Efficient Structure Transformation Based on Sensitivity-Oriented Structure Adjustment for Inverse-Designed Devices

Round 1

Reviewer 1 Report

The Authors present an interesting method to adjusts the structures obtained using inverse design technique, leading to simpler tolopogies which are more robust and effective.

The manuscript is sound, and the proposed method can be of interest for researchers and companies working in the inverse design of photonic integrated devices.

The manuscript is well written and the arguments are well presented. There are a few typos that should be fixed (e.g., line 20: deigned->designed, line 63: "This make the ST inherently necessitates" -> "For this reason the ST inherently necessitates")

I have few questions regarding the method:

* what is its computational cost? Could the Authors give, e.g., an indication of the time required for the different steps in the numerical examples of Section 3?

* in which programming language did you implement the method?

* The ST method is strictly serial and cannot be parallelized. Are there any chances to parallelize at least parts of the new SA method?

The quality of English Language is overall good; minor editing is required.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The article explains how to obtain T and cross-sections through the structure adjustment method t. It can be published once the following issues are addressed.

1. Can the same design be applied to the TM mode?

2. The design may have higher mode excitations that could couple to the TE0 mode and cause reflections during transmission. What do you think is the best way to suppress the higher-order modes?

3. How small is the feature? Will the absence of sharp corners affect performance after fabrication? Can you justify the fabrication or explain the methodology?

The English language is good.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The paper is interesting and relevant to the community. I believe the authors can expand their literature survey further and also discuss more how their method compares to the current state of the art. 

The paper can be improved with grammatical revision.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

numerical procedure is presented and verified for optimisation task in nanophotoncs setting. validation is presented and discussed. authors achieved god verification of their proposed method. 

the abstract should clearly define what is new and less part of it should be used for motivation and introduction. what is solved by this new research? now it reads as we made study and it worked. this is not good enough. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf