Polyolefins (POs) are among the most widely used plastics thanks to their versatility, durability and low cost. However, their extensive use has resulted in a large amount of polymer waste, so efficient recycling methods are essential. A key challenge in PO recycling is the presence of multi-material systems, which complicates mechanical recycling. Overcoming this challenge is vital for the plastics industry to achieve its sustainability goals. This research project explores the use of artificial intelligence (AI) and scientific machine learning (SciML) to accelerate the microstructural and rheological characterisation of polyolefins. Deep learning models will extract key structural features of polymer droplets, with a focus on mono-material POs under different conditions. High-resolution video analysis of droplet deformation has enabled the development of an advanced algorithm for the precise detection of droplet contours, extracting morphological features such as height, shape and deformation time. Initially, the approach will focus on mono-material systems before gradually extending to complex multi-material PO systems. The ultimate goal is to optimize recycling methods, making them more efficient and scalable while ensuring the reuse of high-quality polyolefins.