZE Shop Floor BI, production, quality and production support see line parameters together with order context and recognise connections faster.

Condition data show how heavily the screw is loaded and help optimise its service life. Rosendahl Nextrom collects tension, temperature and object data in the extruder section of its lines via OPC UA for this purpose. Because the information is described semantically, conversions such as Celsius to Fahrenheit are no longer needed. Historical data and experience feed into analyses for predictive maintenance on each line, in future also using artificial intelligence.