Machine learning approach to packaging compatibility testing in the new product development process

Abstract

The paper compares the effectiveness of selected machine learning methods as modelling tools supporting the selection of a packaging type in new product development process. The main goal of the developed model is to reduce the risk of failure in compatibility tests which are preformed to ensure safety, durability, and efficacy of the finished product for the entire period of its shelf life and consumer use. This kind of testing is mandatory inter alia for all aerosol packaging as any mechanical alterations of the packaging can cause the pressurized product to unseal and stop working properly. Moreover, aerosol products are classified as dangerous goods and any leaking of the product or propellent can be a serious hazard to the storage place, environment, and final consumer. Thus, basic compatibility observations of metal aerosol packaging (i.e. general corrosion, pitting corrosion, coating blistering or detinning) and different compatibility factors (e.g. formula ingredients, water contamination, pH, package material and coatings) were discussed. Artificial intelligence methods applied in the design process can reduce the lengthy testing time as well as developing costs and help benefit from the knowledge and experience of technologists stored in historical data in databases.