DETAILED ACTION
This nonfinal action is in response to application 17/948,639 filed on 09/20/2022 with priority to provisional application 63/246,015 filed on 09/20/2021.
Claims 1-29 are pending in the application. Claims 1, 15, and 29 are independent claims.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Information Disclosure Statement
The information disclosure statements (IDS) filed 09/01/2023 have been fully considered by the examiner.
Claim Objections
Claims 13, 14, 27 and 29 are objected to because of the following informalities:
In claims 13 and 27, “convolution neural network (CNN)” should read “convolutional neural network (CNN)” to correct an apparent typographical error.
In claims 14 and 28, “assess whether a location…; assess whether ground on…; and/or calculate a depth….” should read “perform at least one of: assessing whether a location…; assessing whether ground on…; and calculating a depth…” or be likewise amended to make clearer the intended scope of the alternative limitation construction.
In claim 29, “A method for reducing vertical bias and/or root mean square error (RMSE) of an elevation dataset” should read “A method for reducing at least one of vertical bias and root mean square error (RMSE) of an elevation dataset” or be likewise amended to make clearer the intended scope of the alternative limitation construction.
Appropriate corrections are required.
Applicant is advised that should claim 16 be found allowable, claim 25 will be objected to under 37 CFR 1.75 as being a substantial duplicate thereof. When two claims in an application are duplicates or else are so close in content that they both cover the same thing, despite a slight difference in wording, it is proper after allowing one claim to object to the other as being a substantial duplicate of the allowed claim (see MPEP § 608.01(m)).
Claim Rejections - 35 USC § 112
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 1-29 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 applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding claim 1, it recites the limitation “iteratively analyze the input data and learn nonlinear relationships between the input data and actual elevation” and “output a predicted elevation based on the analysis of the input data”,. Although the terms “actual elevation” and “predicted elevation” imply measurement of height of some object or region, the claim does not actually state what object or region the recited elevations are in reference to. It is therefore unclear what the relationship is between the recited elevations and the “input data”, because the claim does not further specify the received input data as comprising elevation data, or any particular type of data with respect to an object or region, and yet later recites outputting predicted elevations based on analysis of the input data, thereby implying some type of correlation that is left unclear. Consequently, one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
For purposes of examination and as best understood in light of the specification, the limitation “iteratively analyze the input data and learn nonlinear relationships between the input data and actual elevation” is interpreted as “for each location of a plurality of locations, iteratively analyze the input data and learn nonlinear relationships between the input data and an actual elevation of the location”, and the limitation “output a predicted elevation based on the analysis of the input data” is interpreted as “for each location of a plurality of locations, output a predicted elevation of the location based on the analysis of the input data”.
Regarding claim 6, it recites the limitation “using data from a NASA ICESat-2 mission as ground truth”. As is commonly understood in the art, “ground truth” is interpreted as referring to known, “correct” values for a set of training data, which are used to evaluate outputted model predictions and calculate error (e.g., via a loss function). However, the claim simply recites training a neural network “using” ground truth data, and does not, for example, recite training the model via calculating a difference between a predicted elevation and an actual elevation (i.e., ground truth), or otherwise provide adequate antecedent basis to clearly illustrate how the recited “ground truth” is interrelated with the existing claim elements (input data, actual elevation, etc). It is thereby unclear if the “ground truth” data is drawn from to determine “actual elevation” values, or is being used in some other manner. Consequently, one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
For purposes of examination and as best understood in light of the specification, the limitation is interpreted as “using data from a NASA ICESat-2 mission as the actual elevations of the locations”.
Regarding claim 7, it recites the limitation “wherein the NN is configured to predict error corrections for pixels on land between a minimum and maximum elevation”. As is commonly understood in the art, “pixels” is interpreted as referring to the smallest discrete units that make up an image displayed on a screen. However, the claim does not recite any type of digital display, and instead simply recites “pixels on land” without providing an explicit definition for the claim term, or adequately explaining the intended scope of a land pixel in the specification. It is thereby further unclear what is meant by the “minimum elevation” or “maximum elevation” of a land pixel. Consequently, one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
For purposes of examination and as best understood in light of the specification, the limitation is interpreted as “wherein the NN is configured to, for areas on land between a minimum and maximum elevation, predict error corrections”.
Regarding claim 8, it recites the limitation “The system according to claim 6, wherein the minimum elevation is -10 m, and the maximum elevation is 120 m”. However, the terms “minimum elevation” and “maximum elevation” have insufficient antecedent basis in the claim, because parent claim 6 does not previously recite a minimum or maximum elevation. Consequently, one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
For purposes of examination, claim 8 is interpreted as being dependent on claim 7 instead of claim 6.
Regarding claim 11, it recites the limitations “The system according to claim 9” and “generate the graphical map”. However, the term “graphical map” has insufficient antecedent basis in the claim, because parent claim 9 does not previously recite a graphical map. Consequently, one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
For purposes of examination, claim 11 is interpreted as being dependent on claim 10 instead of claim 9.
Regarding claim 12, it recites the limitation “further comprising a plurality of remote devices, each remote device configured to display a graphical map generated based on user input sent from the remote device”. However, parent claim 10 already recites an outputted graphical map – it is thereby unclear if claim 12 is referring to the same graphical map with further limitations, or is referencing a separate graphical map element. It is further unclear what is meant by “user input sent from the remote device”, because the claim does not specify to what element the user input is sent to for triggering generation of the graphical map, leaving the interrelation of the limitation with existing claim elements (e.g., one or processors) uncertain. Consequently, one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
For purposes of examination, the limitation is interpreted as “further comprising a plurality of remote devices, wherein for each remote device: the remote device is configured to send user input to the one or more processors, and display the graphical map outputted by the one or more processors, wherein the graphical map was generated based on the user input”.
Regarding claim 14, it recites the limitations “compare each data element of the digital elevation model to a water height or elevation; and for each data element, assess..”. However, the term “each data element” has insufficient antecedent basis in the claim, as parent claim 1 does not previous recite the digital elevation model comprising a plurality of data elements.
Claim 14 further recites “assess whether a location represented by the data element is at or below an elevation expected to flood or be inundated based on the water height or elevation”. However, the claim already previously recites comparing each data element to “a[n] elevation”; it is therefore unclear if the recited “elevation expected to flood” is referencing the previously recited elevation or a separate elevation element. The following limitations “assess whether ground on which an installed infrastructure or environment or planned infrastructure or environment at a location represented by the data element is at or below an elevation expected to flood or be inundated based on the water height or elevation” and “calculate a depth of a flood at a location represented by the data element based on the water height or elevation whether such water height or elevation is the result of a measurement, prediction, or flood model” share similar deficiencies. Further, these limitations appear to have substantial grammatical deficiencies (e.g., “ground on which an installed infrastructure or environment or planned infrastructure or environment at a location represented by the data element is at or below an elevation”, “a depth of a flood at a location represented by the data element based on the water height or elevation whether such water height or elevation is the result of a measurement”) that result in a lack of clarity with respect to the intended interrelation of claim elements, such that one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
For purposes of examination and as best understood in light of the specification, claim 14 is interpreted as follows:
The system according to claim 1, wherein the one or more processors are further configured to:
for each data element of a plurality of data elements within the digital elevation model, compare the data element to a particular water height or elevation; and perform at least one of:
assessing, based on the particular water height or elevation, whether a location represented by the data element is at or below an elevation expected to flood or be inundated,
assessing, based on the particular water height or elevation, whether the ground on which an installed infrastructure, environment, or planned infrastructure at a location represented by the data element is built, is at or below an elevation expected to flood or be inundated; and
calculating, based on the particular water height or elevation, wherein the particular water height or elevation is the result of a measurement, prediction, or flood model, a depth of a flood at a location represented by the data element.
Regarding claims 15, 21-22 and 28-29, they have substantially similar deficiencies to those found in claims 1, 6-7, and 14 above. Consequently, they are rejected for the same reasons and are interpreted as detailed above.
Regarding claims 2-5, 9-10, 13, 16-20, and 23-27, they inherit the deficiencies of their parent claims. Consequently, they are also rejected under 35 U.S.C. 112(b) as being indefinite for depending on an indefinite parent claim.
Applicant is advised to consider other parts of the disclosure, including the specification and drawings, for clarity and indefiniteness issues similar to those detailed above. The examiner notes that any amendments made to the specification, claims, or drawings must only contain subject matter that is supported by the originally filed disclosure in order to avoid being rejected under U.S.C. 112(a) as new matter (see MPEP § 608.04). Additionally, any interpretations of indefinite claim language detailed above are made solely for the purpose of examining the instant application on the merits, and are not attested as being adequately supported by the disclosure or adequately resolving all indefiniteness issues raised.
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 1-29 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The analysis of the claims will follow the 2019 Revised Patent Subject Matter Eligibility Guidance, 84 Fed. Reg. 50 (“2019 PEG”).
Independent Claims (Claim 1, Claim 15, Claim 29)
Step 1: Claim 1 is drawn to a system/apparatus, claim 15 is drawn to a method, and claim 29 is drawn to a method. Therefore, each of these claims falls under one of the four categories of statutory subject matter (process/method, machine/apparatus, manufacture/product, or composition of matter).
Step 2A Prong 1: Claims 1, 15, and 29 each recite a judicially recognized exception of an abstract idea.
Claim 1 recites, inter alia:
A system for creating [an] elevation model, comprising: iteratively analyze the input data; output a predicted elevation based on the analysis of the input data; and generate [an] elevation model based on the predicted elevation – These limitations amount to an abstract procedure of observing elevation data, using reasoning to draw conclusions from the observed data, making predictions on elevation levels based on the drawn conclusions, and then drawing a model of predicted elevation levels. They therefore recite a process of evaluation that a human could reasonably perform in the mind or using pen and paper.
learn nonlinear relationships between the input data and actual elevation – Within the established abstract procedure of analyzing data detailed above, this limitation further amounts to a procedural step of manipulating data through mathematical correlations (nonlinear relationships) between variables (input data and actual elevation), and thereby recites a mathematical relationship.
Claims 15 and 29 recite substantially similar abstract idea limitations to those found in claim 1, and therefore recite the same judicial exception.
Step 2A Prong 2: Claims 1, 15, and 29 do not recite any further additional elements besides those recited in the independent claims, and the following additional elements recited in claims 1, 15, and 29 also do not integrate the recited judicial exceptions into a practical application.
Claim 1 additionally recites:
[A system for creating] a digital elevation model; [generate] a digital elevation model – These limitations amount to mere instructions to implement an abstract idea “digitally”, i.e., on a computer or computer components.
one or more processors configured to: [receive/provide/generate] – This limitation amounts to mere instructions to implement an abstract idea on a computer or computer components.
receive input data; provide the input data to [a neural network] – These limitations amount to steps of merely gathering data to enable further analysis, and therefore recite insignificant extra-solution activity.
a neural network (NN), the NN comprising: an input layer; a plurality of hidden layers connected to the input layer, the plurality of hidden layers configured to [iteratively analyze]; an output layer connected to the plurality of hidden layers, the output layer configured to [output] – These limitations amount to mere instructions to apply an exception, as they do no more than generically invoke a neural network and its components as tools to perform an existing abstract process of analyzing data to draw conclusions and make predictions.
Claim 15 recites substantially similar additional elements to those recited in claim 1, and further recites:
[generate a digital elevation model] for one or more geographic locations – This limitation amounts to no more than generally linking a judicial exception to the technological environment of geographic modeling, as it simply ties the drawn model to being somehow representative of an existing geographic location without providing anything more.
Claim 29 recites substantially similar additional elements to those recited in claim 1, and further recites:
A method for reducing vertical bias and/or root mean square error (RMSE) of an elevation dataset, comprising: – The elements recited in the preamble simply state a purpose or intended use of the claimed procedure (reducing vertical bias and/or root mean square error (RMSE) of an elevation dataset) without having any clear correspondence to, or providing a distinct definition for, any of the limitations present within the body of the claim; they are thereby not considered to be of patentable significance (see MPEP § 2111.02) when considering eligibility of the claim.
Step 2B: The additional elements recited in claims 1, 15, and 29, viewed individually or as an ordered combination, do not provide an inventive concept or otherwise amount to significantly more than the recited abstract ideas themselves.
Claim 1 additionally recites:
[A system for creating] a digital elevation model; [generate] a digital elevation model – Mere instructions to implement an abstract idea “digitally”, i.e., on a computer or computer components, do not provide an inventive concept or significantly more to the recited abstract idea.
one or more processors configured to: [receive/provide/generate] – Mere instructions to implement an abstract idea on a computer or computer components do not provide an inventive concept or significantly more to the recited abstract idea.
receive input data; provide the input data to [a neural network] – Receiving and transmitting data is well-understood, routine, and conventional activity (see MPEP § 2106.05(d); “Receiving or transmitting data over a network”) and therefore does not provide an inventive concept or significantly more to the recited abstract idea.
a neural network (NN), the NN comprising: an input layer; a plurality of hidden layers connected to the input layer, the plurality of hidden layers configured to [iteratively analyze]; an output layer connected to the plurality of hidden layers, the output layer configured to [output] – Generically invoking a neural network and its components as tools to perform an existing abstract process does not provide an inventive concept or significantly more to the recited abstract idea.
Claim 15 recites substantially similar additional elements to those recited in claim 1, and further recites:
[generate a digital elevation model] for one or more geographic locations – Generally linking a judicial exception to the technological environment of geographic modeling does not provide an inventive concept or significantly more to the recited abstract idea.
Claim 29 recites substantially similar additional elements to those recited in claim 1, and further recites:
A method for reducing vertical bias and/or root mean square error (RMSE) of an elevation dataset, comprising: – Elements recited in the preamble that have no clear correspondence to, and do not provide a distinct definition for, any of the limitations present within the body of the claim, do not provide an inventive concept or significantly more to the recited abstract idea.
Even when considered as an ordered combination, the additional elements recited in the claims ultimately do no more than place the claims in the context of generically invoking a neural network to perform an abstract procedure.
As such, claims 1, 15, and 29 are not patent eligible.
Dependent Claims (Claims 2-14, Claims 16-28)
Dependent claims 2-14 and 16-28 narrow the scope of independent claims 1, 15, and 29, and thus merely narrow the recited judicial exceptions. With respect to the independent claims, the recited judicial exceptions are not meaningfully integrated into a practical application, and also do not amount to significantly more than the recited abstract ideas themselves. The dependent claims recite abstract idea limitations similar to those recited within the independent claims, as they also do not provide anything more than mathematical concepts or mental processes that are capable of being performed in the human mind and/or using pen and paper. The dependent claims also do not recite any further additional elements that successfully integrate the recited judicial exceptions into a practical application or amount to significantly more than the recited abstract ideas themselves. Consequently, claims 2-14 and 16-28 are also rejected under 35 U.S.C. 101.
Step 1: Claims 2-14 are drawn to a system/apparatus, and claims 16-28 are drawn to a method. Therefore, each of these claims falls under one of the four categories of statutory subject matter (process/method, machine/apparatus, manufacture/product, or composition of matter).
Step 2A Prong 1: Claims 2-14 and 16-28 each recite a judicially recognized exception of an abstract idea.
Claims 2-6 recite the same judicial exception as claim 1.
Claim 7 recites, inter alia:
predict error corrections for pixels on land between a minimum and maximum elevation – This limitation amounts to a process of using reasoning to update and improve elevation predictions based on observation of data, and therefore recites a process of evaluation that a human could reasonably perform in the mind or using pen and paper.
Claim 8 recites, inter alia:
wherein the minimum elevation is -10 m, and the maximum elevation is 120 m – Wherein parent claim 7 already recites a process of updating elevation predictions, this limitation merely sets a range of possible elevation levels, and therefore further recites a process of evaluation that a human could reasonably perform in the mind or using pen and paper.
Claim 9 recites the same judicial exception as claim 1.
Claim 10 recites, inter alia:
output a graphical map based on the digital elevation model – This limitation amounts to generating some type of graphical representation of spatial information from observing a drawn model, and therefore further recites a process of evaluation that a human could reasonably perform in the mind or using pen and paper.
Claim 11 recites, inter alia:
receive user input, and based on the user input, [generate] – This limitation amounts to merely listening to a user request and performing an action in response, and therefore recites a process of evaluation that a human could reasonably perform in the mind or using pen and paper.
generate the graphical map, where the graphical map shows root mean square error (RMSE) of the digital elevation model – This limitation amounts to manipulating data through mathematical methods (root mean square error) to produce output values, and therefore recites mathematical calculation.
Claim 12 recites, inter alia:
display a graphical map generated based on user input – This limitation amounts to, in response to a user request, creating some type of graphical representation of spatial information for display based on observation of a model, and therefore further recites a process of evaluation that a human could reasonably perform in the mind or using pen and paper.
Claim 13 recites the same judicial exception as claim 1.
Claim 14 recites, inter alia:
compare each data element of the digital elevation model to a water height or elevation – This limitation amounts to a process of generically drawing comparisons between data points based on observation, and therefore recites a process of evaluation that a human could reasonably perform in the mind or using pen and paper.
and for each data element,
assess whether a location represented by the data element is at or below an elevation expected to flood or be inundated based on the water height or elevation – This limitation amounts to making a prediction based on observation of data, and therefore recites a process of evaluation that a human could reasonably perform in the mind or using pen and paper.
assess whether ground on which an installed infrastructure or environment or planned infrastructure or environment at a location represented by the data element is at or below an elevation expected to flood or be inundated based on the water height or elevation – This limitation amounts to making a prediction based on observation of data, and therefore recites a process of evaluation that a human could reasonably perform in the mind or using pen and paper.
calculate a depth of a flood at a location represented by the data element based on the water height or elevation whether such water height or elevation is the result of a measurement, prediction, or flood model – This limitation amounts to using mathematical methods (calculate based on the water height or elevation) to determine an output variable (a depth of a flood), and therefore recites mathematical calculation.
Claim 16 recites substantially similar abstract idea limitations to those found in claim 10, and therefore recites the same judicial exception.
Claims 17-26 and 27-28 recite substantially similar abstract idea limitations to those found in claims 2-11 and 13-14 therefore recite the same judicial exceptions.
Step 2A Prong 2: Claims 7-8, 10, 14, 16, 22-23, 25, and 28 do not recite any further additional elements besides those recited in the independent claims, and the following additional elements recited in claims 2-6, 9, 11-13, 17-21, 24, and 26-27 also do not integrate the recited judicial exceptions into a practical application.
Claim 2 additionally recites:
wherein input data includes vegetation, architecture, and population density information for a plurality of locations – This limitation merely specifies a type of data to manipulated, and therefore recites insignificant extra-solution activity.
Claim 3 additionally recites:
wherein the plurality of hidden layers comprises at least a thousand hidden units – This limitation amounts to an insignificant step with regard to implementation of a neural network, wherein the neural network is merely being invoked as a tool to perform an abstract idea. It therefore recites insignificant extra-solution activity.
Claim 4 additionally recites:
wherein the input layer comprises at least 10 units corresponding to at least 2,000 values of the input data – This limitation amounts to an insignificant step with regard to implementation of a neural network, wherein the neural network is merely being invoked as a tool to perform an abstract idea. It therefore recites insignificant extra-solution activity.
Claim 5 additionally recites:
wherein the output layer comprises one unit – This limitation amounts to an insignificant step with regard to implementation of a neural network, wherein the neural network is merely being invoked as a tool to perform an abstract idea. It therefore recites insignificant extra-solution activity.
Claim 6 additionally recites:
wherein the NN is trained using data from a NASA ICESat-2 mission as ground truth – This limitation merely specifies a type of data to manipulated, and therefore recites insignificant extra-solution activity.
Claim 9 additionally recites:
wherein the input data comprises one or more datasets stored on a database operably coupled to at least one of the one or more processors – This limitation amounts to steps of merely gathering data to enable further analysis, and therefore recite insignificant extra-solution activity.
Claim 11 additionally recites:
where the graphical map shows predicted flood locations, vertical bias of the digital elevation model – This limitation amounts to generally linking a judicial exception to the fields of use of flood risk assessment and digital elevation modeling, as it does no more than specify an intended purpose of the produced graphical map without providing anything more.
Claim 12 additionally recites:
further comprising a plurality of remote devices, each remote device configured to [display]; [based on user input] sent from the remote device – This limitation does no more than provide a generic means (remote device) for obtaining user input, i.e., gathering data, and displaying a result to the user, i.e., outputting data; it therefore recites insignificant pre-solution and post-solution activity.
Claim 13 additionally recites:
wherein the NN is a convolution neural network (CNN) – This limitation amounts to mere instructions to apply an exception, as it does no more than generically invoke a convolutional neural network as a tool to perform an existing abstract process.
Claims 17-21, 24, 26, and 27 recite substantially similar additional elements to those found in claims 2-6, 9, 11, and 13, and therefore also do not integrate the recited judicial exception into a practical application.
Step 2B: The additional elements recited in claims 2-6, 9, 11-13, 17-21, and 24, viewed individually or as an ordered combination, do not provide an inventive concept or otherwise amount to significantly more than the recited abstract ideas themselves.
Claim 2 additionally recites:
wherein input data includes vegetation, architecture, and population density information for a plurality of locations – Considering topographic and demographic variables of a geographical region in evaluating sea-level rise exposure (i.e., flood risk assessment) is well-understood, routine, and conventional activity (see McMichael et al., “A review of estimating population exposure to sea-level rise and the relevance for migration”, [pages 1-2 Introduction]) and therefore does not provide an inventive concept or significantly more to the recited abstract idea.
Claim 3 additionally recites:
wherein the plurality of hidden layers comprises at least a thousand hidden units – Neural networks comprising thousands of neurons (i.e., units) is well-understood, routine, and conventional activity (e.g., deep networks) (see Liu et al., “Towards Better Analysis of Deep Convolutional Neural Networks”, [page 91 Introduction]) and therefore does not provide an inventive concept or significantly more to the recited abstract idea.
Claim 4 additionally recites:
wherein the input layer comprises at least 10 units corresponding to at least 2,000 values of the input data – Neural networks utilizing large amounts of input data (e.g., common benchmark dataset CIFAR10 – see Liu et al., “Towards Better Analysis of Deep Convolutional Neural Networks”, [page 96 Base CNN]), wherein the number of neurons (i.e., units) of the input layer corresponds to the dimensionality of the input data (see Kamali, “Deep Learning (Part 1) – Feedforward neural networks (FNN)”, [page 2, pages 8-9]), is well-understood, routine, and conventional activity, and therefore does not provide an inventive concept or significantly more to the recited abstract idea.
Claim 5 additionally recites:
wherein the output layer comprises one unit – The number of neurons (i.e., units) of the output layer corresponding to the type of learning being performed (e.g., one output neuron for binary classification or regression task) is well-understood, routine, and conventional activity (see Kamali, “Deep Learning (Part 1) – Feedforward neural networks (FNN)”, [page 2, pages 8-9, page 18]) and therefore does not provide an inventive concept or significantly more to the recited abstract idea.
Claim 6 additionally recites:
wherein the NN is trained using data from a NASA ICESat-2 mission as ground truth – Using lidar data (e.g., ICESat-2 data) for flood risk assessment is well-understood, routine, and conventional activity (see Muhadi et al., “The Use of LiDAR-Derived DEM in Flood Applications: A Review” [Abstract]) and therefore does not provide an inventive concept or significantly more to the recited abstract idea.
Claim 9 additionally recites:
wherein the input data comprises one or more datasets stored on a database operably coupled to at least one of the one or more processors – Storing and retrieving information in memory (e.g., database) is well-understood, routine, and conventional activity (see MPEP § 2106.05(d); “Storing and retrieving information in memory”) and therefore does not provide an inventive concept or significantly more to the recited abstract idea.
Claim 11 additionally recites:
where the graphical map shows predicted flood locations, vertical bias of the digital elevation model – Generally linking a judicial exception to the fields of use of flood risk assessment and digital elevation modeling does not provide an inventive concept or significantly more to the recited abstract idea.
Claim 12 additionally recites:
further comprising a plurality of remote devices, each remote device configured to [display]; [based on user input] sent from the remote device – Receiving and transmitting data over a network (i.e., across devices) is well-understood, routine, and conventional activity (see MPEP § 2106.05(d); “Receiving or transmitting data over a network”) and therefore does not provide an inventive concept or significantly more to the recited abstract idea.
Claim 13 additionally recites:
wherein the NN is a convolution neural network (CNN) – Generically invoking a convolutional neural network as a tool to perform an existing abstract process does not provide an inventive concept or significantly more to the recited abstract idea.
Claims 17-21, 24, 26, and 27 recite substantially similar additional elements to those found in claim 2-6, 9, 11, and 13, and therefore also do not provide an inventive concept or significantly more to the recited abstract idea.
Even when considered as an ordered combination, the additional elements recited in the claims ultimately do no more than place the claims in the context of generically applying an abstract procedure on a neural network to perform flood risk assessment.
As such, claims 2-14 and 16-28 also are not patent eligible.
Claim Rejections - 35 USC § 102
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1, 4-5, 7, 9-10, 15-16, 19-20, 22, 24-25, and 29 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Kulp et al. (“CoastalDEM: A global coastal digital elevation model improved from SRTM using a neural network”, available online 29 December 2017, cited in IDS filed 09/01/2023), hereinafter Kulp.
Regarding claim 1, Kulp discloses A system for creating a digital elevation model, (“This paper seeks to substantially advance the techniques introduced in (Wendi et al., 2016) by building a neural network that improves 1-arcsecond (roughly 30 m) horizontal resolution coastal SRTM elevations by employing a more comprehensive selection of inputs and far larger training and testing sets…We assess the post-correction error of the resulting digital elevation model, CoastalDEM, in both the US and Australia using lidar data from both countries that was not used for training, and use ICESat data to assess improvements worldwide (Section 3)” [Kulp page 232 Introduction]) comprising: one or more processors (“In this report, we chose a classical multilayer perceptron neural network model to prove the effectiveness of such a method in improving a DEM…Further, due to computational resource limitations, we chose a relatively small number of hidden layers and nodes in our network” [Kulp page 238 Discussion; It is implicit that the CoastalDEM model is generated on a computer with adequate processing capabilities for performing the disclosed functions) configured to:
receive input data; (“We use data from SRTM, NASA's ICESat mission, two vegetation indices, and the Landscan population database - introducing altogether a total of 23 variables - as input into a multi layer perceptron neural network” [Kulp page 232 Introduction])
provide the input data to a neural network (NN) ([Kulp page 232 Introduction] as detailed above), the NN (“Loosely inspired by the human brain, a traditional multilayer perceptron (MLP) neural network is made up of layers of nodes, or units, arranged in a directed weighted graph. The units in one layer are fully connected to the units in the next layer” [Kulp page 232 Neural network algorithm and architecture]) comprising:
an input layer; (“The sizes of the input…layers are strictly decided by the dimensionality of data (our system uses a 23-unit input layer for each of the 23 variables)” [Kulp page 232 Neural network algorithm and architecture])
a plurality of hidden layers connected to the input layer, the plurality of hidden layers configured to iteratively analyze the input data and learn nonlinear relationships between the input data and actual elevation; (“An artificial neural network (ANN, or NN) is a nonparametric computational model that can be used to fully automatically learn complex, highly nonlinear relationships between any number of different variables…The units in one layer are fully connected to the units in the next layer, starting with a single input layer and ending with a single output layer, and some number of intermediate (“hidden”) layers between them…The sizes of the input and output layers are strictly decided by the dimensionality of the data…while the number and sizes of hidden layers are hyperparameters that largely depend on the complexity
of the problem. Weights between units are learned through iterative backpropagation using a training set of samples with known answers, with the goal of minimizing the difference between predictions and actual answers” [Kulp page 232 Neural network algorithm and architecture]) and
an output layer connected to the plurality of hidden layers, the output layer configured to output a predicted elevation based on the analysis of the input data; (“For each sample, we compute a desired output, SRTM elevation
error” [Kulp page 234 Training and testing with lidar]; “The units in one layer are fully connected to the units in the next layer, starting with a single input layer and ending with a single output layer… The sizes of the input and output layers are strictly decided by the dimensionality of the data (our system uses…a 1-unit output layer for predicted SRTM error)” [Kulp page 232 Neural network algorithm and architecture]) and
generate a digital elevation model based on the predicted elevation (“We assess the post-correction error of the resulting digital elevation model, CoastalDEM, in both the US and Australia using lidar data from both countries that was not used for training” [Kulp page 232 Introduction]; “We apply the trained neural network to every coastal pixel in the 1–20 m SRTM domain in the world” [Kulp page 234 Global correction and assessment]; see Fig. 3 including Original SRTM and Adjusted SRTM – “Fig. 3. Samples of lidar elevation, original SRTM, and corrected SRTM elevation in Charleston, SC and Brisbane, Australia” [Kulp page 235]; It is implicit that the SRTM elevation errors output by the neural network, wherein the neural network is trained on every pixel within the SRTM domain, naturally generate an adjusted elevation model drawn from corrections to the original SRTM model).
Regarding claim 4, Kulp discloses the limitations of parent claim 1 and wherein the input layer comprises at least 10 units corresponding to at least 2,000 values of the input data (“(our system uses a 23-unit input layer for each of the 23 variables)” [Kulp page 232 Neural network algorithm and architecture]; “Training and testing datasets are derived from the coasts of the contiguous United States and Australia… Each pixel has a 1/100 probability of being chosen as a sample, and the resulting 59.3 million US data samples are randomly split into training (70%), validation (15%) and testing (15%)” [Kulp page 234 Training and testing with lidar])
Regarding claim 5, Kulp discloses the limitations of parent claim 1 and wherein the output layer comprises one unit. (“our system uses…a 1-unit output layer for predicted SRTM error)” [Kulp page 232 Neural network algorithm and architecture])
Regarding claim 7, Kulp discloses the limitations of parent claim 1 and wherein the NN is configured to predict error corrections for pixels on land between a minimum and maximum elevation (“Training and testing datasets are derived from the coasts of the contiguous United States and Australia. We must limit pixels targeted for adjustment, as we have empirically found that both training time and optimal NN architecture size grow considerably as the range of possible inputs widens. Here, we concentrate on 1-arcsecond land pixels >=1 m and <=20 m elevation which, given published SRTM error (< 10 m), should cover the vast majority of the low elevation coastal zone (land below 10 m)” [Kulp page 234 Training and testing with lidar]).
Regarding claim 9, Kulp discloses the limitations of parent claim 1 and wherein the input data comprises one or more datasets stored on a database operably coupled to at least one of the one or more processors (“We use data from SRTM, NASA's ICESat mission, two vegetation indices, and the Landscan population database - introducing altogether a total of 23 variables - as input into a multi layer perceptron neural network, training with ground truth elevation data from US coastal lidar data curated by NOAA (2015) (Section 2; Fig. 1)” [Kulp page 232 Introduction]).
Regarding claim 10, Kulp discloses the limitations of parent claim 1 and wherein the one or more processors is further configured to output a graphical map based on the digital elevation model (“Samples from the final improved DEM in Charleston, SC and Brisbane, Australia (Fig. 3) qualitatively suggest that the neural network flattens high frequency error in SRTM (reduces noise) and substantially reduces elevation in areas of large, positive error. A corresponding elevation error map of Charleston (Fig. 4) again suggests that SRTM struggles the most with areas of high population density (downtown Charleston, see yellow area in top left panel of Fig. 5 for location), and places of dense and tall vegetation (northeast of Charleston)” [Kulp page 236 Sample test cases]; see Fig. 4 including map of elevation error in original SRTM (top) and corrected SRTM (bottom) – “Fig. 4. Elevation error in original SRTM (top), and corrected SRTM (bottom) in Charleston, SC and Brisbane, Australia at elevations between 1 m and 20 m” [Kulp page 236])
Regarding claim 15, it is a method claim that largely corresponds to the system/apparatus of claim 1, which is already disclosed by Kulp as detailed above. Kulp further discloses generating a digital elevation model based on the predicted elevation for one or more geographic locations (see Fig. 3 – “Fig. 3. Samples of lidar elevation, original SRTM, and corrected SRTM elevation in Charleston, SC and Brisbane, Australia” [Kulp page 235]). Consequently, claim 15 is rejected for the same reasons as claim 1.
Regarding claims 16 and 25, they are method claims that largely correspond to the system/apparatus of claim 10, which is already disclosed by Kulp as detailed above. Consequently, they are rejected for the same reasons.
Regarding claims 19-20, 22, and 24, they are method claims that largely correspond to the systems/apparatuses of claims 4-5, 7, and 9, which are already disclosed by Kulp as detailed above. Consequently, they are rejected for the same reasons.
Regarding claim 29, it is a method claim that largely corresponds to the system/apparatus of claim 1, which is already disclosed by Kulp as detailed above. Kulp further discloses reducing vertical bias and/or root mean square error (RMSE) of an elevation dataset (“Our adjustment system reduces mean vertical bias in the coastal US from 3.67 m to less than 0.01 m, and in Australia from 2.49 m to 0.11 m. RMSE is cut by roughly one-half at both locations, from 5.36 m to 2.39 m in the US, and from 4.15 m to 2.46 in Australia” [Kulp Abstract]). Consequently, claim 29 is rejected for the same reasons as claim 1.
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.
Claims 2 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Kulp, as applied to claims 1 and 15 above, further in view of Kim et al. ("Simple-Yet-Effective SRTM DEM Improvement Scheme for Dense Urban Cities Using ANN and Remote Sensing Data: Application to Flood Modeling", published 14 Mar 2020), hereinafter Kim.
Regarding claim 2, Kulp teaches the limitations of parent claim 1 and wherein the input data includes vegetation and population density information for a plurality of locations (“Given a (1-arcsecond) SRTM target pixel, we encode a 23-dimensional vector as input into our neural network. This vector is made up of 9 values from SRTM 3.0 (1 arcsecond), 9 values from SRTM 2.1 (3 arcsecond), estimated slope, population density, a local ICESat error rate, tree canopy height, and vegetation density, each described below” [Kulp page 232 Neural network input layer]; “Training and testing datasets are derived from the coasts of the contiguous United States and Australia” [Kulp page 234 Training and testing with lidar]; see Fig. 3 – “Fig. 3. Samples of lidar elevation, original SRTM, and corrected SRTM elevation in Charleston, SC and Brisbane, Australia” [Kulp page 235]).
However, Kulp does not expressly teach wherein the input data includes architecture information for a plurality of locations (The examiner notes that under broadest reasonable interpretation in light of the specification (“The CNN was trained on high-quality global elevation data, using data from NASA's recent ICESat-2 mission, which covers land across the entire world. This choice was aimed at further improving performance in other countries where architecture and population density can be very different than what exists in the US” [¶ 0042]), Kulp may be reasonably interpreted as teaching the limitations of the claims on its own merits, given that the disclosed neural network is similarly trained on global elevation data (see Fig. 1 including Global Datasets – “Fig. 1. Flowchart visualization of adjustment system architecture. Global datasets, including SRTM 3.0, 2.1, Landscan, ICESat, and two vegetation indices, are used to make input samples to the neural network. Local lidar DEM's in the US and Australia are used to build training, validation, and testing sets with known actual SRTM error” [Kulp page 232]) wherein different countries (e.g., US and Australia) inherently have differences in architecture. Nevertheless, for the sake of completeness, the explicit teachings of an additional reference are further relied upon to cover the scope of the claim).
In the same field of endeavor, Kim teaches a means of applying a neural network to generate a digital elevation model (DEM) that improves accuracy over existing DEMs (“DEM data from a publicly accessible satellite, Shuttle Radar Topography Mission (SRTM), and Sentinel 2 multispectral imagery are selected and used to train the artificial neural network (ANN) to improve the quality of SRTM’s DEM… The trained ANN will then be ready to efficiently and effectively generate a high-quality DEM, at low cost, for places where ground truth DEM data is not available” [Kim Abstract]) wherein the data input to the ANN includes architecture information for a plurality of locations (In this paper, the performance of the DEM improvement scheme is evaluated over two dense urban cities, Nice (France) and Singapore” [Kim -Abstract]; “This paper presents significant improvements to the SRTM DEM using an ANN with remote sensing data. The improvement is particularly significant for dense urban areas. Figure 2 demonstrates the schematic diagram of our DEM improvement methodology. Generally, it requires four types of data; multispectral imagery, the DEM to be improved (SRTM DEM in this study), the building footprint for sorting the building areas, and a reference DEM (ground truth elevation)” [Kim page 2 Introduction]; see Figure 2 including Required Data –> Building Footprint (Open Street Map) – “Figure 2. Schematic diagram of DEM improvement methodology” [Kim page 3]; “Two separate ANN trainings were used, one for buildings only and the other for the entire area without buildings. Buildings were classified with building footprints from the Open Street Map (OSM). The elevation of iSRTM DEM was then calculated from the process of ANN” [Kim page 6 Artificial Neural Network Setup]; “The iSRTM DEMwais [sic] obtained from two ANN trainings, one with and one without building heights” [Kim page 7 Proof of Concept and Application of the Approach]; For training the ANN model to generate an improved quality DEM for dense urban cities (e.g., locations including Nice and Singapore), building heights (i.e., architecture information), drawn from building footprints, are included as input data)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein the input data includes architecture information for a plurality of locations as taught by Kim into Kulp because they are both directed towards applying neural networks to generate digital elevation models (DEM) that improve accuracy over existing DEMs. Given that Kulp already discusses applicability of the disclosed system to urban areas (“The methods presented here are flexible and effective, and can be effectively applied to land cover of all types, including dense urban development” [Kulp Abstract]), and flexibility in including additional input variables to the neural network (“This method offers flexibility in input parameters, and can be re-applied to improve future global DEM releases from NASA or other sources. Additionally, performance improvements are not limited to forested regions, as were previous SRTM enhancement efforts, but rather work across all types of land cover, including urban areas” [Kulp page 238 Summary and conclusions]), a person of ordinary skill in the art would recognize the value of incorporating the teachings of Kim to thereby improve performance of the system in said urban areas, with further application to improved accuracy in flood maps drawn from the generated model (“In this paper, the performance of the DEM improvement scheme is evaluated over two dense urban cities, Nice (France) and Singapore; with the performance criteria using various matrices, e.g., visual clarity, scatter plots, root mean square error (RMSE) and flood maps. The DEM resulting from the improved SRTM (iSRTM) showed significantly better results than the original SRTM DEM, with about 38% RMSE reduction. Flood maps from iSRTM DEM show much more reasonable flood patterns than SRTM DEM’s flood map” [Kim Abstract]).
Regarding claim 17, it is a method claim that largely corresponds to the system/apparatus of claim 2, which is already taught by the combination of Kulp and Kim as detailed above. Consequently, it is rejected for the same reasons.
Claims 3, 13-14, 18, and 27-28 are rejected under 35 U.S.C. 103 as being unpatentable over Kulp, as applied to claims 1 and 15 above, further in view of Meadows et al. (“A Comparison of Machine Learning Approaches to Improve Free Topography Data for Flood Modeling”, published 14 Jan 2021, cited in IDS filed 09/01/2023), hereinafter Meadows.
Regarding claim 3, Kulp teaches the limitations of claim 1 and wherein the plurality of hidden layers comprises hidden units (“Our neural network contains 3 hidden layers, with 10, 20, and 10 hidden units respectively” [Kulp page 232 Neural network algorithm and architecture]).
However, Kulp does not expressly teach wherein the plurality of hidden layers comprises at least a thousand hidden units.
In the same field of endeavor, Meadows teaches a means of applying a neural network to generate a digital elevation model (DEM) that improves accuracy over existing DEMs (“Given the high financial and institutional cost of collecting and processing accurate topography data, many large-scale flood hazard assessments continue to rely instead on freely-available global Digital Elevation Models, despite the significant vertical biases known to affect them. To predict (and thereby reduce) these biases, we apply a fully-convolutional neural network (FCN), a form of artificial neural network originally developed for image segmentation which is capable of learning from multi-variate spatial patterns at different scales. We assess its potential by training such a model on a wide variety of remote-sensed input data (primarily multi-spectral imagery), using high-resolution, LiDAR-derived Digital Terrain Models published by the New Zealand government as the reference topography data. In parallel, two more widely used machine learning models are also trained, in order to provide benchmarks against which the novel FCN may be assessed” [Meadows Abstract]) wherein the plurality of hidden layers of the neural network comprises at least a thousand hidden units (“The second machine learning algorithm tested here is of the same type used in the two most recent studies exploring this issue [5,21]–a densely-connected neural network,… This type of model consists of an input layer (with 23 neurons in this case, corresponding to pixel values from each of the selected input features), an output layer (one neuron, the prediction of the correction required for that particular pixel), and a variable number of hidden layers in between… As shown in Figure 7, three hidden layers were selected for this model, respectively containing 3074, 4089 and 1222 neurons (the optimal values found during hyperparameter tuning)” [Meadows pages 11-12 Densely-Connected Neural Networks (DCN)])
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein the plurality of hidden layers comprises at least a thousand hidden units as taught by Meadows into Kulp because they are both directed towards applying neural networks to generate digital elevation models (DEM) that improve accuracy over existing DEMs. Given that Kulp explicitly suggests that adoption of more sophisticated neural network models could allow for improvements in pattern learning (“In this report, we chose a classical multilayer perceptron neural network model to prove the effectiveness of such a method in improving a DEM. More specialized, sophisticated neural network models could allow for deeper and more abstract pattern learning” [Kulp page 238 Discussion]), yet also recognizes the limitations of computational resource constraints (“Further, due to computational resource limitations, we chose a relatively small number of hidden layers and nodes in our network” [Kulp page 238 Discussion]) a person of ordinary skill in the art would recognize the value of incorporating the teachings of Meadows to improve model performance via increased model complexity (“Of the three machine learning techniques tested here, performance was found to improve with increasing model complexity, with the simplest model (RF) reducing test RMSE by 55%, the more complex DCN outperforming that (60%), and the most complex model (FCN) going even further (71%)” [Meadows page 22 Discussion]), without requiring as significant an increase in computational resource requirements as an even more complex model, e.g., the FCN model alternately disclosed in Meadows. Doing so would also require less modification to the existing system than incorporating, e.g., the FCN model, given that the disclosed DCN model is explicitly recognized as being of the same type as the NN model of Kulp ([Meadows page 11 Densely-Connected Neural Networks (DCN)] as detailed above – note that Kulp is the cited reference (see [5] on [page 26 References]) discussed in Meadows).
Regarding claim 13, Kulp teaches the limitations of claim 1. Meadows further teaches wherein the NN is a convolution neural network (CNN) (“To predict (and thereby reduce) these biases, we apply a fully-convolutional neural network (FCN), a form of artificial neural network originally developed for image segmentation which is capable of learning from multi-variate spatial patterns at different scales” [Meadows Abstract]; see Figure 8 – “Figure 8. The FCN architecture used here, a simplification of the U-net model developed by Ronenberger et al. [36], showing how the model is able to learn at different spatial scales” [Meadows page 12]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein the NN is a convolution neural network (CNN) as taught by Meadows into Kulp because they are both directed towards applying neural networks to generate digital elevation models (DEM) that improve accuracy over existing DEMs. Incorporating the teachings of Kulp would improve performance of the system in improving DEM accuracy due to the ability of FCNs to learn from spatial patterns at multiple scales rather than only a pixel-by-pixel basis (“Of the three machine learning techniques tested here, performance was found to improve with increasing model complexity, with the simplest model (RF) reducing test RMSE by 55%, the more complex DCN outperforming that (60%), and the most complex model (FCN) going even further (71%). This is likely due to its ability to learn from spatial patterns at multiple scales, which represents a significant advantage over the other models, which learn on a pixel-by-pixel basis including only very rudimentary information about their spatial context (such as slope values)” [Meadows page 22 Discussion]).
Regarding claim 14, Kulp teaches the limitations of claim 1. Meadows further teaches compar[ing] each data element of the digital elevation model to a water height or elevation (“..we made an initial assessment of performance on flood-related objectives by looking at test residuals according to flood susceptibility zone [39] and by deriving a map of Height Above Nearest Drainage (HAND) values [68]. As shown in Figure 17, performance across these different flood-related classifications did not vary greatly, with the possible exception of the lowest HAND range (grid cells for which the resampled LiDAR topography was within a 2 m vertical distance of the nearest drainage cell), where FCN corrections achieved a reduction in RMSE of 77% (compared to 71% more broadly)” [Meadows page 20 Performance on Flood Related Objectives]; Derivation of height above nearest drainage (HAND) values involves calculating vertical distance between a point on land and its nearest drainage cell (e.g., stream, river), i.e., comparing each pixel (i.e., data element) of the DEM model to a water height); and
for each data element,
assess whether a location represented by the data element is at or below an elevation expected to flood or be inundated based on the water height or elevation (see Figure 17 including: a) By flood susceptibility, wherein pixels (i.e. data elements) of the DEM are split into Flood-prone [70,886] and Not flood-prone [29,482 pixels], and b) By height above nearest drainage (HAND), wherein pixels (i.e., data elements) are split into HAND classes (0-2 m, 2-5 m, 5-10 m, 10-20 m, >20 m), with lower HAND values indicating a higher risk of flooding [Meadows page 20]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein the one or more processors are further configured to: compare each data element of the digital elevation model to a water height or elevation; and for each data element, assess whether a location represented by the data element is at or below an elevation expected to flood or be inundated based on the water height or elevation as taught by Meadows into Kulp because they are both directed towards applying neural networks to generate digital elevation models (DEM) that improve accuracy over existing DEMs. Given that Kulp already discusses applicability of the generated DEM model to evaluating flood risk (“With these improvements, the corrected global coastal elevation model presented here should be a valuable asset for next-generation flood risk modeling projects” [Kulp page 238 Discussion]), a person of ordinary skill in the art would recognize the value of incorporating the teachings of Meadows to enable a means of initially evaluating model performance on flood-related objectives (“While flood hazard modeling using the various corrected DEMs is beyond the scope of this study, we made an initial assessment of performance on flood-related objectives” [Meadows page 20 Performance on Flood-Related Objectives]).
The examiner notes that the above limitations, as mapped to Kulp and Meadows, comprise a selection within the recited alternative expression (see MPEP § 2143.03) and thereby cover the entire scope of the claim. Nevertheless, it is further acknowledged, for the sake of completeness, that the additionally cited references Kim and Vernimmen (see attached PTO-892 Notice of References Cited) further teach assess[ing] whether ground on which an installed infrastructure or environment or planned infrastructure or environment at a location represented by the data element is at or below an elevation expected to flood or be inundated based on the water height or elevation (see Figure 9 – “Figure 9. Flood map comparisons between (a) Flood map with SRTM DEM and (b) Flood map with iSRTM DEM in dense urban area of Nice, France” and Figure 10 – “Figure 10. Flood map comparisons between (a) Flood map with SRTM DEM and (b) Flood map with iSRTM DEM in dense urban area of Singapore” [Kim page 10]; “Figures 9 and 10 illustrate the inundated areas in Nice, France, and Singapore respectively. Different DEMs, SRTM DEM and iSRTM DEM, were used in the flood model to investigate the flood patterns. Flood maps resulting from iSRTM DEM capture the flooding in the flood prone areas, i.e., roads/low-lying areas, while flood maps resulting from the original SRTM DEM do not. Due to the coarse resolution of SRTM DEM (30 30 m) and inaccurate terrain elevation, flood patterns do not follow the real topographic characteristics” [Kim page 11]) and calculat[ing] a depth of a flood at a location represented by the data element based on the water height or elevation whether such water height or elevation is the result of a measurement, prediction, or flood model (“An indication of the impact of differences between DTMs and GDEMs on assessing flood risk is obtained by demonstrating the depth of inundation that would occur if water levels are 2 m above current MSL (Figure 9). Where the local TOPODEM DTM and GLL_DTM_v1 yield limited land area with a flood depth more than 2 m, at 0.5% and 4.7% of the Mekong Delta extent, respectively, this is 24.8%, 1.8%, 79.1%, and 0.3% for SRTM90, MERIT, CoastalDEM, and TanDEM-X. Such large differences are meaningful to flood risk assessments as risk is not only determined from flood extent but also the depth of inundation. Flood depths of 2 m or more are generally considered life-threatening, whereas depths of 1 m or less will mostly damage infrastructure and crops” [Vernimmen page 13 Impact of DTM Accuracy on Flood Risk Assessments]; see Figure 9 – “Figure 9. Flood depth map with 2 m high water level applied to 6 elevation maps for the Mekong Delta. (a) TOPODEM, (b) GLL_DTM_v1, (c) TanDEM-X, (d) SRTM90, (e) CoastalDEM, (f) MERIT” [Vernimmen page 13]) respectively, and would be likewise combinable to one of ordinary skill in the art.
Regarding claims 18 and 27-28, they are method claims that largely correspond to the systems/apparatuses of claims 3 and 13-14, which are already taught by the combination of Kulp and Meadows as detailed above. Consequently, they are rejected for the same reasons.
Claims 6 and 21 are rejected under 35 U.S.C. 103 as being unpatentable over Kulp, as applied to claims 1 and 15 above, further in view of Vernimmen et al. ("New ICESat-2 Satellite LiDAR Data Allow First Global Lowland DTM Suitable for Accurate Coastal Flood Risk Assessment", published 31 Aug 2020), hereinafter Vernimmen.
Regarding claim 6, Kulp teaches the limitations of claim 1 and wherein the NN is trained using lidar data as ground truth (“We use data from SRTM, NASA's ICESat mission, two vegetation indices, and the Landscan population database - introducing altogether a total of 23 variables - as input into a multi layer perceptron neural network, training with ground truth elevation data from US coastal lidar data curated by NOAA (2015)(Section 2; Fig. 1)” [Kulp page 232 Introduction]; “For each sample, we compute a desired output, SRTM elevation error. To assess this error in the US and Australia, we use a similar approach to (Hofton et al., 2006), employing high-quality elevation models based on lidar as a baseline topography (ground truth)” [Kulp page 234 Training and testing with lidar]).
However, Kulp does not expressly teach using data from a NASA ICESat-2 mission.
In the same field of endeavor, Vernimmen teaches a means of improving accuracy over existing digital elevation models for assessing coastal flood risk (“No accurate global lowland digital terrain model (DTM) exists to date that allows reliable quantification of coastal lowland flood risk, currently and with sea-level rise. We created the first global coastal lowland DTM that is derived from satellite LiDAR data…It is accurate within 0.5 m for 83.4% of land area below 10 m above mean sea level (+MSL), with a root-mean-square error (RMSE) value of 0.54 m, compared to three local area DTMs for three major lowland areas: the Everglades, the Netherlands, and the Mekong Delta. This accuracy is far higher than that of four existing global digital elevation models (GDEMs), which are derived from satellite radar data, namely, SRTM90, MERIT, CoastalDEM, and TanDEM-X” [Vernimmen Abstract]) using data from a NASA ICESat-2 mission (“The global LiDAR lowland DTM (GLL_DTM_v1) at 0.05-degree resolution (~5 5 km) is created from ICESat-2 data collected between 14 October 2018 and 13 May 2020” [Vernimmen Abstract]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated using data from a NASA ICESat-2 mission as taught by Vernimmen into Kulp because they are both directed towards improving accuracy over existing digital elevation models for assessing coastal flood risk. Incorporating the teachings of Vernimmen would enable training of the neural network on a global level via a highly accurate dataset (“The low errors and high consistency across regions in GLL_DTM_v1 compared to radar-based GDEMs demonstrates that the ICESat-2 satellite LiDAR data provide a unique opportunity to improve global lowland area mapping with substantially higher accuracy than has been possible so far and, therefore, will allow more accurate environmental and climate impact assessments relying on such data products” [Vernimmen page 10 Impact of DTM Accuracy on Lowland Extent Estimation]) rather than solely on US and Australian locations, which would further alleviate concerns of overfitting as already discussed in Kulp (“To alleviate the concerns of overfitting in the United States, we use the secondary Australian testing set to further validate our methods” [Kulp page 234 Training and testing with lidar]).
Regarding claim 21, it is a method claim that largely corresponds to the system/apparatus of claim 6, which is already taught by the combination of Kulp and Vernimmen as detailed above. Consequently, it is rejected for the same reasons.
Claims 8 and 23 are rejected under 35 U.S.C. 103 as being unpatentable over Kulp, as applied to claims 7 and 22 above, further in view of Vernimmen ("New ICESat-2 Satellite LiDAR Data Allow First Global Lowland DTM Suitable for Accurate Coastal Flood Risk Assessment", published 31 Aug 2020) and Meadows (“A Comparison of Machine Learning Approaches to Improve Free Topography Data for Flood Modeling”, published 14 Jan 2021, cited in IDS filed 09/01/2023).
Regarding claim 8, Kulp teaches the limitations of parent claim 7. Vernimmen further teaches wherein the minimum elevation is -10 m (“The global LiDAR lowland DTM (GLL_DTM_v1) at 0.05-degree resolution (~5 5 km) is created from ICESat-2 data collected between 14 October 2018 and 13 May 2020. It is accurate within 0.5 m for 83.4% of land area below 10 m above mean sea level (+MSL)…Globally, we find 3.23, 2.12, and 1.05 million km2 of land below 10, 5, and 2 m +MSL” [Vernimmen Abstract]). Meadows further teaches wherein the maximum elevation is 120 m (“As shown in Figure 3, the eight sites available include coastal and inland areas, covering elevations ranging from 0 to 1242 metres above mean sea level” [Meadows page 6 Reference Topography Data]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein the minimum elevation is -10 m as taught by Vernimmen into Kulp because they are both directed towards improving accuracy over existing digital elevation models for assessing coastal flood risk. Given that Kulp already discusses applicability of the disclosed system to flood risk assessment for coastal regions (“With these improvements, the corrected global coastal elevation model presented here should be a valuable asset for next-generation flood risk modeling projects” [Kulp page 238 Summary and conclusions]), a person of ordinary skill in the art would recognize the value of incorporating the teachings of Vernimmen to thereby expand considered coastal areas to include those at lowest elevations, i.e., elevations below mean sea level at high risk of flooding (“We determine land areas in the low elevation coastal zone (LECZ), defined as the contiguous area along the coast that is below 10 m +MSL [29,30]. We also present accuracy and extent numbers for land below 2 m, which is most susceptible to river floods and inundation from tropical storms [31], and below 5 m as this is around the maximum rate of sea-level rise projected by 2300 in a high carbon emission scenario [2], which can put lowlands at risk of permanent inundation” [Vernimmen page 2 Introduction]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein the maximum elevation is 120 m as taught by Meadows into the combination because both Kulp and Meadows are directed towards applying neural networks to generate digital elevation models (DEM) that improve accuracy over existing DEMs. Given that Kulp already discusses applicability of the disclosed system to urban areas (“The methods presented here are flexible and effective, and can be effectively applied to land cover of all types, including dense urban development” [Kulp Abstract]), a person of ordinary skill would recognize the value of incorporating the teachings of Meadows to enable expansion of the model to a wider variety of regions, e.g., dense urban areas with tall buildings (“More recently, the deeper DCN used by Kulp and Strauss [5] to correct SRTM elevations in coastal zones (defined as areas with SRTM elevations between 1–20 m) achieved RMSE reductions of 41% and 55% (for two different testing zones, both much larger than those considered here…It is worth noting that their model specifically targeted low-lying coastal zones (i.e., trained only on data from those areas), whereas all models trained in the current study learnt from a much wider variety of input data (including elevations above 1200 m) and so should be more widely applicable” [Vernimmen page 24 Comparison of Test Results Obtained Here with Other Published Work]).
Regarding claim 23, it is a method claim that largely corresponds to the system/apparatus of claim 8, which is already taught by the combination of Kulp, Vernimmen, and Meadows as detailed above. Consequently, it is rejected for the same reasons.
Claims 11-12 and 26 are rejected under 35 U.S.C. 103 as being unpatentable over Kulp, as applied to claims 10 and 25 above, further in view of Rossi et al. ("Evaluating FLOODIS: Mobile Sensing for a Flood Emergency Service in the Cloud", published 30 Nov 2015), hereinafter Rossi.
Regarding claim 11, Kulp teaches the limitations of parent claim 10 and generat[ing] the graphical map, where the graphical map shows vertical bias of the digital elevation model, or root mean square error (RMSE) of the digital elevation model (“The ICESat-baselined global error maps are presented in Fig. 6. Approximately 68% of 1 degree cells see median absolute error reduced. Performance varies spatially, with apparent overcompensation more common in areas where original SRTM deviations are small or negative, such as parts of northern Africa, northern Europe, and northeast Asia. Stratified median global bias is reduced from 1.90 m to 0.24 m (an 87% improvement), and stratified RMSE is reduced from - 4.17 m to 3.10 m (a 26% improvement)” [Kulp page 236 ICESat assessment]; see Figure 6 – “Fig. 6. Median SRTM deviations from ICESat measurements at elevations from 1 m through 20 m over 1-degree cells worldwide, both before correction (top) and after (bottom)” [Kulp page 238]).
However, Kulp does not expressly teach wherein the one or more processors is further configured to receive user input, and based on the user input, generate the graphical map and where the graphical map shows predicted flood locations (it is noted by the examiner that although Kulp teaches the entire scope of the recited alternative limitation construction, for the sake of completeness, an additional reference is further acknowledged to teach remaining possible selections).
In the same field of endeavor, Rossi teaches a means of utilizing digital elevation models for flood risk assessment (“The European Copernicus Programme for Earth Observation, “Copernicus EMS” [2] (formerly known as GIO-EMS, GMES Initial Operations Emergency Management Service) provides geographical mapping services of natural disasters, with accurate geospatial information derived from satellite observations and complemented with in situ or open data sources, whenever available. Copernicus satellites support a wide range of applications and they can collect many measurables for different natural disasters, also providing detailed Digital Elevation Models (DEM) [3] of the earth…To overcome the aforementioned limitations we propose FLOODIS: a novel service that provides a faster, flexible and scalable flood emergency system…We improve flood extent map updates by creating flood nowcast maps, and we also compute flood forecasts though a novel algorithm” [Rossi page 1 Introduction]; “Within FLOODIS, we are mainly interested in getting from Copernicus EMS delineation maps, which provide an assessment of the flood event extent” [Rossi page 2 Copernicus EMS]; “The goal of the FLOODIS Flood Forecast Backend (F3B), which is a component of the GEO Gateway, is to periodically generate both flood nowcast and forecast map layers in the form of shapefiles, so as to allow their ingestion as with Copernicus delineation layers. The flood forecast is computed considering post-processed river flow data from the EFAS [7] Sensor Observation Service (SOS), the EU-DEM [3], the CORINE [9] Land Cover, and Open Street Map (OSM) river information” [Rossi page 3 Flood Nowcast and Forecast Service]) wherein the one or more processors is further configured to receive user input, and based on the user input, generate a graphical map (“To cope with usage bursts, we implement a cloud-based Service Oriented Architecture (SOA). We break down FLOODIS in different sub-systems in a view of allowing each one of them to be deployed independently, automatically scaling according to the usage. The main components of FLOODIS are the following. The GEO Gateway that implements the interfaces with all existing European emergency systems like Copernicus EMS and EFAS [7] in order to provide both flood extent, flood nowcast, and flood forecast maps… The Service Layer (SL) that acts as the FLOODIS centralized service provider, implementing (i) web services to receive/provide geolocalized flood Reports from/to the mobile application” [Rossi page 3 System Architecture]; “The core functionality of the GEO Gateway is to ingest, process, store and distribute flood related map layers, including delineation, nowcast…The provision of the flood delineation maps to the final users is then achieved by using the open-source map server software GeoServer, that implements Open Geospatial Consortium (OGC) web services. We select the web map tiled service (WMTS) as it provides for caching of the tiles that are requested, greatly reducing the load on the spatial database. Note that all flood map layers are delivered with WMTS, and are requested both by the FLOODIS Service Layer and the MAs…. We report in Figure 4a the architecture of the Geo Gateway, in which we specify all its sub-components, the main standards adopted and the information flows” [Rossi page 3 GEO Gateway]; “In order to implement the Service Layer (SL), we use the well-known 3-tier programming pattern. This pattern divides the program into three logically and architecturally different layers: the Data Access (DA), the Business Logic (BL) and the Presentation (P) layer. The DA deals with the interaction with the database, the BL provides for common algorithms and processing, while P implements the User Interface together with the client-side scripts and validation rules” [Rossi page 4 Service Layer and Augmentation Module]; see Figure 4 – “Figure 4: Geo Gateway architecture (a) and Service Layer architecture (b)” [Rossi page 4]; The Service Layer, via the User Interface, processes requests from mobile devices (i.e., user input), and sends these requests to the Geo Gateway, which generates and distributes the flood reports (i.e., graphical maps)) based on a digital elevation model, ([Rossi page 1 Introduction] and [Rossi page 2 Copernicus EMS] and [Rossi page 3 Flood Nowcast and Forecast Service] as detailed above) where the graphical map shows predicted flood locations ([Rossi page 1 Introduction] as detailed above; Flood forecast maps implicitly show predicted flood locations).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein the one or more processors is further configured to receive user input, and based on the user input, generate a graphical map, where the graphical map shows predicted flood locations as taught by Rossi into Kulp because they are both directed towards utilizing digital elevation models for flood risk assessment. Given that Kulp already discusses applicability of the generated DEM model to evaluating flood risk (“With these improvements, the corrected global coastal elevation model presented here should be a valuable asset for next-generation flood risk modeling projects” [Kulp page 238 Discussion]), a person of ordinary skill in the art would recognize the value of incorporating the teachings of Rossi to enable usage of the generated DEM model in a real-time flood emergency system (“We stress that due to lacking of real-time flood emergency systems, prompt alerting and actuation of evacuation procedures represents a huge challenge for Civil Protections (CPs) and Disaster Management Centers (DMCs), who struggle to avoid causalities despite of the huge efforts of their emergency teams. To overcome the aforementioned limitations we propose FLOODIS: a novel service that provides a faster, flexible and scalable flood emergency system” [Rossi page 1 Introduction]).
Regarding claim 12, Kulp teaches the limitations of parent claim 10. Rossi further teaches a plurality of remote devices, each remote device configured to display a graphical map generated based on user input sent from the remote device (“To cope with usage bursts, we implement a cloud-based Service Oriented Architecture (SOA). We break down FLOODIS in different sub-systems in a view of allowing each one of them to be deployed independently, automatically scaling according to the usage. The main components of FLOODIS are the following. The GEO Gateway that implements the interfaces with all existing European emergency systems like Copernicus EMS and EFAS [7] in order to provide both flood extent, flood nowcast, and flood forecast maps… The Service Layer (SL) that acts as the FLOODIS centralized service provider, implementing (i) web services to receive/provide geolocalized flood Reports from/to the mobile application” [Rossi page 3 System Architecture]; “The core functionality of the GEO Gateway is to ingest, process, store and distribute flood related map layers, including delineation, nowcast…The provision of the flood delineation maps to the final users is then achieved by using the open-source map server software GeoServer, that implements Open Geospatial Consortium (OGC) web services. We select the web map tiled service (WMTS) as it provides for caching of the tiles that are requested, greatly reducing the load on the spatial database. Note that all flood map layers are delivered with WMTS, and are requested both by the FLOODIS Service Layer and the MAs…. We report in Figure 4a the architecture of the Geo Gateway, in which we specify all its sub-components, the main standards adopted and the information flows” [Rossi page 3 GEO Gateway]; “In order to implement the Service Layer (SL), we use the well-known 3-tier programming pattern. This pattern divides the program into three logically and architecturally different layers: the Data Access (DA), the Business Logic (BL) and the Presentation (P) layer. The DA deals with the interaction with the database, the BL provides for common algorithms and processing, while P implements the User Interface together with the client-side scripts and validation rules” [Rossi page 4 Service Layer and Augmentation Module]; see Figure 4 – “Figure 4: Geo Gateway architecture (a) and Service Layer architecture (b)” [Rossi page 4]; The Service Layer, via the User Interface, processes requests from mobile devices (i.e., user input sent from remote devices), and sends these requests to the Geo Gateway, which generates and distributes the flood reports (i.e., graphical maps) that are sent to user’s mobile devices for display)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated a plurality of remote devices, each remote device configured to display a graphical map generated based on user input sent from the remote device as taught by Rossi into Kulp because they are both directed towards utilizing digital elevation models for flood risk assessment. Given that Kulp already discusses applicability of the generated DEM model to evaluating flood risk (“With these improvements, the corrected global coastal elevation model presented here should be a valuable asset for next-generation flood risk modeling projects” [Kulp page 238 Discussion]), a person of ordinary skill in the art would recognize the value of incorporating the teachings of Rossi to enable usage of the generated DEM model in a real-time flood emergency system (“We stress that due to lacking of real-time flood emergency systems, prompt alerting and actuation of evacuation procedures represents a huge challenge for Civil Protections (CPs) and Disaster Management Centers (DMCs), who struggle to avoid causalities despite of the huge efforts of their emergency teams. To overcome the aforementioned limitations we propose FLOODIS: a novel service that provides a faster, flexible and scalable flood emergency system” [Rossi page 1 Introduction]).
Regarding claim 26, it is a method claim that largely corresponds to the system/apparatus of claim 11, which is already taught by the combination of Kulp and Rossi as detailed above. Consequently, it is rejected for the same reasons.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Liu et al. ("Towards Better Analysis of Deep Convolutional Neural Networks", published 09 Aug 2016) discloses a visual analytics approach for better understanding, diagnosing, and refining deep convolutional neural networks (CNNs).
McMichael et al. ("A review of estimating population exposure to sea-level rise and the relevance for migration", published 27 Nov 2020) discloses a review of publications analyzing estimates of population exposure to sea-level rise and related hazards (e.g., coastal flood events).
Muhadi et al. ("The Use of LiDAR-Derived DEM in Flood Applications: A Review", published 18 July 2020) discloses a review of known applications of LiDAR-derived digital elevation models (DEM) in flood studies, aiming to provide insight into the operating principles of different LiDAR systems and the advantages and disadvantages of each system.
Kamali ("Deep Learning (Part 1) - Feedforward Neural Networks (FNN)", published 28 Apr 2021) discloses a review of fundamental concepts underlying the architecture of feedforward neural networks (FNN).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to VIJAY M BALAKRISHNAN whose telephone number is (571) 272-0455. The examiner can normally be reached 10am-5pm EST Mon-Thurs.
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, JENNIFER WELCH can be reached on (571) 272-7212. 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.
/V.M.B./
Examiner, Art Unit 2143
/JENNIFER N WELCH/Supervisory Patent Examiner, Art Unit 2143
Prosecution Insights