Breakthrough in Artificial Intelligence: Researchers Develop Machine Learning Approach for Beam Shaping and Polarization Control

Strategic Insights -

Scientists at Technische Universitat Braunschweig have made a groundbreaking discovery in the field of artificial intelligence, presenting a machine learning-based approach to design binary metasurfaces for beam shaping, angle, and polarization state control. This innovative method utilizes Lumerical FDTD and Non-Dominated Sorting Genetic Algorithm III (NSGA-III) to optimize the topology of the outcoupling structure composed of subwavelength pixels, enabling precise control over the emitted light field.

Key Takeaways:

  • Researchers from Technische Universitat Braunschweig have developed a machine learning-based approach to design binary metasurfaces for beam shaping, angle, and polarization state control.
  • The research concludes that the generated pattern maintains the desired beam shape and angle while modulating the right/left circular and linear polarization states, allowing scalability for different wavelengths without significant distortion.
  • The design promises low fabrication complexity and scalability for chip-integrated quantum applications.
  • Sorokina Anastasiia, the lead researcher, was involved in the study, along with an additional 10 authors from the Institute of Semiconductor Technology at Technische Universitat Braunschweig.
  • The research utilizes Lumerical FDTD and Non-Dominated Sorting Genetic Algorithm III (NSGA-III) to optimize the topology of the outcoupling structure composed of subwavelength pixels.
  • The study presents a QR-Code Structure for Beam Shaping and Polarization Control via Machine Learning for Chip-Integrated Quantum Applications, published in EPJ Web of Conferences (2025, 335():05013).

Statistics:

  • The researchers implemented Lumerical FDTD and Non-Dominated Sorting Genetic Algorithm III (NSGA-III) to optimize the topology of the outcoupling structure.
  • The generated pattern maintains the desired beam shape and angle while modulating the right,left circular and linear polarization states.
  • The design shows promising low fabrication complexity and scalability for chip-integrated quantum applications.
  • The study involves 12 authors from the Institute of Semiconductor Technology at Technische Universitat Braunschweig.

Sources:

  • QR-Code Structure for Beam Shaping and Polarization Control via Machine Learning for Chip-Integrated Quantum Applications. EPJ Web of Conferences, 2025, 335():05013. (EPJ Web of Conferences - http://www.epj-conferences.org/).
  • EDP Sciences, publisher of EPJ Web of Conferences.