Enhancing property prediction and process optimization in building materials through machine learning: A review
Konstantinos I. Stergiou, Charis Ntakolia, Paris Varytis, Elias P. Koumoulos, Patrik Karlsson, Serafeim Moustakidis
Abstract
Analysis and design, as the most critical components in material science, require a highly rigorous approach to assure long-term success. Due to a recent increase in the amount of available experimental data, large databases now contain a depth of knowledge on important properties of materials. The use of this information, combined with Machine Learning (ML) solutions, can enhance the materials’ manufacturing process and efficiency. Indeed, ML can predict material properties, minimize the time and cost of laboratory testing, as well as optimize critical manufacturing processes. This paper aims to give an up-to-date review of the literature on how ML models are used to predict buildings’ material properties (thermal, mechanical, and optical) and optimize the production lines for: a) Phase Change Materials (PCMs), b) Thermoelectric generators (TEGs), c) Customizable 3D-components, d) Advanced cement/concrete-based materials, e) Aerogels, f) Insulation components made from waste materials, g) Multifunctional component materials (MCs), h) Solar active building envelopes (SAE), i) Omniphobic coatings. The review showed that ML-driven approaches for materials’ properties prediction in buildings and process optimization have grown rapidly, providing information and insights that can be utilized in the industry to maximize the materials’ production and efficiency while reducing CO2 emissions, resulting in a more productive and environmentally friendly era.