An investigation of the utilisation of different data sources in manufacturing with application in injection moulding

Abstract

This work focuses on the effective utilisation of varying data sources in injection moulding for process improvement through a close collaboration with an industrial partner. The aim is to improve productivity in an injection moulding process consisting of more than 100 injection moulding machines. It has been identified that predicting quality through Machine Process Data is the key to increase productivity by reducing scrap. The scope of this work is to investigate whether a sufficient prediction accuracy (less than 10% of the specification spread) can be achieved by using readily available Machine Process Data or additional sensor signals obtained at a higher cost are needed. The latter comprises Machine Profile and Cavity Profile Data. One of the conclusions is that the available Machine Process Data does not capture the variation in the raw material that impacts element quality and therefore fails to meet the required prediction accuracy. Utilising Machine Profiles or Cavity Profiles have shown similar results in reducing the prediction error. Since the cost of implementing cavity sensors in the entire production is higher than utilising the Machine Profiles, further exploration around improving the utilisation of Machine Profile Data in a setting where process variation and labelled data are limited is proposed.

Keywords:

• Industry 4.0
• data evaluation
• machine learning
• deep learning
• predictive quality
• injection moulding

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 PRIAMUS System Technologies is a manufacturer of cavity sensors.

2 New supply of material is added in the top of the silo and material used in the production is emptied out though the bottom. This means the materials from the two suppliers are layered in the silo.

3 This can be explained by small changes in viscosity of the melt depending on material properties.

Additional information

Notes on contributors

G. Ø. Rønsch

Georg Ø. Rønsch research is mainly focused on efficient utilisation of data from production equipment with the aim of improving production productivity. The work is centred around injection moulding, but the approaches can be utilised in other industries. The interest for this topic has been matured through 15 years of work within production optimisation in different industries. Currently he is enrolled in an industrial PhD project titled ‘Model-based Process Surveillance and Optimisation for Fault Detection and Diagnosis’, where effective use of techniques within chemometrics, machine learning, and deep learning is explored.

M. Kulahci

Murat Kulahci's research currently focuses primarily on large data analytics for descriptive, inferential and predictive purposes. Many of his research applications involve high dimensional, high frequency data demanding analysis methods in chemometrics and machine learning. He has been collaborating with various industries in many industrial statistics projects and digital manufacturing. He has more than 100 scientific publications in refereed international journals and conference proceedings and three books published with co-authors from Europe and the U.S.A.

M. Dybdahl

Martin Dybdahl is concerned with the practical application of data analytics grounded primarily in the field of manufacturing. With a background in both engineering and business development, the interests have consolidated to focus on making use of data analytics and machine learning to improve manufacturing processes and other business critical processes.