Cracked phone screens and shattered windscreens may soon be easier to address with researchers at IIT-Delhi having developed a first-of-its-kind machine-learning software for “predicting and optimising glass compositions”. The software, Python for Glass Genomics (PyGGi), has been made to enable researchers and companies to “easily predict glasses with superior properties like scratch resistance and crack resistance”.
One of the project investigators (PI), Professor N M Anoop Krishnan from the Civil Engineering Department, said, “Understanding and predicting the composition-structure-property relationship is key… Through this software, once it is determined if the glass is breakable and scratch-free or not, it can be used for things like phones, car screens, window panes etc. This will then help in designing unbreakable glass or bulletproof glass. This is a first-of-its-kind software in the world,” he said.
Professor Hariprasad Kodamana from the Mechanical Engineering Department, who is the second PI, said, “PyGGi will be constantly updated to meet all challenges in the field.”