Young researcher Aram Sargsyan shapes the future of computational systems

Aram Sargsyan, a PhD student at the YSU Institute of Physics, is conducting research within the framework of a scientific project aimed at developing a high-precision computational system capable of delivering stable and reliable results even under the influence of various environmental disturbances. "This topic is particularly important and timely in an era marked by the rapid development of artificial intelligence (AI). It will serve as a foundation for the creation of the next generation of computational systems," emphasizes Aram.

PhD student at the YSU Institute of Physics and a researcher at the Photonics AI Lab, Aram Sargsyan began his academic journey in computer and information sciences, focusing on artificial intelligence. Over time, he recognized the fundamental connections between these disciplines and physics and decided to dedicate himself to science by integrating knowledge from physics, AI, and information technology. Aram is naturally inquisitive, and it is precisely those questions that have yet to find definitive answers that motivate and inspire him. His goal is to uncover the unknown and drive innovation, developing new approaches and solutions. 

"It is this pursuit of discovering the unknown and creating something new that motivates and captivates me. I often draw inspiration from cross-disciplinary and field-specific interactions. Just as many AI approaches have been inspired by neuroscience, I try to project concepts and structures from different fields into my work," he said.

Nature is another source of inspiration. Systems and mechanisms inherent in the natural world often provide insights that lead to optimal solutions. 

Aram is among the winners of the "Research Support Program for PhD Students and Young Candidates – 2025", organized by the Armenian Ministry of Education, Science, Culture and Sport. His project, titled "Arbitrary Transmission Matrix Engineering with Physics-Informed Neural Network Pipelines", explores the precision of optical computational systems and compares them with conventional approaches. His research also examines the conditions and application areas in which optical systems may be most effective, as well as the additional advantages they offer.

– Aram, how relevant is your research topic in today's scientific landscape?

– Just as quantum computers have attracted widespread attention, optical computational systems—also known as optical computers—also have the potential to achieve significant progress, sometimes outperforming the best existing conventional technologies. Matrices and their multiplication play a fundamental role in many computational processes, particularly in AI. The primary objective of our project is to implement these operations using optical systems. A key requirement is the ability to generate arbitrary matrices within the system, enabling a wide range of arbitrary calculations. This is why our work focuses on addressing this specific challenge. While we also apply AI methods, our approach differs from conventional ones in that the models are grounded in physical laws and concepts. The physics-informed neural network approach enables us to achieve more accurate and robust results. 

Research groups around the world are working on similar challenges, and optical computing systems are already being applied in fields such as medicine and defense technologies. Our research builds on leading global achievements, combining them with our own ideas and approaches to advance the field worldwide. The innovative aspect of our work is largely determined by the approach we choose.

– What are the expected outcomes of your research?

– The project aims to develop a system capable of constructing and managing arbitrary matrices with high precision. Such functionality could later be applied during various computations, especially in large-scale data processing and AI-related problem-solving.

– What potential applications could your findings have?

– The results could have a wide range of applications. They may be used in fundamental science—for example, in optics, to develop more controllable and automated experimental systems—as well as in other fields. Such systems would enable faster and more energy-efficient training of AI models. There are also potential applications in medicine, such as the development of non-invasive, high-precision imaging technologies. Additionally, these approaches could improve the imaging capabilities of unmanned aerial vehicles (UAVs), particularly under challenging weather conditions. 

– What phase is your research currently in?

– The research is currently in an intensive experimental phase. We have already obtained a number of related results, including a recently published article on this topic. In addition, we have presented our interim findings at several conferences. At this stage, we are preparing the main paper, for which we have already gathered a substantial portion of the necessary results. We also plan further expansions and experimental refinements, applying our system to different problems and setups, testing the flexibility of our approach, and generalizing its capabilities.

– Do you see potential for commercial applications?

– Yes, I believe there is strong potential for commercializing the research results. Although the work is still in an active experimental and fundamental research phase, it is already evident that the approaches we are developing could form the basis for new technological solutions, as I mentioned earlier. With further development, these technologies could form the foundation for new devices, photonic systems, or software-hardware solutions, which may attract interest from both industry and high-tech companies.

When you love the work you do, challenges become manageable over time. I try to follow this principle.

Aram also emphasized the importance of support programs for young scientists, noting that they help cover research-related expenses and play a crucial role in motivating early-career researchers, encouraging them to work actively and engage in scientific activity. According to him, supporting science—and particularly young scientists—is a valuable investment not only in scientific advancement but also in societal and technological progress.

One of Aram's key goals is to develop science in line with contemporary advances, leveraging the latest tools and technologies. He is currently working to strengthen and expand his doctoral research while participating in initiatives within the AI field, keeping abreast of the latest developments to ensure his research remains relevant and impactful.

Aram was a member of a research team that won third place in the interdisciplinary project competition at the Armenian scientific community's Annual Review Conference 2024. The team's project focused on creating a new type of superconducting material.