Using a form of unsupervised brain-like AI called Deep Learning they’ve increased the speed of metamaterial identification and learned how to better optimize them for interaction with different elements.
The inter-disciplinary team of scientists from Israels Tel Aviv University published their research in the journal Light: Science and Applications.
Metamaterials are a kind of microscopic man-made crystal whose variations can be so information dense and complex that at times it is difficult to categorize or define them. By leveraging AI scientists believe they can automate time consuming repetitive aspects of identification and optimization.
It turns out that when you arrange certain particles in to very specific geometric formations in relationship to one another the entire material transforms at a larger scale. Unique functions arise from finding one of those sweet spots including a significant change to the materials level of conductivity, refractive index, density and even what frequencies it can absorb and reflect.
Metamaterials facilitate a lot of diverse applications including nano-scale memory storage, optical technology, functional invisibility cloaks, and better solar power generation/storage.
A team lead by Dr Haim Suchowski and Professor Liof Wolf from Tel Aviv focused on nanophotonic metamaterial elements in their particular experiment
“The process of designing metamaterials consists of carving nanoscale elements with a precise electromagnetic response,” Dr. Mrejen says. “But because of the complexity of the physics involved, the design, fabrication and characterization processes of these elements require a huge amount of trial and error, dramatically limiting their applications.”
Since meta-material construction and categorization require so much work a large margin of error pulls in to question the long term feasibility of affordable metamaterial products. So with that being said, scientists here at Tel Aviv insist that artificial intelligence, and in particular – Deep learning are vital to the continued existence of the meta-material industry.
“Our new approach depends almost entirely on Deep Learning, a computer network inspired by the layered and hierarchical architecture of the human brain,” Prof. Wolf explains. “It’s one of the most advanced forms of machine learning, responsible for major advances in technology, including speech recognition, translation and image processing. We thought it would be the right approach for designing nanophotonic, metamaterial elements.”
In this experiment their deep learning network is supplied with a total of 15 000 artificial experiments that are supposed to serve as an example for the machine to draw conclusions about the relationship between nano-element shape and function as far as the meta materials electromagnetic behavior is concerned.
“We demonstrated that a ‘trained’ Deep Learning network can predict, in a split second, the geometry of a fabricated nanostructure,” Dr. Suchowski says.
The algorithm also generates a whole swathe of specialized designs for nanoelement structure so that each formation is better suited to specific chemicals or proteins.
“These results are broadly applicable to so many fields, including spectroscopy and targeted therapy, i.e., the efficient and quick design of nanoparticles capable of targeting malicious proteins,” says Dr. Suchowski. “For the first time, a novel Deep Neural Network, trained with thousands of synthetic experiments, was not only able to determine the dimensions of nanosized objects but was also capable of allowing the rapid design and characterization of metasurface-based optical elements for targeted chemicals and biomolecules.
“Our solution also works the other way around. Once a shape is fabricated, it usually takes expensive equipment and time to determine the precise shape that has actually been fabricated. Our computer-based solution does that in a split second based on a simple transmission measurement.”
The team at Tel Aviv has filed a patent for their technology and next they plan to explore the chemical structure of nanoparticles