DETAILED ACTION
This action is responsive to Applicant’s reply filed 12 December 2025. This action is made non-final.
Status of the Claims
Claims 1, 8 and 15 are currently amended.
Claims 5-7 and 12-14 are canceled.
Claim status is currently pending and under examination for claims 1-4, 8-11 and 15, of which independent claims are 1, 8 and 15.
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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on September 29, 2025 has been entered.
Response to Amendment
Applicant’s arguments regarding the art rejections are moot in view of the new grounds of rejection necessitated by applicant’s amendment.
Claim Rejections - 35 USC § 103
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 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 1, 3-4, 8, 10-11 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Wang et al. (US 20150356402 A1), hereinafter Wang (‘402), in view of Wang et al. (US 8214182 B2), hereinafter Wang (‘182), further in view of Luo et al. (“Automated Visual Defect Classification for Flat Steel Surface”), hereinafter Luo, and Nehmadi et al. (US 20060269120 A1), hereinafter Nehmadi, and further in view of Fulton et al. (US 6772821 B1), hereinafter Fulton, and Mizota et al. (JP 4536197 B2), hereinafter Mizota.
With respect to Claim 1, Wang (‘402) teaches:
A prediction system implemented on a computing apparatus (Wang (‘402’) discloses “the system includes … a processing unit (for example, one or more microprocessors) and a computer-readable medium that has computer-readable program code embodied therein. The computer-readable medium is cooperative with the input, output and processing unit to operate as an ANN to provide the substantially real-time prediction” [0009].)
and configured to train a model for predicting a defect of a target product, the prediction system comprising (Wang (‘402) discloses “a neural network system to provide substantially real-time prediction of at least one of a residual stress and distortion of a quenched aluminum casting is disclosed. The system includes an input configured to receive data relating to at least one of topological features, geometrical features and quenching process parameters associated with the casting, an information output configured to convey data relating to at least one of the residual stress and distortion of the aluminum casting predicted by the system” [0009].):
processing circuitry configured to: receive, from a storage device, information on an existing product including (Wang (‘402) discloses “The system includes an input configured to receive data relating to at least one of topological features, geometrical features and quenching process parameters associated with the casting … a processing unit (for example, one or more microprocessors) and a computer-readable medium that has computer-readable program code embodied therein. The computer-readable medium is cooperative with the input, output and processing unit to operate as an ANN to provide the substantially real-time prediction” [0009].):
a defect characteristic value indicating a defect … in the existing product (Wang (‘402) discloses “computational simulation is one alternate way to predict residual stresses, where analytical or numerical methods can be used in place of the mechanical or non-destructive approaches mentioned above. Finite element analysis (FEA) is one conventional numerical approach, where the large-scale partial differential equations that explain the mechanics of continuous medium can be modeled as an aggregate of discrete points within the medium. One such system that performs residual stress and distortion predictions with a good accuracy can be found in U.S. Pat. No. 8,214,182 entitled METHODS OF PREDICTING RESIDUAL STRESSES AND DISTORTION IN QUENCHED ALUMINUM CASTINGS” [0005].
Wang (‘402) discloses analytical residual stress and distortion data (‘defect characteristic values’) is obtained from a FEA model, “aluminum casting analysis approach is achieved through the use of artificial neural networks (ANNs, also referred to herein more simply as neural networks) and more particularly, multilayer feedforward (MFF, also referred to as feedforward) neural network models that use analytical residual stress and distortion data predicted by a sophisticated FEA model together with part (a) geometry information …” [0007].),
shape information indicating a three-dimensional shape of the existing product (Wang (‘402) discloses “aluminum casting analysis approach is achieved through the use of artificial neural networks (ANNs, also referred to herein more simply as neural networks) and more particularly, multilayer feedforward (MFF, also referred to as feedforward) neural network models that use analytical residual stress and distortion data predicted by a sophisticated FEA model together with part (a) geometry information such as curvature and maximum dihedral angle” [0007].
Wang (‘402) further discloses “in neural network 100, an input layer 200 is made up of numerous casting- and quench-specific features, including casting topological features 210, quenching process features 220 and casting geometric features 230. Output layer 400 includes both three-dimensional distortion information 410 and maximum principal stress information 420 of the cast component” [0017].),
and conditional information indicating a manufacturing condition of the existing product (Wang (‘402) discloses manufacturing conditions as quench process parameters, “neural network models that use analytical residual stress and distortion data predicted by a sophisticated FEA model together with part (a) geometry information … and (c) quench parameters such as quench temperature and quench media” [0007].);
train a defect prediction model using the defect characteristic value, [shape features], and the manufacturing condition of the existing product (Wang (‘402) discloses “aluminum casting analysis approach is achieved through the use of artificial neural networks (ANNs, also referred to herein more simply as neural networks) and more particularly, multilayer feedforward (MFF, also referred to as feedforward) neural network models that use analytical residual stress and distortion data predicted by a sophisticated FEA model together with part (a) geometry information such as curvature and maximum dihedral angle, (b) topological (i.e., topographic) features such as nodal neighbor topologies and (c) quench parameters such as quench temperature and quench media” [0007].
Wang (‘402) further discloses “In neural network 100, an input layer 200 is made up of numerous casting- and quench-specific features, including casting topological features 210, quenching process features 220 and casting geometric features 230” [0017].),
the defect prediction model including a machine learning model configured to associate spatial shape features and manufacturing parameters with … defect predictions (See [0017-0018] of the Wang (‘402) disclosure discussing training a neural network with geometric features (‘spatial shape features’) and quenching process features (‘manufacturing parameters’) to predict residual stress and distortion (‘defect’) of a quenched aluminum-based casting. By training a neural network with geometric and quenching process features, a neural network learns to associate these features with stress and distortion.);
input a feature … representing a three-dimensional shape of the target product to the trained defect prediction model (Wang (‘402) discloses “an input layer 200 is made up of numerous casting- and quench-specific features, including casting topological features 210, quenching process features 220 and casting geometric features 230. Output layer 400 includes both three-dimensional distortion information 410” [0017]. See Figures 3B and 4B depicting predicted residual stress and distortions, respectively, on a section of cast aluminum part (‘target product’) by using a trained neural network.);
output, via the defect prediction model, a defect characteristic value indicating a defect … in the target product (See Figures 3B and 4B depicting predicted residual stress and distortions, respectively, on a section of cast aluminum part (‘target product’). The predictions are a result of using a trained neural network, see [0030].);
However, Wang (‘402) does not teach a defect associated with a location, which Wang (‘182) does:
a defect associated with a location (Wang (‘182) discloses “the computer readable-medium, using the calculated changes in strain and strain rates, may then calculate at least one of a residual stress and distortion at the respective integration points of the dimensional elements of the respective virtual zones of the virtual aluminum casting to predict at least one residual stress and distortion of the quenched aluminum casting” (Col. 18, lines 22-27).
Wang (‘182’) further discloses “In quenching, the induced residual stresses and distortion in aluminum castings generally are due to differences in cooling rates and, thus, temperatures and thermal contraction rates from location to location of the aluminum casting geometric structure. In accordance with embodiments, a strain and stress analysis module predicts at least one of residual stresses and distortion in quenched aluminum castings. Using a strain and stress analysis, the residual stresses and distortion may be predicted through incorporating thermal strains induced in the quenching with a nonlinear constitutive behavior of as-quenched microstructures of aluminum castings, the thermal strains generally arising from non-uniform transient temperature distributions across the up to entire sections of an aluminum casting during quenching. The transient temperature distribution of aluminum castings during quenching may be calculated based on surface node-specific, element-specific, and/or zone-specific heat transfer coefficients” (Col. 9, lines 15-32).
Wang (‘182) further discloses “Data representing the aluminum casting is provided to enable predictions of residual stresses and/or distortion of an aluminum casting determined through simulation and computation by the embodiments. As such, an aluminum casting is provided having a defined geometric structure, composition, and material properties. The exterior surface of the geometric structure of the aluminum casting may be divided into a plurality of nodes, elements, and/or zones that may be affected differently by a quenching process thereof. For example, the zones may comprise a top surface, a bottom surface, and a side surface of the aluminum casting” (Col. 9, line 61 to Col. 10, line 4).).
Wang (‘182) teaches using strain and strain rates to calculate residual stress and distortions in virtual zones of an aluminum casting is a known method in the art. Wang (‘182) further teaches that due to differences in cooling temperatures in different zones of a casting, strains may occur that lead to residual stress and distortions. Wang (‘402) teaches that the technique used by Wang (‘182) accurately performs residual stress and distortion predictions. Before the effective filing date, it would have been obvious to a person of ordinary skill in the art to combine the artificial neural network disclosed by Wang (‘402) for predicting residual stress and distortion in aluminum castings with the technique of Wang (‘182) to accurately predict the zones where residual stress and distortion occurs. By including zone quench temperatures in residual stress and distortion prediction, the strains derived from the quenching process can be used to accurately predict zones where residual stress and distortion occur, which can lead to discovering underlying issues in the casting process related to zones and cooling.
Furthermore, the combination of Wang (‘402) in view of Wang (‘182) does not teach inputting shape information to a shape-processing model and outputting a feature vector representing the shape of the existing product which Luo does:
input the shape information to a shape-processing model implemented using a convolutional neural network (Luo discloses Figure 7 on P. 9342 (reproduced below) depicting defect classification by a convolutional neural network (CNN). Images depicting three dimensional defects (‘shape information’) on flat steel surfaces are input to a CNN (‘shape-processing model’).
PNG
media_image1.png
546
1362
media_image1.png
Greyscale
Luo discloses “a CNN was adopted as a classifier for the defect classification of flat steel surfaces, and the experimental results indicate that the CNN is powerful and robust for the classification task. Generally, a dense layer is placed at the end of the network that executes the final classification or regression based on the extracted features, and the schematic of defect classification by CNN is shown in Fig. 7” (P. 9341, Sec. VII-A, Second Paragraph).),
the shape-processing model configured to perform a convolution operation on the three-dimensional shape (Luo discloses Figure 7 depicting pooling layers where images of defects are pooled (‘perform a convolution operation’) to extract features.)
and output a feature vector representing the shape of the existing product (Luo discloses a CNN processing images of defects (‘shape of an existing product’) outputs extracted features (‘feature vector’), “a dense layer is placed at the end of the network that executes the final classification or regression based on the extracted features” (P. 9341, Sec. VII-A, Second Paragraph).);
train a … [neural network layer] using … the feature vector output from the shape-processing model … (Luo discloses Figure 7 depicting fully connected layers (a dense layer) at the end of the convolutional neural network. The fully connected layers take the extracted features (‘feature vector output’) from the CNN (‘shape-processing model’) to classify defects. It is implied that the fully connected layers are trained if they are able to output defect classifications.),
input a feature vector representing a three-dimensional shape of the target product to the trained … [neural network layer] (See Figure 7 depicting inputting extracted features (‘feature vector’) into fully connected neural network layers to classify defects in images of flat steel surfaces (‘three-dimensional shape of a target product’).);
Luo teaches inputting images of flat steel surfaces (‘shape information’) into a convolutional neural network (‘shape-processing model’) to classify defects is a known method in the art. Before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to combine the artificial neural network disclosed by Wang (‘402) for predicting residual stress and distortion in aluminum castings with the convolutional neural network of Luo to extract features from three-dimensional shapes. By using a CNN to extract features from a three-dimensional shape, a CNN can efficiently extract the most important features since CNNs are designed to recognize and learn complex features. Furthermore, by training a model with features generated by a CNN, less computational resources would be used since the model would be trained on lower-dimension data thus reducing model complexity.
Furthermore, the combination of Wang (‘402) in view of Wang (‘182) and in further view of Luo does not teach displaying a defect associated with a location in a target product, which is taught by Nehmadi:
and display, via a display device, the defect characteristic value indicating the defect associated with the location in the target product (Nehmadi discloses “as previously described, for some embodiments, different dice may be printed with different focus and exposure combinations (e.g., creating an FEM wafer). For such embodiments, different software simulations may be performed to predict defect locations for dies printed with different combination of focus and exposure levels. Different sample sets may then be selected according to the techniques described above” [0063].
Nehmadi further discloses “For some embodiments, corresponding defect maps similar to those described above with reference to FIG. 7 may be generated. For example, a GUI may be provided that allows a user to quickly display a defect map showing the location on a wafer or die of detected, but not predicted defects, as well as locations for which defects were predicted, but not detected” [0065].).
Nehmadi teaches a GUI displaying a defect map showing the predicted defect locations of a wafer is a known method in the art. Before the effective filing date, it would have been obvious to a person of ordinary skill in the art to combine the artificial neural network disclosed by Wang (‘402) for predicting residual stress and distortion in aluminum castings with the method of Nehmadi to provide a visual representation of defects for users. By providing a visual representation to users on a display device, users are able to easily see the exact locations of the predicted defects on a product, and can therefore make reasonable guesses or take actions to find the root cause of the predicted defects.
Furthermore, the combination of Wang (‘402) in view of Wang (‘182) and in further view of Luo and Nehmadi does not teach molten metal type, molten metal temperature, die temperature, die surface treatment, cycle time, spray application amount, and spray time manufacturing conditions, which are taught by Fulton:
wherein, when the target product is a casting, the manufacturing condition includes all of molten metal type, molten metal temperature, …, die temperature, die surface treatment, cycle time, … spray application amount, spray time … (Fulton discloses metal alloy chemical composition (‘molten metal type’), “Prior to manufacture of the die castings, a desired chemical composition for the metal or metal alloy and a desired shot profile are selected. Typically, a process designer employed by the manufacturer selects values for these physical parameters as those values which are believed to minimize the likelihood that die castings, manufactured under those physical parameters, would contain defects. Thus, deviations from the selected values for these physical parameters are deemed as increasingly the likelihood that die castings manufactured under such conditions are more likely to contain defects” (Col. 6, lines 57-67).
Fulton discloses “Physical parameters affecting die casting integrity and surface quality are of primary concern since it is these factors which are generally considered to affect the occurrence of defects in die castings. In the past, the physical parameters which were deemed as affecting die casting integrity and surface quality included metal or metal alloy temperature, die lube spray, fast shot velocity and intensification pressure. For the development of the disclosed processes, the physical parameters deemed as affecting die casting integrity and surface quality were expanded to include die steel chemistry, die steel toughness, die steel hardness, die steel polishing, heat treatment of the die steel, die temperature, alloy cleanliness, alloy gas content, porosity level of the manufactured die castings, vacuum level applied to the die cavity, in-cavity metal pressure, die lube dilution ratio, die lube flow rate, die spray pattern, and amount of plunger lube on a per shot basis” (Col. 18, line 61 to Col. 19, line 10).
Metal alloy temperature (‘molten metal temperature’), die temperature, die lube spray (‘die surface treatment’), and die lube dilution ratio / amount of plunger lube on a per shot basis (‘spray application amount’) are physical parameters (‘manufacturing conditions’) that affect die casting integrity and surface quality.
Fulton discloses “some of the physical parameters suitable for inclusion in the series of physical parameters to be measured each time that a die casting is formed during a die casting machine cycle include die ejector plate temperature, die cover plate temperature, die cavity pressure, die lube ratio, die lube spray volume per shot, die spray pattern, die spray time, shot profile (which, as previously set forth, includes slow shot velocity, fast shot velocity, transition time and intensification pressure), total die casting machine cycle time” (Col. 11, Lines 6-15).).
Fulton teaches physical parameters that affect die casting integrity and surface quality are known in the art. Before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to combine the artificial neural network of Wang (‘402) with the physical parameters disclosed by Fulton to train a machine learning model using physical parameters. By training a machine learning model with physical parameters, the trained model can be used to predict casting defects, thereby allowing manufacturers to identify issues early and adjust physical parameters to reduce the likelihood of producing defective castings.
Furthermore, the combination of Wang (‘402) in view of Wang (‘182), in further view of Luo and Nehmadi, and in further view of Fulton does not teach water flow time, die time, die opening sequence, and air blow sequence manufacturing conditions, which are taught by Mizota:
wherein, when the target product is a casting, the manufacturing condition includes all of … water flow time, … die time, die opening sequence, …, spray time, and air blow sequence (Mizota discloses “As casting conditions that affect the mold temperature, for example, there are various flow rates such as the flow rate of cooling water flowing through the mold, the injection speed of the molten metal, the spraying time of the release agent sprayed onto the mold, and the present invention, Casting is performed such that the mold temperature approaches the target temperature by actively changing the casting conditions that affect these mold temperatures” [0024].
To calculate the flow rate of cooling water flowing through a mold, water flow time must be used in the calculation. Therefore, the cooling water flow rate includes water flow time.
Mizota discloses “if the mold temperature is lower than the target temperature, the solidification rate of the injected molten metal will be too fast, causing problems such as not supplying a sufficient amount of molten metal to the thin part of the product, resulting in a defective product” [0003].
Mizota discloses a good casting can be obtained from optimizing mold opening time (‘die opening sequence’), “Furthermore, even if the mold temperature deviates somewhat from the target temperature, it is possible to obtain a good cast product by optimizing the injection conditions such as the injection speed, mold opening time, and operation timing of the pressure pin” [0004].
Mizota discloses adjusting die time, “the die time defines the time (cycle time) from the start of casting by the die casting machine 2 to the completion of casting. The longer the die time, the longer the time that the molten metal exists in the mold, so the mold temperature rises. … Further, if the die time is long, the time until the next casting operation is started becomes long, and if the die temperature is lower than the target temperature, the die temperature may be further lowered. For this reason, the die time value is shortened” [0043].
Mizota discloses air blowing time (‘air blow sequence’) affects casting quality, “The air spraying time defines the spraying time of compressed air for the purpose of removing moisture remaining on the surface of the mold by the above-described spraying of the release agent. This air blowing time is an example of casting conditions that affect the quality of the casting of the present invention. This is because if the casting is performed in a state where residual moisture exists, the quality of the cast product is deteriorated” [0048].).
Mizota teaches casting conditions that affect casting quality are known in the art. Before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to combine the artificial neural network of Wang (‘402) with the casting conditions disclosed by Mizota to train a machine learning model using casting conditions. By training a machine learning model with casting conditions, the trained model can be used to predict casting defects, thereby allowing manufacturers to identify issues early and adjust casting conditions to reduce the likelihood of producing defective castings.
With respect to claim 3, the combination of Wang (‘402) in view of Wang (‘182), in further view of Luo and Nehmadi, and in further view of Fulton and Mizota teaches:
the prediction system according to claim 1, wherein the defect characteristic value includes a value indicating a degree of the defect of the target product (Wang (‘402) discloses Figures 3B and 4B (reproduced below) depicting prediction results from an artificial neural network. Figure 3B depicts the predicted residual stress (‘defect characteristic value’) results of a section of a cast aluminum part (‘target product’) on a contour plot. Figure 4B depicts the predicted distortion (‘defect characteristic value’) results of the same section of the cast aluminum part on a contour plot. The predicted residual stresses and distortions are represented as shadings on a cast aluminum part, which indicate the areas the defects span, and therefore their severity.
PNG
media_image2.png
541
1327
media_image2.png
Greyscale
See [0030] (discussing Figures 3B and 4B).).
With respect to claim 4, the combination of Wang (‘402) in view of Wang (‘182), in further view of Luo and Nehmadi, and in further view of Fulton and Mizota teaches:
the prediction system according to claim 1, wherein, when the target product is a casting, the defect prediction model is further trained by at least one of a die volume of the casting, a casting volume, a casting surface area, or a thickness of the casting (Wang (‘182) discloses “the heat transfer module calculates a plurality of heat transfer coefficients specific to the respective virtual surface nodes, elements, and/or zones and calculates a plurality of virtual node-specific, element-specific, and/or zone-specific temperatures using the heat transfer coefficients, the virtual node-specific, element-specific, and/or zone-specific temperatures respectively specific to a time of the simulated quenching. The strain and stress analysis module calculates a total strain of at least a node, an element, and/or a zone in the aluminum casting using the virtual node-specific, element-specific, and/or zone-specific temperatures and a coefficient of thermal expansion/contraction. The strain and stress analysis module also calculates a plurality of thermal stresses and strains at integration points defining the dimensional elements of the respective virtual aluminum casting and calculates a strain rate and a change in strain at the respective integration points. The material constitutive model defined in the user material subroutine module which is in communication with strain and stress analysis module calculates at least one of a residual stress and distortion at the respective integration points to predict at least one residual stress and distortion of the aluminum casting” (Col. 3, lines 1-26).
Wang (‘182) further discloses “the virtual surface zones of the virtual aluminum casting may comprise at least one top surface of the virtual aluminum casting and at least one bottom surface of the virtual aluminum casting respective to quench orientation of the aluminum casting. The virtual surface zones may respectively comprise a plurality of nodes and dimensional elements on the surfaces respectively defined by a length (x), a width (y), and a depth (z)” (Col. 3, lines 48-55).
Wang (‘182) further discloses “Data representing the aluminum casting is provided to enable predictions of residual stresses and/or distortion of an aluminum casting determined through simulation and computation by the embodiments. As such, an aluminum casting is provided having a defined geometric structure, composition, and material properties. The exterior surface of the geometric structure of the aluminum casting may be divided into a plurality of nodes, elements, and/or zones that may be affected differently by a quenching process thereof. For example, the zones may comprise a top surface, a bottom surface, and a side surface of the aluminum casting” (Col. 9, line 61 to Col. 10, line 4).).
Wang (‘182) teaches collecting temperatures of zones comprised of a plurality of casting surfaces to calculate strains and stresses at integration points. The strains and stresses are then used to calculate and predict residual stresses and distortions. Before the effective filing date, it would have been obvious to a person of ordinary skill in the art to combine the artificial neural network disclosed by Wang (‘402) for predicting residual stress and distortion in aluminum castings with the technique of Wang (‘182) to predict location-specific residual stresses and distortions. By using zones defined by a surface’s length, width, and depth, the locations where residual stress and distortion occurs can be known, which can help reveal manufacturing errors in the quenching process specific to that zone.
With respect to claim 8, the rejection of claim 1 is incorporated. The difference in scope being:
A prediction method implemented by a computing apparatus for training a model for (Wang (‘402) discloses “a computer-implemented method of rapidly predicting at least one of residual stress and distortion of a quenched aluminum casting is disclosed. The method includes receiving computer input data corresponding to at least one of topological features, geometrical features and quenching process parameters associated with the casting, and then operating the computer as a neural network to determine output data corresponding to at least one of residual stress and distortion values” [0008].).
With respect to claim 10, the claim recites similar limitations corresponding to claim 3, therefore the same rationale of rejection is applicable.
With respect to claim 11, the claim recites similar limitations corresponding to claim 4, therefore the same rationale of rejection is applicable.
With respect to claim 15, the rejection of claim 1 is incorporated. The difference in scope being
a non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to (Wang (‘402) discloses “a computer-readable medium that has computer-readable program code embodied therein. The computer-readable medium is cooperative with the input, output and processing unit” [0009].).
Claims 2 and 9 are rejected under 35 U.S.C. 103 as being unpatentable over Wang (‘402) in view of Wang (‘182), further in view of Luo and Nehmadi, further in view of Fulton and Mizota, and in further view of Nieves Acedo (EP 3316061 A1), hereinafter Nieves Acedo.
With respect to claim 2, the combination of Wang (‘402) in view of Wang (‘182), in further view of Luo and Nehmadi, and in further view of Fulton and Mizota teaches:
the prediction system according to claim 1, wherein: the target product is a casting (See Wang (‘402), [0007] (discussing aluminum castings).);
However, the combination does not teach “the defect of the target product, indicated by the defect characteristic value, includes at least one of seizure, shrinkage …,” which Nieves Acedo does teach:
and the defect of the target product, indicated by the defect characteristic value, includes at least one of seizure, shrinkage, flow line, galling, raw material deformation, die cracking, or entrapment of the target product (Nieves Acedo discloses “using these variables, measured in real time during the process, an attempt is made to predict the following characteristics:
- appearance of microporosity or microshrinkage. Microshrinkage, also known as secondary contraction, is a defect or irregularity in castings that tends to appear in the cooling stage. Specifically, this defect consists of a form of shrinkage involving a large number of very small cavities which they may be distributed across a large area of the casting” [0011].
See also Col. 6, lines 25-51 (discussing the classifiers used to predict output variables (microshrinkage).).
Nieves Acedo teaches predicting microshrinkage in castings using classifiers is a known method in the art. Before the effective filing date, it would have been obvious to a person of ordinary skill in the art to combine the artificial neural network disclosed by Wang (‘402) for predicting residual stress and distortion in aluminum castings with the technique disclosed by Nieves Acedo to predict microshrinkage defects in castings. By predicting microshrinkages, issues in the cooling process can be identified and resolved earlier. Identifying cooling issues early on would prevent castings from forming the microshrinkages that weaken them, thereby reducing manufacturing costs.
With respect to claim 9, the claim recites similar limitations corresponding to claim 2, therefore the same rationale of rejection is applicable.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PEDRO J MORALES whose telephone number is (571)272-6106. The examiner can normally be reached 8:30 AM - 6:00 PM.
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, MIRANDA M HUANG can be reached at (571)270-7092. 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.
/PEDRO J MORALES/Examiner, Art Unit 2124
/VINCENT GONZALES/Primary Examiner, Art Unit 2124