Inverse Design Methods for Novel, High-Performance and Manufacturable Components for Photonic Integrated Circuits

Hosted By: Optoelectronics Technical Group

20 February 2020, 12:00 - 13:00 - Eastern Daylight Time (UTC - 04:00)

Download Presentation Slides

Scaling up the complexity of photonic integrated circuits (PICs) requires the individual photonic components to become more compact, more efficient and more tolerant against manufacturing defects or variations, and for the design process to become more capable and less complex. Inverse design methods are a novel solution that allow the designer to directly specify performance targets and manufacturing constraints. The method can then automatically find optimal structures that fulfill the specified criteria.

In this webinar hosted by the OSA Optoelectronics Technical Group, James Pond and Jens Niegemann will present two complementary inverse design methods:

1) Parametrized geometry optimization, which makes it possible to automatically improve an already existing design with respect to performance and robustness.

2) Topology optimization, which enables the automatic generation of high-performance devices with a small footprint without the need for an initial design idea. It requires minimal human input and often converges on unintuitive but highly efficient structures.

The webinar will discuss the basic theory behind these inverse design methods and will demonstrate how to generate high-performance and easily manufacturable devices.

What You Will Learn:

  • The basic theory behind inverse design methods
  • How to use inverse design in practice
  • How to enforce size constraints to ensure manufacturability
  • How to extend the inverse design methods to generate robust designs that are optimally insensitive to manufacturing imperfections

Who Should Attend:

  • PhD students, research scientists and engineers working in university, government laboratories or designing commercial applications
  • Anyone wanting to learn more about the inverse design methods including topology optimization

About the Presenters: James Pond and Jens Niegemann, Lumerical Inc.

Dr. James Pond is the CTO and co-founder of Lumerical Inc. and is a driving force behind the company’s core software algorithms, technology, and advanced photonic modeling capabilities. He has almost two decades of experience in optical and photonic simulation, and is the author of numerous papers, patents and conference presentations.




Dr. Jens Niegemann received his PhD in theoretical physics in 2008 from the University of Karlsruhe, Germany. In 2015 he became Principal Scientist at Lumerical Inc. where he focuses on the development and implementation of efficient algorithms for photonic simulations. Dr. Niegemann has contributed to more than 50 peer-reviewed publications and conference presentations.