The computer architecture that makes generative ‘AI’ systems such as ChatGPT possible has been modified to develop more effective solar cells or even ‘smart’ windows.
It’s the optical properties of the materials involved that are key to such improved designs, according to the scientists at Michigan University, and that’s where the co-called OptoGPT comes in.
In essence, it works backwards from the desired optical properties – be they for telescopes, windows, or solar cells – to come up with the recipes and material structures requited to achieve them.
The phenomenal computing power needed to achieve this borrow – and adapts – from the large language model (LLM) techniques that underpin genAI systems.
According to the Michigan team, the new algorithm designs optical multilayer film structures – stacked thin layers of different materials – that can serve a variety of purposes. Well-designed multilayer structures can maximize light absorption in a solar cell or optimize reflection in a telescope.
They can also improve semiconductor manufacturing with extreme UV light, and make buildings better at regulating heat with smart windows that become more transparent or more reflective depending on temperature.
OptoGPT is said to produce designs for multilayer film structures within 0.1 seconds, almost instantaneously. In addition, its designs contain six fewer layers on average compared to previous models, meaning its designs are easier to manufacture.
L Jay Guo, professor of electrical and computer engineering, was the corresponding author of the study, which was published in the journal Opto-Electronic Advances.
“Designing these structures usually requires extensive training and expertise as identifying the best combination of materials, and the thickness of each layer, is not an easy task,” he said.
To automate the design process for optical structures, the research team tailored a transformer architecture – the machine learning framework used in large language models like OpenAI’s ChatGPT and Google’s Bard – for their own purposes.
“In a sense, we created artificial sentences to fit the existing model structure,” Guo added.
Words make light work
The model treats materials at a certain thickness as words, also encoding their associated optical properties as inputs. Seeking out correlations between these ‘words’, the model predicts the next word to create a ‘phrase’ – in this case a design for an optical multilayer film structure – that achieves the desired property such as high reflection.
Researchers then tested the new model’s performance using a validation dataset containing 1,000 known design structures including their material composition, thickness and optical properties.
When comparing OptoGPT’s designs to the validation set, the difference between the two was only 2.58%; lower than the closest optical properties in the training dataset at 2.96%.
Similar to how large language models are able to respond to any text-based question, OptoGPT is trained on a large amount of data and is said to be able to respond well to general optical design tasks across the field.
Taking analysis a step further, the researchers used a statistical technique to map out associations that OptoGPT makes.
“The high-dimensional data structure of neural networks is a hidden space, too abstract to understand. We tried to poke a hole in the black box to see what was going on,” Guo said.
When mapped in a 2D space, materials cluster by type such as metals and dielectric materials, which are electrically insulating but can support an internal electric field. All dielectrics, including semiconductors, converge upon a central point as the thickness approaches 10 nanometres.
From an optics perspective, the pattern makes sense as light behaves similarly regardless of material as they approach such small thicknesses, helping further validate OptoGPT’s accuracy.
Known as an inverse design algorithm because it starts with the desired effect and works backward to a material design, OptoGPT is said to offer more flexibility than previous inverse design algorithm approaches, which were developed for specific tasks.
It enables researchers and engineers to design optical multilayer film structures for a wide breadth of applications.
Read next: Balfour Beatty to deliver North Carolina airport expansion
Are you a building professional? Sign up for a FREE MEMBERSHIP to upload news stories, post job vacancies, and connect with colleagues on our secure social feed.
