Development of artificial neural network-based lifetime prediction models for polymers under environmental and industrial conditions

Project summary

Plastics are everywhere in daily life – from cars and buildings to packaging and energy systems. However, estimating the service life of these materials under real-world conditions remains a challenge. Over time, plastic components can slowly deform or crack due to use, temperature changes, or environmental effects such as sunlight, humidity, or chemicals. Current test methods do not accurately reflect what happens to materials after years of use. This can lead to premature failure or over-design, wasting resources and increasing costs. The aim of our project is to solve these problems by developing a modelling framework that can more accurately predict the lifetime of plastics under real-world conditions. We are combining mechanical testing, accelerated ageing, and artificial intelligence (AI) tools to create a methodology that improves as more data becomes available. The model will be adaptable to specific environments, such as marine conditions or areas with high radiation. This will support industries like automotive, construction, and energy storage, where safe, durable, and reliable plastic components are essential.

Project results