October_2022_AMP_Digital

A D V A N C E D M A T E R I A L S & P R O C E S S E S | O C T O B E R 2 0 2 2 5 ALGORITHM FIXES 3D PRINTING ERRORS Engineers at the University of Cambridge, U.K., developed a machine learning algorithm that can find and fix a wide range of different 3D printing errors in real time, and can be easily added to new or existing machines to enhance their capabilities. 3D printers using the algorithm could also learn how to print new materials by themselves. Researchers have been working on automated 3D printing monitoring, but existing systems can only detect a limited range of errors in one part, one material, and one printing system. The Cambridge team trained a deep learning computer vision model by showing it nearly one million images captured automatically during the production of 192 printed objects. Each image was labeled with the printer’s settings, such as the speed and temperature of the printing nozzle and flow rate of the printing material. The model also received information about how far those settings were from good values, enabling the algorithm to learn how errors arise. Using this approach, the team was able to build an algorithm that is generalizable and can be applied to identify and correct errors RESEARCH TRACKS in unfamiliar objects or materials, or even in new printing systems. In the future, the trained algorithm could be more efficient and reliable than a human operator at spotting errors, say researchers. With the support of Cambridge Enterprise, the university’s commercialization department, researcher Douglas Brion formed Matta, a spin-off company that will develop the technology for commercial applications. www.matta.ai. INSIGHT INTO DISORDERED MATERIALS Researchers at the National University of Singapore (NUS) developed a human-explainable machine learning system that quickly identifies previously unseen novel structures in disordered materials without help from humans. The team created a machine learning framework that can learn the universal vocabulary and grammar used to describe disordered systems. Using this framework, they discovered that a wide range of disordered materials can be logically decomposed into recurring motifs and related compositional rules. These motifs are the building blocks Example image of the 3D printer nozzle used by the machine learning algorithm to detect and correct errors in real time. Courtesy of Douglas Brion. that can vastly simplify how to understand and classify complex disorders in real materials. The scientists used a sequence of mathematical expressions known as Zernike polynomials to quantify the subtle structural and chemical features within atomic arrangements. These special mathematical expressions can effectively model the features despite different atomic orientations. To overcome the limited signal fromeach atom, the team generalized a single-particle imaging approach that automatically reveals distinct building blocks (i.e., motifs) within disordered materials. Having learned the motifs from tens of thousands of atoms in an automated manner, the team could now discover how these motifs self-assemble into complex but disordered hierarchies. They found that some disordered materials can be described by just a handful of motifs, yet these few motifs create diverse structures due to complex motif-motif hierarchies. Other materials start with a continuous range of motifs, thereby blurring the boundaries between their motifs and hierarchies. The team hopes to turn this framework into a companion artificial intelligence application for microscopes to rapidly make sense of disordered materials. www.science.nus.edu.sg. Fundamental pentagonal motifs follow a three-level hierarchy to form increasingly complex larger motifs. Courtesy of Science Advances.

RkJQdWJsaXNoZXIy MTYyMzk3NQ==