DETAILED ACTION
Claims 1-20 are presented for examination.
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Drawings
The drawings received on 14 September 2021 are accepted.
Information Disclosure Statement
The listing of references in the specification is not a proper information disclosure statement. 37 CFR 1.98(b) requires a list of all patents, publications, or other information submitted for consideration by the Office, and MPEP § 609.04(a) states, "the list may not be incorporated into the specification but must be submitted in a separate paper." Therefore, unless the references have been cited by the examiner on form PTO-892, they have not been considered.
Specification page 17 lists: Hinton, G. E. & Salakhutdinov, R. R. “Reducing the dimensionality of data with neural networks” Science 313, 504-507 (2006)), and Barber, C. B., Dobkin, D. P., Dobkin, D. P. & Huhdanpaa, H. “The quickhull algorithm for convex hulls” ACM Transactions on Mathematical Software (TOMS) 22, 469-483 (1996)).
Specification pages 21-22 list: Kiarashinejad, Y., Abdollahramezani, S., Zandehshahvar, M., Hemmatyar, 0. & Adibi, A Deep learning reveals underlying physics of light-matter interactions in nanophotonic devices. Advanced Theory and Simulations (2019).
Specification
Applicant is reminded of the proper language and format for an abstract of the disclosure.
The abstract should be in narrative form and generally limited to a single paragraph on a separate sheet within the range of 50 to 150 words in length. The abstract should describe the disclosure sufficiently to assist readers in deciding whether there is a need for consulting the full patent text for details.
The language should be clear and concise and should not repeat information given in the title. It should avoid using phrases which can be implied, such as, “The disclosure concerns,” “The disclosure defined by this invention,” “The disclosure describes,” etc. In addition, the form and legal phraseology often used in patent claims, such as “means” and “said,” should be avoided.
The abstract of the disclosure is objected to because:
The abstract includes phrases which can be implied.
Examiner suggests amending the abstract as follows:
Identifying a first plurality of data points comprising a design space and a response space.Computing a first convex hull of all data points in the plurality of data points, merging the first convex hull with previous convex hulls to form an optimized convex hull.The previous convex hulls comprising a second plurality of data points comprising previous design spaces and previous response spaces, and determining, by the optimized convex hull, a feasible optical response performance within an electromagnetic nanostructure.
A corrected abstract of the disclosure is required and must be presented on a separate sheet, apart from any other text. See MPEP § 608.01(b).
Claim Rejections - 35 USC § 112(b) Indefiniteness
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-20 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, or for pre-AIA the applicant regards as the invention.
Claims 1 and 18 recites “cascading the second neural network with the decoding layer of the first multi-layer neural network.” It is unclear what is meant by cascading the neural network with a decoding layer of a different neural network. Specification paragraph 83 merely repeat this same language and fails to define what is meant by cascading here. For purposes of compact prosecution, Examiner is interpreting any combination of the two neural networks or subnetworks is a cascading of the respective networks.
Claim 11 recites “wherein the response space and the reduced response space have a one-to-one dimensional relationship.” However, claim 1 upon which claim 11 depends recites “generate a reduced response space (20) having reduced dimensionality compared to the response space.” This is a contradiction because having a one-to-one dimensional relationship shows they have the same number of dimensions while parent claim 1 requires the reduced response space to have a reduced dimensionality. A person of ordinary skill in the art would not know how to resolve these two contradicting limitations.
Claim 13 third clause recites “previous convex hulls.” There is a lack of antecedent basis for previous convex hulls.
Dependent claims 2-12, 14-17, 19, and 20 are rejected for depending from a rejected claim.
Claim Rejections - 35 USC § 112(d) Improper Dependent
The following is a quotation of 35 U.S.C. 112(d):
(d) REFERENCE IN DEPENDENT FORMS.—Subject to subsection (e), a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
The following is a quotation of pre-AIA 35 U.S.C. 112, fourth paragraph:
Subject to the following paragraph [i.e., the fifth paragraph of pre-AIA 35 U.S.C. 112], a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
Claim 12 is rejected under 35 U.S.C. 112(d) or pre-AIA 35 U.S.C. 112, 4th paragraph, as being of improper dependent form for failing to further limit the subject matter of the claim upon which it depends, or for failing to include all the limitations of the claim upon which it depends.
Claim 12 recites “the reduced design space and the reduced response space have a one-to-one dimensional relationship.” This is a tautological truth which is always true. A space shares a one-to-one dimensional relationship with itself via the identify relationship. Reciting a mathematical tautology fails to further limit the parent claim.
Applicant may cancel the claim(s), amend the claim(s) to place the claim(s) in proper dependent form, rewrite the claim(s) in independent form, or present a sufficient showing that the dependent claim(s) complies with the statutory requirements.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
To determine if a claim is directed to patent ineligible subject matter, the Court has guided the Office to apply the Alice/Mayo test, which requires:
1. Determining if the claim falls within a statutory category;
2A. Determining if the claim is directed to a patent ineligible judicial exception consisting of a law of nature, a natural phenomenon, or abstract idea; and
2B. If the claim is directed to a judicial exception, determining if the claim recites limitations or elements that amount to significantly more than the judicial exception.
See MPEP §2106.
Step 2A is a two prong inquiry. MPEP §2106.04(II)(A). Under 2A(i), the first prong, examiners evaluate whether a law of nature, natural phenomenon, or abstract idea is set forth or described in the claim. Abstract ideas include mathematical concepts, certain methods of organizing human activity, and mental processes. MPEP §2106.04(a)(2). Under 2A(ii), the second prong, examiners determine whether any additional limitations integrates the judicial exception into a practical application. MPEP §2106.04(d).
Claim 1 step 2A(i):
The claim(s) recite:
1. A system (10) for detecting a feasible optical response performance from a structure, the system comprising:
…
identify, based on the input design data, limitation data (14);
generate, based on the limitation data, simulation data (15) comprising a design space (16) and a response space (17);
train, utilizing the response space, a first multi-layer neural network (18) to generate a reduced response space (20) having reduced dimensionality compared to the response space, the first multi-layer neural network comprising an encoding layer and a decoding layer;
train, utilizing the design space and the response space, a second neural network (22) to generate a reduced design space (24) having reduced dimensionality compared to the design space;
generate, by cascading the second neural network with the decoding layer of the first multi-layer neural network, an optimization convex hull (30); and
invert, using the optimization convex hull, the design space and the response space to generate the feasible optical response performance from the structure.
Detecting a feasible optical response performance corresponds with respective mathematical calculations of the performance and/or mental processes in the form of observation of the respective results.
Identifying the limitation data corresponds with a mathematical definition of respective constraints. See further Specification paragraph 79 discussing the limitation data.
Generating the simulation data of the design space and response space is further recitation of mathematical definition of respective variables.
Training the first and second neural networks is a recitation of respective mathematical algorithms encompassed by Specification paragraph 81. While training a machine learning algorithm is not always interpreted as reciting a mathematical concept, when the Specification recites specific mathematical algorithms, such as backpropagation, then the claim recitation of training is properly interpreted as a mathematical concept. See “July 2024 Subject Matter Eligibility Examples” page 6 Example 47 claim 2 analysis (“Step (c) recites training an ANN using a selected algorithm. The training algorithm is a backpropagation algorithm and a gradient descent algorithm. When given their broadest reasonable interpretation in light of the background, the backpropagation algorithm and gradient descent algorithm are mathematical calculations. The plain meaning of these terms are optimization algorithms, which compute neural network parameters using a series of mathematical calculations.”).
Cascading the second neural network with the decoding layer of the first neural network to generate an optimized convex hull is interpreted in light of the Specification. Specification paragraph 104 states “dimensionality reduction (DR) implementation is based on … the Quickhull [citing Barber “The quickhull algorithm for convex hulls].”) The dimensionality reduction corresponds with the claimed optimization. Similar to above, when the Specification describes a specific mathematical algorithm for performing a respective step that step is interpreted as being directed towards the specifically disclosed mathematical algorithm. Therefore, the generation of the optimization convex hull using the quickhull algorithm is found to be a recitation of mathematical algorithm.
Inverting the design space and the response space to generate feasible optical response performance is a mathematical inversion of the respective mathematical structure. Further recitation of mathematical operation is further recitation of mathematical concept.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 1 step 2A(ii):
This judicial exception is not integrated into a practical application because:
The claim(s) recite:
one or more processors;
at least one memory in communication with the one or more processors and configured to store instructions that, when executed by the one or more processors, are configured to cause the system to:
collect input design data (12);
The processor and memory are recited at a high-level of generality (i.e., as a generic processor performing generic computer functions) such that it amounts no more than mere instructions to apply the exception using a generic computer components. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. See MPEP §2106.05(b) (“Merely adding a generic computer, generic computer components, or a programmed computer to perform generic computer functions does not automatically overcome an eligibility rejection. Alice Corp. Pty. Ltd. v. CLS Bank Int’l, 573 U.S. 208, 223-24, 110 USPQ2d 1976, 1983-84 (2014).”).
Collecting data, recited at a high level of generality, is insignificant extra solution activity in the form of a generic data gathering. See MPEP §2106.05(g).
Claim 1 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Regarding the processor and memory, limitations analyzed under MPEP §2106.05(b) in step 2A(ii) above are analyzed the same here for step 2B.
Regarding the data collection, MPEP §2106.05(d) provides examples:
i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information);
iv. Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015)
These data gathering examples are encompassed by the generic recitation of data gathering recited by the claim. Accordingly, the claim recitation here is at least as abstract as the examples given in the MPEP.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 2 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
2. The system of claim 1, wherein the instructions are further configured to cause the system to determine, by using the optimization convex hull, a designation of overlapping or non-overlapping of a desired design space of a desired structure.
Determining whether a design point of a desired structure is overlapping, or not overlapping with a convex hull is a mathematical operation. Specification paragraph 141 states “Find the number of intersection of the line xa and every vertex of the convex hull. If the number of intersections is odd, the point lies inside the convex hull. Otherwise, if the number of intersections is even or zero, this point is outside the convex hull.” This mathematical algorithm is further recitation of mathematical concept related to the mathematically defined convex hull.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 2 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 2 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 3 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
3. The system of claim 1, wherein the instructions are further configured to cause the system to validate, by using validation data (40), the optimization convex hull.
Validating with validation data is recitation of performing further mathematical calculation of respective validation metrics.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 3 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 3 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 4 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
4. The system of claim 1, wherein the input design data comprises a plurality of randomly generated patterns of a simulated structure.
Randomly generating patterns for a design space of the structure corresponds with the start of the mathematical algorithm. Randomly generating input patterns is common for any Markov process and thus corresponds with respective mathematical algorithm steps.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 4 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 4 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 5 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 5 step 2A(ii):
This judicial exception is not integrated into a practical application because:
The claim(s) recite:
5. The system of claim 4, wherein limitation data comprises structural limitation data relating to physical properties of a photonic nanostructure.
The structural limitation data relating to physical properties of a photonic nanostructure is a general linking to a field of use. Merely linking the use of an abstract idea to a field of use fails to integrate the claim into a practical application. See MPEP §2106.05(h).
Claim 5 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Regarding the photonic nanostructure, limitations analyzed under MPEP §2106.05(h) in step 2A(ii) above are analyzed the same here for step 2B.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 6 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 6 step 2A(ii):
This judicial exception is not integrated into a practical application because:
The claim(s) recite:
6. The system of claim 5, wherein the photonic nanostructure comprises a metasurface.
The nanostructure being a metasurface is a general linking to a field of use. Merely linking the use of an abstract idea to a field of use fails to integrate the claim into a practical application. See MPEP §2106.05(h).
Claim 6 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Regarding the photonic nanostructure, limitations analyzed under MPEP §2106.05(h) in step 2A(ii) above are analyzed the same here for step 2B.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 7 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
7. The system of claim 1, wherein the first multi-layer neural network is an autoencoder.
Training and executing an autoencoder remains mathematical subject matter according to the analysis under claim 1. Being an autoencoder instead of a more generalized multi-layer neural network does not alter the previous analysis of the training and generating steps.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 7 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 7 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 8 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
8. The system of claim 7, wherein the autoencoder utilizes mean squared error as a cost function.
A mean squared error cost function is an explicit recitation of mathematical subject matter.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 8 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 8 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 9 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
9. The system of claim 1, wherein the simulation data comprises a multi-dimensional response space.
A multi-dimensional response space is a mathematical description of the mathematically defined numerical space.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 9 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 9 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 10 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
10. The system of claim 9, wherein the simulation data comprises at least a six-dimensional response space.
A six-dimensional response space is a mathematical description of the mathematically defined numerical space.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 10 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 10 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 11 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
11. The system of claim 1, wherein the response space and the reduced response space have a one-to-one dimensional relationship.
The dimensional relationship of the reduced response space is a mathematical description of the mathematically defined numerical space.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 11 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 11 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 12 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
12. The system of claim 1, wherein the reduced design space and the reduced response space have a one-to-one dimensional relationship.
The dimensional relationship of the reduced design space and response space is a mathematical description of the mathematically defined numerical space.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 12 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 12 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 13 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
13. A method comprising:
identifying a first plurality of data points comprising a design space and a response space;
computing a first convex hull of all data points in the first plurality of data points;
merging the first convex hull with previous convex hulls to form an optimized convex hull, the previous convex hulls comprising a second plurality of data points comprising previous design spaces and previous response spaces; and
determining, by the optimized convex hull, a feasible optical response performance ….
Identifying a design space and response space corresponds with a mathematical definition of respective mathematical numerical spaces.
Computing a convex hull from data points is an explicit mathematical operation.
Merging convex hulls is an explicit mathematical operation or mathematically defined entities.
Determining the feasible optical response performance space of the optimized convex hull design space corresponds with mathematical determination of the mathematically defined feasible space. Alternatively, determining the feasible space encompasses mental processes in the form of observation and evaluation.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 13 step 2A(ii):
This judicial exception is not integrated into a practical application because:
The claim(s) recite:
determining, … within an electromagnetic nanostructure.
The designed structure being an electromagnetic nanostructure is generally linking to a field of use. Merely linking the use of an abstract idea to a field of use fails to integrate the claim into a practical application. See MPEP §2106.05(h).
Claim 13 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Regarding the electromagnetic nanostructure, limitations analyzed under MPEP §2106.05(h) in step 2A(ii) above are analyzed the same here for step 2B.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 14 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
14. The method of claim 13, further comprising:
inverting, using the optimized convex hull, the design space and the response space to generate the feasible optical response performance from the electromagnetic nanostructure.
Inverting the design space and the response space to generate feasible optical response performance is a mathematical inversion of the respective mathematical structure. Further recitation of mathematical operation is further recitation of mathematical concept.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 14 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 14 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 15 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
15. The method of claim 13, wherein inverting, using the optimized convex hull, the design space and the response space further comprises applying a one-class support vector machine algorithm.
Applying a support vector machine algorithm is recitation to perform a respective mathematical algorithm.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 15 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 15 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 16 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
16. The method of claim 13, wherein inverting, using the optimized convex hull, the design space and the response space further comprises designating a desired design space of a desired electromagnetic nanostructure as overlapping or non-overlapping the optimized convex hull.
Designating a desired design space is mental process in the form of judgment or opinion. Combining mental process steps with a mathematical concept is a combination of abstract ideas which remains a recitation of abstract ideas.
The design space which is overlapping or non-overlapping is a mathematical definition of a respective numerical space.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 16 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 16 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 17 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
17. The method of claim 13, further comprising:
merging, using the optimized convex hull and third plurality of data points from a desired electromagnetic nanostructure structure, a re-optimized convex hull when the desired electromagnetic nanostructure comprises a desired design space designated as non-overlapping.
Merging convex hulls is an explicit mathematical operation or mathematically defined entities.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 17 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 17 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 18 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
18. A system for detecting a feasible optical response performance …
identify, based on the input electromagnetic nanostructure design data, structural limitation data comprising material properties, potential nanostructure geometry, periodic/non-periodic, unit-cell structure, and fabrication limitations;
generate, based on the structural limitation data, electromagnetic simulation data comprising a design space, the design space comprising a set of design patterns and a corresponding response space comprising a corresponding set of response patterns;
train, utilizing the corresponding response space, a first multi-layer neural network to generate a reduced response space having reduced dimensionality compared to the corresponding response space, the first multi-layer neural network comprising an encoding layer and a decoding layer;
train, utilizing the design space and the corresponding response space, a second neural network to generate a reduced design space having reduced dimensionality compared to the design space;
generate, by cascading the second neural network with the decoding layer of the first multi-layer neural network, an optimization convex hull; and
invert, using the optimization convex hull, the design space and the corresponding response space to generate the feasible optical response performance from the electromagnetic nanostructure.
Detecting a feasible optical response performance corresponds with respective mathematical calculations of the performance and/or mental processes in the form of observation of the respective results.
Identifying the limitation data corresponds with a mathematical definition of respective constraints. See further Specification paragraph 79 discussing the limitation data.
Generating the simulation data of the design space and response space is further recitation of mathematical definition of respective variables.
Training the first and second neural networks is a recitation of respective mathematical algorithms encompassed by Specification paragraph 81. While training a machine learning algorithm is not always interpreted as reciting a mathematical concept, when the Specification recites specific mathematical algorithms, such as backpropagation, then the claim recitation of training is properly interpreted as a mathematical concept. See “July 2024 Subject Matter Eligibility Examples” page 6 Example 47 claim 2 analysis (“Step (c) recites training an ANN using a selected algorithm. The training algorithm is a backpropagation algorithm and a gradient descent algorithm. When given their broadest reasonable interpretation in light of the background, the backpropagation algorithm and gradient descent algorithm are mathematical calculations. The plain meaning of these terms are optimization algorithms, which compute neural network parameters using a series of mathematical calculations.”).
Cascading the second neural network with the decoding layer of the first neural network to generate an optimized convex hull is interpreted in light of the Specification. Specification paragraph 104 states “dimensionality reduction (DR) implementation is based on … the Quickhull [citing Barber “The quickhull algorithm for convex hulls].”) The dimensionality reduction corresponds with the claimed optimization. Similar to above, when the Specification describes a specific mathematical algorithm for performing a respective step that step is interpreted as being directed towards the specifically disclosed mathematical algorithm. Therefore, the generation of the optimization convex hull using the quickhull algorithm is found to be a recitation of mathematical algorithm.
Inverting the design space and the response space to generate feasible optical response performance is a mathematical inversion of the respective mathematical structure. Further recitation of mathematical operation is further recitation of mathematical concept.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 18 step 2A(ii):
This judicial exception is not integrated into a practical application because:
The claim(s) recite:
… from an electromagnetic nanostructure, the system comprising:
one or more processors;
at least one memory in communication with the one or more processors and configured to store instructions that, when executed by the one or more processors, are configured to cause the system to:
collect input electromagnetic nanostructure design data;
The nanostructure being a metasurface is a general linking to a field of use. Merely linking the use of an abstract idea to a field of use fails to integrate the claim into a practical application. See MPEP §2106.05(h).
The processor and memory are recited at a high-level of generality (i.e., as a generic processor performing generic computer functions) such that it amounts no more than mere instructions to apply the exception using a generic computer components. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. See MPEP §2106.05(b) (“Merely adding a generic computer, generic computer components, or a programmed computer to perform generic computer functions does not automatically overcome an eligibility rejection. Alice Corp. Pty. Ltd. v. CLS Bank Int’l, 573 U.S. 208, 223-24, 110 USPQ2d 1976, 1983-84 (2014).”).
Collecting data, recited at a high level of generality, is insignificant extra solution activity in the form of a generic data gathering. See MPEP §2106.05(g).
Claim 18 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Regarding the photonic nanostructure, limitations analyzed under MPEP §2106.05(h) in step 2A(ii) above are analyzed the same here for step 2B.
Regarding the processor and memory, limitations analyzed under MPEP §2106.05(b) in step 2A(ii) above are analyzed the same here for step 2B.
Regarding the data collection, MPEP §2106.05(d) provides examples:
i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information);
iv. Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015)
These data gathering examples are encompassed by the generic recitation of data gathering recited by the claim. Accordingly, the claim recitation here is at least as abstract as the examples given in the MPEP.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 19 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
19. The system of claim 18, wherein the instructions are further configured to cause the system to determine, by using the optimization convex hull, a designation of overlapping or non-overlapping of a desired design space of a desired structure.
Designating a desired design space is mental process in the form of judgment or opinion. Combining mental process steps with a mathematical concept is a combination of abstract ideas which remains a recitation of abstract ideas.
The design space which is overlapping or non-overlapping is a mathematical definition of a respective numerical space.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 19 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 19 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim 20 step 2A(i):
Dependent claims recite at least the identified judicially excepted subject matter of their parent claim(s).
The claim(s) recite:
20. The system of claim 18, wherein the electromagnetic simulation data comprises a multi-dimensional response space ranging from about 2-dimensional to about 6-dimensional.
A multi-dimensional response space is a mathematical description of the mathematically defined numerical space.
This falls within the mathematical concept grouping of abstract ideas. See MPEP §2106.04(a)(2).
Claim 20 step 2A(ii):
This judicial exception is not integrated into a practical application because:
Claim(s) do not recite any “additional” limitations.
Claim 20 step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception, when considered individually and in combination, because:
Claim(s) do not recite any “additional” limitations.
When further considering the claims as a whole and as an ordered combination the claims fail to amount to significantly more than the judicially excepted abstract idea.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1, 2, 4-12 and 18-20
Claims 1, 2, 4-12 and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Liu, Z., et al. “A Hybrid Strategy for the Discovery and Design of Photonic Nanostructures” arXiv:1902.02293v2 [physics.optics] (Feb. 2019) [herein “Liu”] in view of US patent 11,593,533 B2 Grossman, et al. [herein “Grossman”].
Claim 1 recites “1. A system (10) for detecting a feasible optical response performance from a structure.” Liu page 4 first paragraph last sentence discloses “the inverse design of complicated photonic structures with multiple dimensional parameters concurrently optimized.” The inverse design of photonic structures with optimized parameters corresponds with determining optical performance for the photonic structures.
Liu page 9 second paragraph disclose “We also note that although the transmittance spectra are set as the design objective in these experiments, any photonic responses such as the diffraction behavior, optical chirality, and field localization can be used as the intended design criteria without further adjustment of our framework.” The photonic response corresponds with an optical response performance.
Claim 1 further recites “the system comprising: one or more processors; at least one memory in communication with the one or more processors and configured to store instructions that, when executed by the one or more processors.” Liu does not explicitly disclose a processor or memory; however, in analogous art of exploration of large-scale generate design spaces, Grossman column 4 lines 1-4 teach “computer system, including a desktop computer, a laptop computer, a mobile device, a virtualized instance of a computing device, a distributed and/or cloud based computer system.” Grossman column 4 lines 9-10 teach “a processor 112, input/output (I/O) devices 114, and a memory 116, coupled together.”
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Liu and Grossman. One having ordinary skill in the art would have found motivation to use generating a convex hull of design options into the system of discovery and design of photonic nanostructures for the advantageous purpose of visualizing and exploring large-scale generative design datasets “as well as various design characteristics associated with the different design options.” See Grossman column 3 lines 21-30.
Claim 1 further recites “are configured to cause the system to: collect input design data (12).” Liu page 4 first paragraph “The designing process only takes customer-defined, on-demand physical properties as the input.” The input properties correspond with an input design data.
Claim 1 further recites “identify, based on the input design data, limitation data (14).” Liu page 4 last paragraph discloses:
we first construct a compact representation of all possible candidates of the geometry. Derived from deep neural networks, this compact representation can be computed by encoding the geometric data into a latent space with the help of a deep generative model, such as a generative adversarial network (GAN) or a variational autoencoder (VAE). The latter is adopted here for the representation of geometrical shapes of photonic nanostructures.
The possible candidates of geometry are respective identified bounds on the input design data.
Liu page 7 first paragraph discloses “restrictions with regard to the frequency range, the lattice period, the thickness of the patterned layer, and the type of constituent materials are predefined by the simulator.” The predefined restrictions on, i.e. thickness, are identified limitation data.
Claim 1 further recites “generate, based on the limitation data, simulation data (15) comprising a design space (16) and a response space (17).” Liu page 4 last paragraph discloses:
we first construct a compact representation of all possible candidates of the geometry. Derived from deep neural networks, this compact representation can be computed by encoding the geometric data into a latent space with the help of a deep generative model, such as a generative adversarial network (GAN) or a variational autoencoder (VAE). The latter is adopted here for the representation of geometrical shapes of photonic nanostructures.
The VAE representation of geometrical shapes of the photonic nanostructures is a generated design space for the photonic structures. The geometric data for the nanostructures corresponds with a design space.
Liu page 5 second paragraph disclose “The encoder transforms the input geometric data into mean vectors µ and standard deviation vectors
σ
, and latent vectors v are sampled from the Gaussian distribution parametrized by µ and
σ
.” The entire distribution parameterized by the mean and standard deviation is a first design space. Liu page 6 last paragraph discloses “The simulator accepts the image of a nanostructure and approximates the four components of the transmittance of the metasurface.” The approximated components of transmittance correspond to a simulated response space.
Claim 1 further recites “train, utilizing the response space, a first multi-layer neural network (18) to generate a reduced response space (20) having reduced dimensionality compared to the response space, the first multi-layer neural network comprising an encoding layer and a decoding layer.” Liu page 9 second paragraph teaches “We also note that although the transmittance spectra are set as the design objective in these experiments, any photonic responses such as the diffraction behavior, optical chirality, and field localization can be used as the intended design criteria without further adjustment of our framework.”
Liu page 7 second paragraph discloses “objective is to minimize the
L
2
and/or
L
∞
norm of the difference between the input spectra and the approximated transmittance, a reasonable choice of the fitness function can be written as [equation (1)].”
Liu page 5 first paragraph discloses:
and latent vectors v are sampled from the Gaussian distribution parametrized by μ and σ. The decoder then reconstructs v back to the geometric information. When the training of this VAE is completed, the decoder can be operated as a geometric data generator as indicated in Fig. 1(b), so that when fed with a randomly sampled vector v, the corresponding pattern of the nanostructure s can be reconstructed.
The latent vectors are a reduced response space compared with the entire continuous Gaussian distribution which is sampled.
Liu page 14 figure 1 shows:
PNG
media_image1.png
200
400
media_image1.png
Greyscale
Liu page 14 figure 1 encoder and decoder combination corresponds with the first neural network.
Claim 1 further recites “train, utilizing the design space and the response space, a second neural network (22) to generate a reduced design space (24) having reduced dimensionality compared to the design space.” Liu page 5 first paragraph discloses:
and latent vectors v are sampled from the Gaussian distribution parametrized by μ and σ. The decoder then reconstructs v back to the geometric information. When the training of this VAE is completed, the decoder can be operated as a geometric data generator as indicated in Fig. 1(b), so that when fed with a randomly sampled vector v, the corresponding pattern of the nanostructure s can be reconstructed.
The latent space vectors v and randomly sampled vectors v are a reduced dimensionality as compared with the entire continuous Gaussian distribution defined by µ and
σ
.
Liu page 14 figure 1 b) decoder corresponds with a second neural network with a reduced dimensionality. Encircling and separating the decoder b) corresponds with generating the second neural network. Training of the first neural network correspond with a training of this second reduced neural network.
Claim 1 further recites “generate, by cascading the second neural network with the decoding layer of the first multi-layer neural network, an optimization convex hull (30).” Liu page 14 figure 1 caption teaches “(b) After the training, the decoder encircled in (a) can be separated and treated as a generator of geometric data.” Encircling and separating the decoder of the first neural network corresponds with cascading the second neural network with the decoding layer of the first multi-layer neural network.
But Liu does not explicitly disclose a convex hull; however, in analogous art of exploration of large-scale generate design spaces, Grossman column 12 lines 43-45 teach “GUI engine 210 may generate a convex hull that intersects design options 126 that minimize a total distance to both ranking axes.” Generating a convex hull intersecting design options corresponds with computing a first convex hull of data points of the design space. The generated convex hull which intersects the design options minimizing distance to the raking axes corresponds with an optimized convex hull.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Liu and Grossman. One having ordinary skill in the art would have found motivation to use generating a convex hull of design options into the system of discovery and design of photonic nanostructures for the advantageous purpose of visualizing and exploring large-scale generative design datasets “as well as various design characteristics associated with the different design options.” See Grossman column 3 lines 21-30.
Claim 1 further recites “and invert, using the optimization convex hull, the design space and the response space to generate the feasible optical response performance from the structure.” Liu page 4 first paragraph last sentence teaches “scalability of the framework enables our approach to be implementable to the inverse design of complicated photonic structures with multiple dimensional parameters concurrently optimized.” An inverse design of the structure from the parameters is an inversion of the design space and the response space to generate the photonic structure from the respective transmittance spectra. See further Liu page 7 second paragraph fitness function.
Claim 2 further recites “2. The system of claim 1, wherein the instructions are further configured to cause the system to determine, by using the optimization convex hull, a designation of overlapping or non-overlapping of a desired design space of a desired structure.” From the above list of alternatives Examiner is selecting “overlapping.”
Liu does not explicitly disclose a convex hull; however, in analogous art of exploration of large-scale generate design spaces, Grossman column 12 lines 43-45 teach “GUI engine 210 may generate a convex hull that intersects design options 126 that minimize a total distance to both ranking axes.” Generating a convex hull intersecting design options corresponds with designating a desired space as overlapping with the generated convex hull.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Liu and Grossman. One having ordinary skill in the art would have found motivation to use generating a convex hull of design options into the system of discovery and design of photonic nanostructures for the advantageous purpose of visualizing and exploring large-scale generative design datasets “as well as various design characteristics associated with the different design options.” See Grossman column 3 lines 21-30.
Claim 4 further recites “4. The system of claim 1, wherein the input design data comprises a plurality of randomly generated patterns of a simulated structure.” Liu page 8 second paragraph discloses “500 random test patterns of metasurfaces s from all classes of geometry are selected, and FEM simulations are performed to obtain the spectra T of these metasurfaces, which are subsequently fed to the framework as the target spectra.” The random test patterns of metasurfaces correspond with a plurality of randomly generated patterns of simulated structures.
Claim 5 further recites “5. The system of claim 4, wherein limitation data comprises structural limitation data relating to physical properties of a photonic nanostructure.” Liu page 4 last paragraph discloses:
we first construct a compact representation of all possible candidates of the geometry. Derived from deep neural networks, this compact representation can be computed by encoding the geometric data into a latent space with the help of a deep generative model, such as a generative adversarial network (GAN) or a variational autoencoder (VAE). The latter is adopted here for the representation of geometrical shapes of photonic nanostructures.
The possible candidates of geometry are respective identified bounds on the input design data.
Liu page 7 first paragraph discloses “restrictions with regard to the frequency range, the lattice period, the thickness of the patterned layer, and the type of constituent materials are predefined by the simulator.” A restriction on the type of constituent material is a structural limitation relating to physical properties of a photonic nanostructure.
Claim 6 further recites “6. The system of claim 5, wherein the photonic nanostructure comprises a metasurface.” Liu page 2 last line to page 3 first line discloses “Designing metasurfaces to obtain aimed optical properties.”
Claim 7 further recites “7. The system of claim 1, wherein the first multi-layer neural network is an autoencoder.” Liu abstract discloses “Through a variational autoencoder, all potential patterns of unit nanostructures are encoded into a continuous latent space.”
Claim 8 further recites “8. The system of claim 7, wherein the autoencoder utilizes mean squared error as a cost function.” Liu page 7 second paragraph discloses “objective is to minimize the
L
2
and/or
L
∞
norm of the difference between the input spectra and the approximated transmittance.” Minimizing the
L
2
norm difference corresponds with using mean squared error as a cost function. The fitness function corresponds with a cost function.
Claim 9 further recites “9. The system of claim 1, wherein the simulation data comprises a multi-dimensional response space.” Liu page 4 first paragraph last sentence discloses “the inverse design of complicated photonic structures with multiple dimensional parameters concurrently optimized.”
Claim 10 further recites “10. The system of claim 9, wherein the simulation data comprises at least a six-dimensional response space.” Liu page 6 second paragraph discloses “the encoded vector v of the images by the VAE has the dimension of 10.” A ten dimensional latent space is at least six-dimensional. The continuous dimensional space of the Gaussian distribution of the non-reduced response space is even larger than the reduced dimensional space of the vector v.
Claim 11 further recites “11. The system of claim 1, wherein the response space and the reduced response space have a one-to-one dimensional relationship.” This claim is rejected under §112(b) above. Because of the contradictory limitations a claim scope is impossible to determine. However, for purposes of compact prosecution the following teachings are noted:
Liu page 5 second paragraph disclose “The encoder transforms the input geometric data into mean vectors µ and standard deviation vectors
σ
, and latent vectors v are sampled from the Gaussian distribution parametrized by µ and
σ
.” The sampling of the vectors v from the Gaussian distribution mean there is a relationship where each vector of v corresponds with one sampled point of the overall distribution.
Claim 12 further recites “12. The system of claim 1, wherein the reduced design space and the reduced response space have a one-to-one dimensional relationship.” This is tautologically true. Any space has a one-to-one dimensional relationship with itself via self-identity. This claim is rejected under §112(d) above. Because of the contradictory limitations a claim scope is impossible to determine. However, for purposes of compact prosecution the following teachings are noted:
Liu page 5 second paragraph disclose “The encoder transforms the input geometric data into mean vectors µ and standard deviation vectors
σ
, and latent vectors v are sampled from the Gaussian distribution parametrized by µ and
σ
.” The sampling of the vectors v from the Gaussian distribution mean there is a relationship where each vector of v corresponds with one sampled point of the overall distribution.
Claim 18 recites “18. A system for detecting a feasible optical response performance from an electromagnetic nanostructure.” Liu page 4 first paragraph last sentence discloses “the inverse design of complicated photonic structures with multiple dimensional parameters concurrently optimized.” The inverse design of photonic structures with optimized parameters corresponds with determining optical performance for the photonic structures. A photonic nanostructure corresponds with an electromagnetic nanostructure.
Liu page 9 second paragraph disclose “We also note that although the transmittance spectra are set as the design objective in these experiments, any photonic responses such as the diffraction behavior, optical chirality, and field localization can be used as the intended design criteria without further adjustment of our framework.” The photonic response corresponds with an optical response performance.
Claim 18 further recites “the system comprising: one or more processors; at least one memory in communication with the one or more processors and configured to store instructions that, when executed by the one or more processors.” Liu does not explicitly disclose a processor or memory; however, in analogous art of exploration of large-scale generate design spaces, Grossman column 4 lines 1-4 teach “computer system, including a desktop computer, a laptop computer, a mobile device, a virtualized instance of a computing device, a distributed and/or cloud based computer system.” Grossman column 4 lines 9-10 teach “a processor 112, input/output (I/O) devices 114, and a memory 116, coupled together.”
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Liu and Grossman. One having ordinary skill in the art would have found motivation to use generating a convex hull of design options into the system of discovery and design of photonic nanostructures for the advantageous purpose of visualizing and exploring large-scale generative design datasets “as well as various design characteristics associated with the different design options.” See Grossman column 3 lines 21-30.
Claim 18 further recites “are configured to cause the system to: collect input electromagnetic nanostructure design data.” Liu page 4 first paragraph “The designing process only takes customer-defined, on-demand physical properties as the input.” The input properties correspond with an input design data.
Claim 18 further recites “identify, based on the input electromagnetic nanostructure design data, structural limitation data comprising material properties, potential nanostructure geometry, periodic/non-periodic, unit-cell structure, and fabrication limitations.” Liu page 4 last paragraph discloses:
we first construct a compact representation of all possible candidates of the geometry. Derived from deep neural networks, this compact representation can be computed by encoding the geometric data into a latent space with the help of a deep generative model, such as a generative adversarial network (GAN) or a variational autoencoder (VAE). The latter is adopted here for the representation of geometrical shapes of photonic nanostructures.
The possible candidates of geometry corresponds with potential nanostructure periodic or non-periodic geometry data.
Liu page 7 first paragraph discloses “restrictions with regard to the frequency range, the lattice period, the thickness of the patterned layer, and the type of constituent materials are predefined by the simulator.” The predefined restrictions on, i.e. thickness, are identified limitation data. The type of constituent materials corresponds with material property limitation data. The thickness and type of material correspond with fabrication limitation data.
Liu page 5 first line disclose “the latent space of all unit-cell patterns in this work.” The unit-cell patterns correspond with unit-cell structure.
Claim 18 further recites “generate, based on the structural limitation data, electromagnetic simulation data comprising a design space, the design space comprising a set of design patterns and a corresponding response space comprising a corresponding set of response patterns.” Liu page 4 last paragraph discloses:
we first construct a compact representation of all possible candidates of the geometry. Derived from deep neural networks, this compact representation can be computed by encoding the geometric data into a latent space with the help of a deep generative model, such as a generative adversarial network (GAN) or a variational autoencoder (VAE). The latter is adopted here for the representation of geometrical shapes of photonic nanostructures.
The VAE representation of geometrical shapes of the photonic nanostructures is a generated design space for the photonic structures. The geometric data for the nanostructures corresponds with a design space.
Liu page 5 second paragraph disclose “The encoder transforms the input geometric data into mean vectors µ and standard deviation vectors
σ
, and latent vectors v are sampled from the Gaussian distribution parametrized by µ and
σ
.” The entire distribution parameterized by the mean and standard deviation is a first design space. Liu page 6 last paragraph discloses “The simulator accepts the image of a nanostructure and approximates the four components of the transmittance of the metasurface.” The approximated components of transmittance correspond to a simulated response space.
Claim 18 further recites “train, utilizing the corresponding response space, a first multi-layer neural network to generate a reduced response space having reduced dimensionality compared to the corresponding response space, the first multi-layer neural network comprising an encoding layer and a decoding layer.” Liu page 9 second paragraph teaches “We also note that although the transmittance spectra are set as the design objective in these experiments, any photonic responses such as the diffraction behavior, optical chirality, and field localization can be used as the intended design criteria without further adjustment of our framework.”
Liu page 7 second paragraph discloses “objective is to minimize the
L
2
and/or
L
∞
norm of the difference between the input spectra and the approximated transmittance, a reasonable choice of the fitness function can be written as [equation (1)].”
Liu page 5 first paragraph discloses:
and latent vectors v are sampled from the Gaussian distribution parametrized by μ and σ. The decoder then reconstructs v back to the geometric information. When the training of this VAE is completed, the decoder can be operated as a geometric data generator as indicated in Fig. 1(b), so that when fed with a randomly sampled vector v, the corresponding pattern of the nanostructure s can be reconstructed.
The latent vectors are a reduced response space compared with the entire continuous Gaussian distribution which is sampled.
Liu page 14 figure 1 shows:
PNG
media_image1.png
200
400
media_image1.png
Greyscale
Liu page 14 figure 1 encoder and decoder combination corresponds with the first neural network.
Claim 18 further recites “train, utilizing the design space and the corresponding response space, a second neural network to generate a reduced design space having reduced dimensionality compared to the design space.” Liu page 5 first paragraph discloses:
and latent vectors v are sampled from the Gaussian distribution parametrized by μ and σ. The decoder then reconstructs v back to the geometric information. When the training of this VAE is completed, the decoder can be operated as a geometric data generator as indicated in Fig. 1(b), so that when fed with a randomly sampled vector v, the corresponding pattern of the nanostructure s can be reconstructed.
The latent space vectors v and randomly sampled vectors v are a reduced dimensionality as compared with the entire continuous Gaussian distribution defined by µ and
σ
.
Liu page 14 figure 1 b) decoder corresponds with a second neural network with a reduced dimensionality. Encircling and separating the decoder b) corresponds with generating the second neural network. Training of the first neural network correspond with a training of this second reduced neural network.
Claim 18 further recites “generate, by cascading the second neural network with the decoding layer of the first multi-layer neural network, an optimization convex hull.” Liu page 14 figure 1 caption teaches “(b) After the training, the decoder encircled in (a) can be separated and treated as a generator of geometric data.” Encircling and separating the decoder of the first neural network corresponds with cascading the second neural network with the decoding layer of the first multi-layer neural network.
But Liu does not explicitly disclose a convex hull; however, in analogous art of exploration of large-scale generate design spaces, Grossman column 12 lines 43-45 teach “GUI engine 210 may generate a convex hull that intersects design options 126 that minimize a total distance to both ranking axes.” Generating a convex hull intersecting design options corresponds with computing a first convex hull of data points of the design space. The generated convex hull which intersects the design options minimizing distance to the raking axes corresponds with an optimized convex hull.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Liu and Grossman. One having ordinary skill in the art would have found motivation to use generating a convex hull of design options into the system of discovery and design of photonic nanostructures for the advantageous purpose of visualizing and exploring large-scale generative design datasets “as well as various design characteristics associated with the different design options.” See Grossman column 3 lines 21-30.
Claim 18 further recites “and invert, using the optimization convex hull, the design space and the corresponding response space to generate the feasible optical response performance from the electromagnetic nanostructure.” Liu page 4 first paragraph last sentence teaches “scalability of the framework enables our approach to be implementable to the inverse design of complicated photonic structures with multiple dimensional parameters concurrently optimized.” An inverse design of the structure from the parameters is an inversion of the design space and the response space to generate the photonic structure from the respective transmittance spectra. See further Liu page 7 second paragraph fitness function.
Claim 19 further recites “19. The system of claim 18, wherein the instructions are further configured to cause the system to determine, by using the optimization convex hull, a designation of overlapping or non-overlapping of a desired design space of a desired structure.” From the above list of alternatives Examiner is selecting “overlapping.”
Liu does not explicitly disclose a convex hull; however, in analogous art of exploration of large-scale generate design spaces, Grossman column 12 lines 43-45 teach “GUI engine 210 may generate a convex hull that intersects design options 126 that minimize a total distance to both ranking axes.” Generating a convex hull intersecting design options corresponds with designating a desired space as overlapping with the generated convex hull.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Liu and Grossman. One having ordinary skill in the art would have found motivation to use generating a convex hull of design options into the system of discovery and design of photonic nanostructures for the advantageous purpose of visualizing and exploring large-scale generative design datasets “as well as various design characteristics associated with the different design options.” See Grossman column 3 lines 21-30.
Claim 20 further recites “20. The system of claim 18, wherein the electromagnetic simulation data comprises a multi-dimensional response space ranging from about 2-dimensional to about 6-dimensional.” Liu page 6 second paragraph discloses “the encoded vector v of the images by the VAE has the dimension of 10.” A ten dimensional latent space is close to a 2- to 6-dimensional response space. MPEP §2144.05 states “a prima facie case of obviousness exists where the claimed ranges or amounts do not overlap with the prior art but are merely close.”
A showing of unexpected results or criticality will rebut Examiner’s finding of obviousness here. MPEP §2144.05(III).
Dependent Claim 3
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over Liu and Grossman as applied to claim 1 above, and further in view of US patent 6,961,719 B1 Rai [herein “Rai”].
Claim 3 further recites “3. The system of claim 1, wherein the instructions are further configured to cause the system to validate, by using validation data (40), the optimization convex hull.” Neither Liu nor Grossman explicitly disclose validation data; however, in analogous art of neural network based multidimensional response surface modeling for selected objective functions, Rai column 7 lines 49-52 teach “If validation data are available, one can select the connection weight vector and resulting NN/SVM system with the smallest validation error.” Selecting a smallest validation error corresponds with using validation data to validate.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Liu, Grossman, and Rai. One having ordinary skill in the art would have found motivation to use validation data into the system of discovery and design of photonic nanostructures for the advantageous purpose of selecting between “different weight vectors may provide acceptably low training errors.” See Rai column 7 lines 45-52.
Claims 13, 14, 16, and 17
Claims 13, 14, 16, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over US patent 11,593,533 B2 Grossman, et al. [herein “Grossman”] in view of Liu, Z., et al. “A Hybrid Strategy for the Discovery and Design of Photonic Nanostructures” arXiv:1902.02293v2 [physics.optics] (Feb. 2019) [herein “Liu”].
Claim 13 further recites “13. A method comprising: identifying a first plurality of data points comprising a design space and a response space.” Grossman column 5 lines 12-16 teach “Generative design engine 200 processes problem definition 122 to generate design space 124. In particular, generative design engine 200 performs a generative design process based on problem definition 122 to generate each design option 126 included in design space 124.”
Grossman column 7 lines 22-26 disclose:
Design characteristics 430 includes single-attribute controllers 432 that can be manipulated by the user to change the values of various design characteristics. Each design characteristics generally corresponds to an emergent property associated with design options 126.
Emergent properties of the design options correspond broadly with a response space.
Grossman does not explicitly disclose an optical response performance for the response space; however, in analogous art of design of photonic nanostructures, Liu page 4 first paragraph last sentence teaches “the inverse design of complicated photonic structures with multiple dimensional parameters concurrently optimized.” The inverse design of photonic structures with optimized parameters corresponds with determining optical performance for the photonic structures.
Liu page 9 second paragraph teaches “We also note that although the transmittance spectra are set as the design objective in these experiments, any photonic responses such as the diffraction behavior, optical chirality, and field localization can be used as the intended design criteria without further adjustment of our framework.” The photonic response corresponds with an optical response performance.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Grossman and Liu. One having ordinary skill in the art would have found motivation to use design of photonic structures with the generative design system of Grossman for the advantageous purpose “[t]o automate the inverse design of optical structures with minimal intervention of human, and to discover arbitrarily-shaped photonic building blocks with minimal predefined restrictions.” See Liu page 3 last paragraph.
Claim 13 further recites “computing a first convex hull of all data points in the first plurality of data points.” Grossman column 12 lines 43-45 teach “GUI engine 210 may generate a convex hull that intersects design options 126 that minimize a total distance to both ranking axes.” Generating a convex hull intersecting design options corresponds with computing a first convex hull of data points of the design space.
Claim 13 further recites “merging the first convex hull with previous convex hulls to form an optimized convex hull, the previous convex hulls comprising a second plurality of data points comprising previous design spaces and previous response spaces.” Grossman column 12 lines 43-45 teach “GUI engine 210 may generate a convex hull that intersects design options 126 that minimize a total distance to both ranking axes.” Generating a convex hull which intersects design options correspond with a combination of design spaces and responses. Without loss of generality the spaces encompassed by at least some of the design options are considered previous spaces.
Claim 13 further recites “and determining, by the optimized convex hull, a feasible optical response performance within an electromagnetic nanostructure.” Grossman column 13 lines 16-18 teach “GUI engine 210 determines any technically feasible set of emergent properties for each design option 126.” The feasible emergent properties correspond with the feasible response performance.
Grossman does not explicitly disclose an optical response performance for the response space; however, in analogous art of design of photonic nanostructures, Liu page 4 first paragraph last sentence teaches “the inverse design of complicated photonic structures with multiple dimensional parameters concurrently optimized.” The inverse design of photonic structures with optimized parameters corresponds with determining optical performance for the photonic structures. Photonic structures of Liu correspond with electromagnetic nanostructures.
Liu page 9 second paragraph teaches “We also note that although the transmittance spectra are set as the design objective in these experiments, any photonic responses such as the diffraction behavior, optical chirality, and field localization can be used as the intended design criteria without further adjustment of our framework.” The photonic response corresponds with an optical response performance.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Grossman and Liu. One having ordinary skill in the art would have found motivation to use design of photonic structures with the generative design system of Grossman for the advantageous purpose “[t]o automate the inverse design of optical structures with minimal intervention of human, and to discover arbitrarily-shaped photonic building blocks with minimal predefined restrictions.” See Liu page 3 last paragraph.
Claim 14 further recites “14. The method of claim 13, further comprising: inverting, using the optimized convex hull, the design space and the response space to generate the feasible optical response performance from the electromagnetic nanostructure.” Grossman does not explicitly disclose an inverting the design space and response space; however, in analogous art of design of photonic nanostructures, Liu page 4 first paragraph last sentence teaches “scalability of the framework enables our approach to be implementable to the inverse design of complicated photonic structures with multiple dimensional parameters concurrently optimized.” An inverse design of the structure from the parameters is an inversion of the design space and the response space to generate the photonic structure from the respective transmittance spectra. See further Liu page 7 second paragraph fitness function.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Grossman and Liu. One having ordinary skill in the art would have found motivation to use inverse design of photonic structures with the generative design system of Grossman for the advantageous purpose “[t]o automate the inverse design of optical structures with minimal intervention of human, and to discover arbitrarily-shaped photonic building blocks with minimal predefined restrictions.” See Liu page 3 last paragraph.
Claim 16 further recites “16. The method of claim 13, wherein inverting, using the optimized convex hull, the design space and the response space further comprises designating a desired design space of a desired electromagnetic nanostructure as overlapping or non-overlapping the optimized convex hull.” From the above list of alternatives Examiner is selecting “overlapping.”
Grossman column 12 lines 43-45 teach “GUI engine 210 may generate a convex hull that intersects design options 126 that minimize a total distance to both ranking axes.” Generating a convex hull intersecting design options corresponds with designating a desired space as overlapping with the generated convex hull.
Claim 17 further recites “17. The method of claim 13, further comprising: merging, using the optimized convex hull and third plurality of data points from a desired electromagnetic nanostructure structure, a re-optimized convex hull when the desired electromagnetic nanostructure comprises a desired design space designated as non-overlapping.” For claim interpretation, MPEP §2111.04(II) Contingent Limitations states:
See Ex parte Schulhauser, Appeal 2013-007847 (PTAB April 28, 2016) for an analysis of contingent claim limitations in the context of both method claims and system claims. In Schulhauser, both method claims and system claims recited the same contingent step. When analyzing the claimed method as a whole, the PTAB determined that giving the claim its broadest reasonable interpretation, "[i]f the condition for performing a contingent step is not satisfied, the performance recited by the step need not be carried out in order for the claimed method to be performed" (quotation omitted). Schulhauser at 10.
(bold added).
Here, Grossman column 12 lines 43-45 teach “GUI engine 210 may generate a convex hull that intersects design options 126 that minimize a total distance to both ranking axes.” Generating a convex hull intersecting design options corresponds with designating a desired space as overlapping with the generated convex hull.
Because the generated convex hull of Grossman always intersects the design options, the claimed contingency of “when the desired electromagnetic nanostructure comprises a desired design space designated as non-overlapping” never occurs. Therefore, in accordance with Schulhauser Grossman teaches this contingent limitation method.
Dependent Claim 15
Claim 15 is rejected under 35 U.S.C. 103 as being unpatentable over Grossman and Liu as applied to claim 13 above, and further in view of US patent 6,961,719 B1 Rai [herein “Rai”].
Claim 15 further recites “15. The method of claim 13, wherein inverting, using the optimized convex hull, the design space and the response space further comprises applying a one-class support vector machine algorithm.” Grossman and Liu does not explicitly disclose a one-class support vector machine algorithm; however, in analogous art of neural network based multidimensional response surface modeling for selected objective functions, Rai column 6 lines 39-43 teaches:
In step 42, output signals from the hidden layer are computed to define the feature space for the [Support Vector Machine (SVM)]. The NN component of the system will provide appropriate combinations of the parameter space coordinates as new coordinates in a feature space for the SVM.”
See further Rai figure 3.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Grossman, Liu, and Rai. One having ordinary skill in the art would have found motivation to use a SVM feature space with the generative design system of Grossman for the advantageous purpose of “provides a nonlinear transformation from the input space variables pm into a feature space that contains the original variables.” See Rai column 5 lines 33-37 and column 6 lines 8-14.
Conclusion
Prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US 20210390396 A1 Fan; Jonathan Albert et al.
teaches
Generative Models for Design
Yao, H., et al. “Intelligent nanophotonics: merging photonics and artificial intelligence at the nanoscale” Nanophotonics, vol. 8, issue 3, pp. 339-366 (Jan. 2019)
Neural network inverse design of nanophotonics. DNN for design and characterization of metasurfaces.
Hinton, G.E. & Salakhutdinov “Reducing the Dimensionality of Data with Neural Networks” Science, vol. 313, pp. 504-507 (2006)
“backpropagation through deep autoencoders would be very effective for nonlinear dimensionality reduction”
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Jay B Hann whose telephone number is (571)272-3330. The examiner can normally be reached M-F 10am-7pm EDT.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Renee Chavez can be reached at (571) 270-1104. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/Jay Hann/Primary Examiner, Art Unit 2186 11 May 2026
Prosecution Insights