DETAILED ACTION
Claims 1-24 are presented for examination. This office action is response to the submission on 11/15/2023.
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 statement (IDS) submitted on 11/15/2023 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Drawings
The drawings filed on 11/15/2023 are acceptable for examination proceedings.
Claim Objections
Claim 22 is objected to because of the following informalities:
Line 3 reads “…are contained in a file obtained by means CAM software…” Examiner believes this should read “…are contained in a file obtained by means of CAM software…” (Typo).
Appropriate correction is required.
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.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-8, 14-18 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Matusik et al. (US20200143006A1).
Claim 1:
Matusik teaches “A method for manufacturing three-dimensional articles by deposition of a plurality of overlapping or adjoining layers of a material for additive manufacturing, comprising the following steps: (i) manufacturing test samples having forms, dimensions and geometric characteristics different from each other by means of deposition of a plurality of overlapping or adjoining layers of material;” (Matusik teaches fabricating multiple parts which have modified input specificiations i.e. they are different from each other in Matusik [0102] "In a more detailed description of the printing process, a series of one or more desired input specifications (x1 . . . xn) 105 of some or all of an object to be fabricated is provided to the machine learning based predistorter G 215 that is configured according to parameters ΘG. The predistorter G 215 processes the input specifications 105 to generate a corresponding series of one or more modified input specifications (y1 . . . yn) 120. The series of modified input specifications 120 is provided to a printer H 125 which prints a series of one or more fabricated parts ({tilde over (Z)}1 . . . {tilde over (Z)}N) according to the series of one or more modified input specifications 120. The fabricated parts 175 are scanned using a scanner 150 (e.g., an optical scanner) producing the successive scans Z1 . . . ZN." and in Matusik [0104] "In some examples, a series of scans Z1 . . . Z N 155 of the fabricated parts 175 is fed back to the beginning of the printing process. At the beginning of the printing process the scans 155 may be combined with object specifications X1 . . . Xiv to form adjusted input specifications 105 that represent the increments (x1 . . . xN) to be printed. It is those increments that are provided to the predistorter G 215."; Matusik teaches that the invention relates to additive manufacturing i.e. deposition of a plurality of overlapping layers of material in Matusik [0002-0003] "This invention relates to an intelligent additive manufacturing approach, and more particularly an approach that makes use of one or more of machine learning, feedback using machine vision, and determination of machine state. Additive manufacturing (AM) is a set of methods that allows objects to be fabricated via selective addition of material. A typical additive manufacturing process works by slicing a digital model (for example, represented using an STL file) into a series of layers. Then the layers are sent to a fabrication apparatus that deposits the layers one by one from the bottom to the top."),
“(ii) detecting and collecting, at least during said step (i) of manufacturing the test samples, data regarding process parameters relating to one or more of the deposition of the material and data regarding one or more of geometric and dimensional and qualitative and structural characteristics of the deposited layers of material;” (Matusik teaches measuring characteristics of the printer which would affect the characteristics of the material in Matusik [0093] "In some implementations, the characteristics of the printer H depend on various environmental or state variables. Examples may include ambient temperature, individual jet characteristics (e.g., degree of clogging) etc. In some implementations, changes in such variables may be captured by updating the parameters ΘĤ and ΘG. In other implementations they are explicitly measured or estimated, and may be represented as mn or {circumflex over (m)}n, and provided as a further input to Ĥ and G."; Matusik teaches scanning the object as it is being manufactured and adjusting parameters based on the measurements in Matusik [0105-0106] "In FIG. 3, the use of both a machine learning based predistorter G 215 and a machine learning based simulator Ĥ 216 may improve the fidelity of manufactured objects. In some examples, to train the predistorter G 215 and the simulator Ĥ 216, training data is collected from the printing of previous objects and the collected training data is used to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216."),
“(iii) analysis and processing of the data detected during step (ii) in order to derive and obtain optimized reference values of the process parameters;” (Matusik teaches that training data including scans are processed in a training procedure to determine updated predistorter parameters in Matusik [0106-0107] "For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. Referring to FIG. 4, in one example of such a training procedure, a combined loss 430 is determined as a sum (or weighted sum) of a simulation loss, an inverse design loss, a cycle input 3D model loss, and a cycle manufactured 3D model loss. That combined loss is provided to parameter optimization module 431 that determines the updated predistorter parameters ΘG and the updated simulation parameters ΘĤ."),
“(iv) manufacturing the three-dimensional article by means of deposition of a plurality of overlapping or adjoining layers of material for additive manufacturing based on the optimized reference values of the process parameters;” (Matusik teaches that the additive manufacturing process accounts for errors from printing pass to printing pass i.e. a three-dimensional article is manufactured using the optimized parameters in Matusik [0117] "The additive manufacturing process described above improves manufacturing accuracy by accounting for systematic errors that are consistent from printing pass to printing pass. These errors are determined by the physical behavior of print materials and the manufacturing process."), and
“wherein said data analysis and processing step (iii) is performed by means of a process for training a software based on at least one artificial intelligence algorithm.” (Matusik teaches that the model parameters may be updated using an artificical neural network in Matusik [0106-108] "For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. Referring to FIG. 4, in one example of such a training procedure, a combined loss 430 is determined as a sum (or weighted sum) of a simulation loss, an inverse design loss, a cycle input 3D model loss, and a cycle manufactured 3D model loss. That combined loss is provided to parameter optimization module 431 that determines the updated predistorter parameters ΘG and the updated simulation parameters ΘĤ. In some examples, the parameter optimization module 431 determines the parameters ΘG and ΘĤ by using an optimization algorithm to iteratively update the parameters ΘG and ΘĤ until some stopping criteria (e.g., number of iterations, convergence). One common way of updating model parameters involves computing/estimating gradients of the combined loss function with respect to the model parameters. For example, a backpropagation algorithm for an artificial neural networks uses such gradient estimates. Of course, other parameter update schemes can be used.").
Claim 2:
Matusik teaches “The method according to claim 1, characterized in that said detection and collection step ii) is performed by means of automatic detection means.” (Matusik teaches that the fabricated parts are scanned N number of times i.e. the scanner automatically scans the part in Matusik [0102] "In a more detailed description of the printing process, a series of one or more desired input specifications (x1 . . . xn) 105 of some or all of an object to be fabricated is provided to the machine learning based predistorter G 215 that is configured according to parameters ΘG. The predistorter G 215 processes the input specifications 105 to generate a corresponding series of one or more modified input specifications (y1 . . . yn) 120. The series of modified input specifications 120 is provided to a printer H 125 which prints a series of one or more fabricated parts ({tilde over (Z)}1 . . . {tilde over (Z)}N) according to the series of one or more modified input specifications 120. The fabricated parts 175 are scanned using a scanner 150 (e.g., an optical scanner) producing the successive scans Z1 . . . ZN.").
Claim 3:
Matusik teaches “The method according to claim 2, characterized in that said automatic detection means comprise one or more of at least one or more of at least (Matusik teaches a scanner which determines scan data representing the surface of the object in Matusik [0082] "In other embodiments, as illustrated n FIG. 1 (with dotted lines representing optional paths associated with feedback), a scanner 150 is schematically illustrated in a configuration that is used to determine scan data 155 representing the result of the fabrication of the object, with this data representing the geometry and/or material of the surface or near-surface section (e.g., material composition or density as a function of depth). In an example of such a feedback arrangement, the scanner 150 determines scan data 155 representing the surface 171 of a previously fabricated part 170 of the object.").
Claim 4:
Matusik teaches “The method according to claim 1, further comprising a step for analysis of the three-dimensional article to be manufactured and decomposition of the geometry of the three-dimensional article into one or more portions performed upstream of step (iv).” (Matusik teaches separating the object into a set of slices i.e. portions with each slice being scanned as the object is produced in Matusik [0087-0088] "In the following further discussion, the following terminology, which is consistent with the terminology used above, is generally used. An object with specification X for the entire completed object is fabricated as a sequence of partial objects with specifications X1, X2, . . . such that an increment (“slice” or set of slices) xn between two successive partial objects is ideally deposited essentially as an increment to form a combination (represented by an addition symbol) Xn=. Each slice xn is used to control a printer to deposit material, thereby forming the sequence of partial objects, with scans Z1, Z2, . . . and corresponding inferred fabricated slices (i.e., the achieved increments) zn=Zn−Zn-1. In the ideal case, the printer exactly produces zn=xn for each slice. The physical printer can be represented in a first form as a transformation H of a commanded slice xn (which may be printed in one or more layers by the printer—that is, a slice is not in general limited to a single printing layer) yielding a fabricated slice with an incremental scan zn, represented in a functional form as zn=H(xn). In some contexts the physical printer is viewed in a second form as a transformation that inputs the previous partially fabricated object, which has a scan Zn-1, as well as the next slice xn and yields the scan of the next partial object, Zn.").
Claim 5:
Matusik teaches “The method according to claim 4, characterized in that the test samples have forms, dimensions and geometric characteristics corresponding to the forms, dimensions and geometric characteristics of the portions of the article to be to be manufactured.” (Matusik teaches separating the object into a set of slices i.e. each slice or layer of the object is unique and has its own form, dimensions, and geometric characteristics in Matusik [0087-0088] "In the following further discussion, the following terminology, which is consistent with the terminology used above, is generally used. An object with specification X for the entire completed object is fabricated as a sequence of partial objects with specifications X1, X2, . . . such that an increment (“slice” or set of slices) xn between two successive partial objects is ideally deposited essentially as an increment to form a combination (represented by an addition symbol) Xn=. Each slice xn is used to control a printer to deposit material, thereby forming the sequence of partial objects, with scans Z1, Z2, . . . and corresponding inferred fabricated slices (i.e., the achieved increments) zn=Zn−Zn-1. In the ideal case, the printer exactly produces zn=xn for each slice. The physical printer can be represented in a first form as a transformation H of a commanded slice xn (which may be printed in one or more layers by the printer—that is, a slice is not in general limited to a single printing layer) yielding a fabricated slice with an incremental scan zn, represented in a functional form as zn=H(xn). In some contexts the physical printer is viewed in a second form as a transformation that inputs the previous partially fabricated object, which has a scan Zn-1, as well as the next slice xn and yields the scan of the next partial object, Zn.").
Claim 6:
Matusik teaches “The method according to claim 4, characterized in that the step for analysis and decomposition of the three-dimensional article is performed using 3D CAD software or solid modelling” (Matusik teaches that the specifications of objects can be represented as a solid model in Matusik [0193] "In some examples, the specifications of objects (or parts of objects) can be represented as any one of a mesh: e.g., a list of vertices and triangles; a triangle soup, a boundary representation (brep), a solid model: a voxel grid (uniform and adaptive), a tetrahedral model, a solid model with material assignment, a procedural model, and implicit model."),
“so as to obtain instructions regarding the operations to be performed by means of CAM software for manufacture of the three-dimensional article,” (Matusik teaches that the predistorter 215 provides a modified specification such as printer commands i.e. operations to be performed for manufacture of the three-dimensional article in Matusik [0081] "Continuing to refer to FIG. 1, the printing system 100 is characterized by a parameterized predistorer 115, which takes as input the specification 105 and produces a modified specification 120 (or other suitable input to the printer 125, such as printer commands or some sort of printing plan). The terms “predistorter” or “distort” do not connote particular types of operations, and should not be viewed as limiting (i.e., operations other than geometric modification are encompassed). In some or all embodiments, the predistorter is configured to essentially form an inverse (a “pre-inverse”) of a printing process, such that when the modified specification 120 is provided to a printer 125 and fabricated to form a corresponding fabricated part 175 of the object, that part 175 matches the originally desired specification 105 that is provided as input."), and
“the CAM and CAD software being interfaced with the software based on at least one artificial intelligence algorithm so as to be trained in such a way as to optimize the procedures for analysis and decomposition of the article and for obtaining instructions for the manufacture of the article.” (Matusik teaches that the model parameters may be updated using an artificical neural network in Matusik [0106-108] "For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. Referring to FIG. 4, in one example of such a training procedure, a combined loss 430 is determined as a sum (or weighted sum) of a simulation loss, an inverse design loss, a cycle input 3D model loss, and a cycle manufactured 3D model loss. That combined loss is provided to parameter optimization module 431 that determines the updated predistorter parameters ΘG and the updated simulation parameters ΘĤ. In some examples, the parameter optimization module 431 determines the parameters ΘG and ΘĤ by using an optimization algorithm to iteratively update the parameters ΘG and ΘĤ until some stopping criteria (e.g., number of iterations, convergence). One common way of updating model parameters involves computing/estimating gradients of the combined loss function with respect to the model parameters. For example, a backpropagation algorithm for an artificial neural networks uses such gradient estimates. Of course, other parameter update schemes can be used."; Matusik Fig. 3 teaches that the updated parameter ΘG is used to determine the predistorter settings which affects the input specifications 120 supplied to the printer
[AltContent: rect]
PNG
media_image1.png
800
269
media_image1.png
Greyscale
).
Claim 7:
Matusik teaches “The method according to claim 1, characterized in that the material for additive manufacturing is chosen from the group comprising thermoplastic materials, (Matusik teaches that the aspects described are applicable to fused deposition modeling, where thermoplastic material is used in Matusik [0194] "Aspects described herein are applicable to any number of additive manufacturing processes, including but not limited to inkjet-based AM, powder-bed, Fused deposition modeling (FDM), Stereolithography (SLA), SLS, DMLS, and Electron-beam melting.").
Claim 8:
Matusik teaches “The method according to claim 5, characterized in that the geometric characteristics of said test samples and of said portions of the three-dimensional article comprise the radii of curvature,” (Matusik teaches that the specifications of objects can be represented as a solid model, which would include the radii of curvature of the object in Matusik [0193] "In some examples, the specifications of objects (or parts of objects) can be represented as any one of a mesh: e.g., a list of vertices and triangles; a triangle soup, a boundary representation (brep), a solid model: a voxel grid (uniform and adaptive), a tetrahedral model, a solid model with material assignment, a procedural model, and implicit model."),
“the widths and the thicknesses of the deposited layers of material,” (Matusik teaches that inputs may include thickness of a location in Matusik [0094] "As discussed in more detail below, the transformations G and Ĥ may be parameterized with the parameters adjusted according to the design criteria, which may be referred to as optimization or loss functions. For example, these transformations may be implemented as artificial neural networks, such as deep convolutional neural networks. In the representation above, an input xn may take various representations, including but not limited to samples of thickness over a grid of two-dimensional locations, or a vector of material composition over the grid, with each position in the vector being associated with an amount or concentration of a different material. In some implementations, an input xn or Xn is represented over a three-dimensional array of locations, for example, representing individual voxels. In some implementations, the output of the predistorter G has the same characteristics as its input. For example, it may form a mapping from desired thickness as a function of location to a commanded thickness as a function of location. However, in other implementations, an input yn to the printer H or its simulations Ĥ may be in the form of printer commands, such as print-head commands, velocity, etc., and the “predistorter” effectively implements a transformation from a desired structure to printer commands."),
“the overlapping of different or adjoining portions,” (Matusik teaches separating the object into a set of slices which would define how the different layers are overlapping in Matusik [0087-0088] "In the following further discussion, the following terminology, which is consistent with the terminology used above, is generally used. An object with specification X for the entire completed object is fabricated as a sequence of partial objects with specifications X1, X2, . . . such that an increment (“slice” or set of slices) xn between two successive partial objects is ideally deposited essentially as an increment to form a combination (represented by an addition symbol) Xn=. Each slice xn is used to control a printer to deposit material, thereby forming the sequence of partial objects, with scans Z1, Z2, . . . and corresponding inferred fabricated slices (i.e., the achieved increments) zn=Zn−Zn-1. In the ideal case, the printer exactly produces zn=xn for each slice. The physical printer can be represented in a first form as a transformation H of a commanded slice xn (which may be printed in one or more layers by the printer—that is, a slice is not in general limited to a single printing layer) yielding a fabricated slice with an incremental scan zn, represented in a functional form as zn=H(xn). In some contexts the physical printer is viewed in a second form as a transformation that inputs the previous partially fabricated object, which has a scan Zn-1, as well as the next slice xn and yields the scan of the next partial object, Zn."), and
“and the interspaces between adjoining portions.” (Matusik teaches that the specification 105 may include the density of a slab i.e. the density would determine the interspaces between adjoining portions in Matusik [0080] "In general, the specification 105 specifies a desired 3D structure (e.g., a solid model), and may be represented, for instance without limitation: as a surface (e.g., in micropolygon form, such as an STL file, or analytically as a functional form of the surface); as a set of quantized volume regions (e.g., “voxels”) indicating which regions are to be filled, the material(s) to use, the density; as a thickness (e.g., in a z dimension) of a thin planar slab as a function of location on the plane (e.g., the x-y plane) of the slab; or as a desired resulting top surface after adding the part to the object. In some implementations, the specification 105 is a product of a planning (e.g., a “slicing”) phase (not shown in FIG. 1) that processes a specification of an entire object and forms specifications of layers to be fabricated in turn to form the entire object.").
Claim 14:
Matusik teaches “The method according to claim 3, characterized in that the one or more geometric and dimensional and qualitative and structural characteristics of the layers of material for additive manufacturing detected in said step (ii) comprise the presence of (Matusik teaches checking for defects in scans of partial 3d prints and incorporates the machine state to plan the next layers in Matusik [0141-0142] "Referring to FIGS. 22 and 23, the printing process described above for FIGS. 21 and 22 is augmented a feedback loop that provides scans 155 to the processing program 2215. The machine state is estimated continuously with the feedback loop control. The processing program 2215 checks for any defects in the scans 155 of partial 3D prints and incorporates the machine state to plan the next layers. For example, if an area of the partial 3D print is lower than the expected height, the processing program can assign over-jetting nozzles to the area for the next few layers. Referring to FIG. 24, in some examples, the printing process includes the predistorter G 2415 (e.g., predistorter of FIG. 2) the and a processing program 2416 configured to determine a fabrication plan 2421 according to an estimated machine state 2417."; [AltContent: rect]
PNG
media_image2.png
809
259
media_image2.png
Greyscale
).
Claim 15:
Matusik teaches “The method according to claim 1, characterized in that said data detection and collection step (ii) is performed during said step (iv) for manufacturing the three-dimensional article,” (Matusik teaches scanning the object as it is being manufactured and adjusting parameters based on the measurements in Matusik [0105-0106] "In FIG. 3, the use of both a machine learning based predistorter G 215 and a machine learning based simulator Ĥ 216 may improve the fidelity of manufactured objects. In some examples, to train the predistorter G 215 and the simulator Ĥ 216, training data is collected from the printing of previous objects and the collected training data is used to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216."), and
“the three-dimensional article manufactured in said step (iv) being a test sample for the subsequent manufacture of three-dimensional articles.” (Matusik teaches that training data collected from printing of previous objects is used to determine parameters in Matusik [0105] "In FIG. 3, the use of both a machine learning based predistorter G 215 and a machine learning based simulator Ĥ 216 may improve the fidelity of manufactured objects. In some examples, to train the predistorter G 215 and the simulator Ĥ 216, training data is collected from the printing of previous objects and the collected training data is used to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216.").
Claim 16:
Matusik teaches “The method according to claim 1, characterized in that said steps (i)-(iii) are repeated a predefined number of times so as to store said data and train said at least one artificial intelligence algorithm to process and improve the optimized reference values of the process parameters.” (Matusik teaches that the parameter optimization module may determine the parameters using an optimization algorithm until a number of iterations is reached and that the updating may be done using an artificial neural network in Matusik [0106-108] "For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. Referring to FIG. 4, in one example of such a training procedure, a combined loss 430 is determined as a sum (or weighted sum) of a simulation loss, an inverse design loss, a cycle input 3D model loss, and a cycle manufactured 3D model loss. That combined loss is provided to parameter optimization module 431 that determines the updated predistorter parameters ΘG and the updated simulation parameters ΘĤ. In some examples, the parameter optimization module 431 determines the parameters ΘG and ΘĤ by using an optimization algorithm to iteratively update the parameters ΘG and ΘĤ until some stopping criteria (e.g., number of iterations, convergence). One common way of updating model parameters involves computing/estimating gradients of the combined loss function with respect to the model parameters. For example, a backpropagation algorithm for an artificial neural networks uses such gradient estimates. Of course, other parameter update schemes can be used.").
Claim 17:
Matusik teaches “The method according to claim 1, characterized in that the at least one algorithm of the artificial intelligence software is of the machine learning, deep learning and reinforcement learning type or a combination of these three types.” (Matusik teaches that the parameter optimization module may be done using an artificial neural network in Matusik [0106-108] "For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. Referring to FIG. 4, in one example of such a training procedure, a combined loss 430 is determined as a sum (or weighted sum) of a simulation loss, an inverse design loss, a cycle input 3D model loss, and a cycle manufactured 3D model loss. That combined loss is provided to parameter optimization module 431 that determines the updated predistorter parameters ΘG and the updated simulation parameters ΘĤ. In some examples, the parameter optimization module 431 determines the parameters ΘG and ΘĤ by using an optimization algorithm to iteratively update the parameters ΘG and ΘĤ until some stopping criteria (e.g., number of iterations, convergence). One common way of updating model parameters involves computing/estimating gradients of the combined loss function with respect to the model parameters. For example, a backpropagation algorithm for an artificial neural networks uses such gradient estimates. Of course, other parameter update schemes can be used.").
Claim 18:
Matusik teaches “The method according to claim 1, characterized in that the data detected and collected in said step (ii) forms an input for the process of training the at least one artificial intelligence algorithm” (Matusik teaches that scan data of the part of the object that is already fabricated may be input into the predistorter 115 in Matusik [0085] "As introduced above, in some “feedback” arrangements, the predistorter 115 can have two inputs: the specification 105 of the part to be fabricated, and scan data 155 of the part of the object that is already fabricated. For such arrangements, the learner 210 determines the parameters 225 to process these two inputs to yield the modified specification 120. To do so, the training data includes triples, each including: scan data 155 for the part of the object already fabricated; the specification 220 provided to the printer 125; and the resulting scan data 256 after the part corresponding to specification 220 has been fabricated."), and
“and the optimized reference values of the process parameters form an output of the process for training of the at least one artificial intelligence algorithm.” (Matusik teaches that the output of the machine learning model can be a fabrication plan in Matusik [0208] "The output of the machine learning models described above can be a 3D model (representation as described before) or part of the model. Typically, only a subsection of the model is used to simplify training process. For example, the data can be a set of voxel layers, where each layer can be represented as an image data. Alternatively, the output can be a fabrication plan."; Matusik teaches that a fabrication plan includes instructions on how to deposit material and that the sensing data can be used to adapt the fabrication plan in Matusik [0197-0198] "The processing program can be specified as follows. The input includes a 3D model. The output includes a fabrication plan that includes instructions to the manufacturing hardware on how/where to move, how to deposit material; how to process material (e.g., curing unit on, off, power; laser power, etc.); auxiliary devices (e.g., rollers, cleaning stations, thermal management). Tasks performed in the processing program include determining part orientation, generating support material, computing digital layers (e.g., slices), computing infill pattern, computing contour, and generating machine instructions. Sensing data (e.g., a 3D scan and material IDs for each voxel) can be used as additional input to the processing program. For example, the processing program might use sensor data to adapt the fabrication plan. Machine state can be used to adjust: the amount of material that is deposited from each deposition device; the motion offset for deposition devices, e.g., adjusting the offset in printhead (typically y-axis) to obtain flat print, high quality print, faster print.").
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 9-11 and 19-23 are rejected under 35 U.S.C. 103 as being unpatentable over Matusik et al. (US20200143006A1, in view of Susnjara et al. (US20190322044A1).
Claim 9:
Matusik teaches “The method according to claim 7, characterized in that said step (iv) for manufacturing the three-dimensional article by means of deposition of the layers of thermoplastic material is performed by means of at least one extruder device for extruding the thermoplastic material” (Matusik teaches an extruder for deposition in Matusik [0199] "In some examples, the fabrication plans described above refer to the sequence of operations required to be executed by the hardware in order to manufacture an object or a set of objects. It should include how the movements of the positioning system, the deposition timing and amount of deposition from the deposition device(s) (e.g., extruders, dispensers, nozzles)."), and
“and a nozzle.” (Matusik teaches a nozzle in Matusik [0199] "In some examples, the fabrication plans described above refer to the sequence of operations required to be executed by the hardware in order to manufacture an object or a set of objects. It should include how the movements of the positioning system, the deposition timing and amount of deposition from the deposition device(s) (e.g., extruders, dispensers, nozzles).").
Matusik does not appear to explicitly teach “comprising at least one screw extruder element,” or “a pump” However, Susnjara does teach these claim limitations.
Susnjara teaches “comprising at least one screw extruder element,” (Susnjara teaches an extruder comprising a heavy duty screw in Susnjara [0050] "With reference now to FIG. 4, there is illustrated, a cross-sectional schematic representation of a thermoplastic extrusion and application system, along with a block diagram of an exemplary servo control circuit, according to aspects of the present disclosure. FIG. 4 depicts an extruder 60, comprising a heavy duty screw 63, rotatably mounted inside a barrel 64, and driven by a servomotor 61 through a gearbox 62."
[AltContent: rect]
PNG
media_image3.png
647
533
media_image3.png
Greyscale
), and
“a pump” (Susnjara teaches a pump in Susnjara [0045] "As shown in FIG. 2A, machine 1A may include a carrier 25A provided with a positive displacement gear pump 66, driven by a servomotor 67 through a gearbox 68. Gear pump 66 may receive molten plastic from extruder 60, as shown in FIG. 1A. Material may be pushed out of gear pump 66 to an applicator head 43A. "
[AltContent: rect]
PNG
media_image4.png
655
556
media_image4.png
Greyscale
).
Matusik and Susnjara are analogous art because they are from the same field of endeavor of additive manufacturing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having teachings of Matusik and Susnjara before him/her, to modify the teachings of an Intelligent additive manufacturing method of Matusik to include the pump and extruder including a heavy duty screw of Susnjara because adding the Methods and apparatus for processing and dispensing material during additive manufacturing of Susnjara would allow for maintaining an optimal pressure within the extruder without a breaker plate or screen as described in Susnjara [0058-0059] “The ability to generate the required pressure may be accomplished with a lower-cost system that reduces mechanical complexity without the need for a breaker plate or a screen (such as a filter) between an end of the screw 63 and gear pump 66, as shown in FIG. 4A, for example. Control of gear pump 66 may be performed without unduly restricting throughput, resulting in higher flow rates for extruder 60. In an exemplary configuration, nozzle 51 may have an open round shape (FIG. 6) which offers little resistance to material flow. Gear pump 66 may restrict flow to the nozzle 51, thereby avoiding the need to provide a nozzle having significant resistance to material flow. A desired or optimal pressure within extruder 60 may be created and maintained by controlling the relative speeds of the extruder 60 and gear pump 66.”
Claim 10:
Matusik in view of Susnjara teaches “The method according to claim 9, characterized in that said at least one extruder device is mounted on a machine with Cartesian or (Susnjara teaches a CNC machine for displacing the application nozzle along the x, y, or z axis i.e. Cartesian movements in Susnjara [0041] "With reference now to FIG. 1 of the drawings, there is illustrated a programmable computer numeric control (CNC) machine 1 embodying aspects of the present disclosure. A controller (not shown) may be operatively connected to machine 1 for displacing an application nozzle along a longitudinal line of travel or x-axis, a transverse line of travel or a y-axis, and a vertical line of travel or z-axis, in accordance with a program inputted or loaded into the controller for performing an additive manufacturing process to replicate a desired component. CNC machine 1 may be configured to print or otherwise build 3D parts from digital representations of the 3D parts (e.g., AMF and STL format files) programmed into the controller. For example, in an extrusion-based additive manufacturing system, a 3D part may be printed from a digital representation of the 3D part in a layer-by-layer manner by extruding a flowable material. The flowable material may be extruded through an extrusion tip carried by a print head of the system, and is deposited as a sequence of beads or layers on a substrate in an x-y plane. The extruded flowable material may fuse to previously deposited material, and may solidify upon a drop in temperature. The position of the print head relative to the substrate is then incrementally advanced along a z-axis (perpendicular to the x-y plane), and the process is then repeated to form a 3D part resembling the digital representation."; Susnjara teaches that the carriage 24 is displaceable along guide rails by a servomotor in Susnjara [0042] "Machine 1 includes a bed 20 provided with a pair of transversely spaced side walls 21 and 22, a gantry 23 supported on side walls 21 and 22, carriage 24 mounted on gantry 23, a carrier 25 mounted on carriage 24, an extruder 60, and an applicator assembly 26 mounted on carrier 25. Supported on bed 20 between side walls 21 and 22 is a worktable 27 provided with a support surface disposed in an x-y plane, which may be fixed or displaceable along an x-axis. In the displaceable version, the worktable 27 may be displaceable along a set of rails mounted on the bed 20 by means of servomotors and rails 28 and 29 mounted on the bed 20 and operatively connected to the worktable 27. Gantry 23 is disposed along a y-axis, supported at the ends thereof on end walls 21 and 22, either fixedly or displaceably along an x-axis on a set of guide rails 28 and 29 provided on the upper ends of side walls 21 and 22. In the displaceable version, the gantry 23 may be displaceable by a set of servomotors mounted on the gantry 23 and operatively connected to tracks provided on the side walls 21 and 22 of the bed 20. Carriage 24 is supported on gantry 23 and is provided with a support member 30 mounted on and displaceable along one or more guide rails 31, 32 and 33 provided on the gantry 23. Carriage 24 may be displaceable along a y-axis on one or more guide rails 31, 32 and 33 by a servomotor mounted on the gantry 23 and operatively connected to support member 30. Carrier 25 is mounted on a set of spaced, vertically disposed guide rails 34 and 35 supported on the carriage 24 for displacement of the carrier 25 relative to carriage 24 along a z-axis. Carrier 25 may be displaceable along the z-axis by a servomotor mounted on carriage 24 and operatively connected to carrier 25."), and
“said dedicated software being adjusted and implemented on the basis of said optimized reference values of the process parameters.” (Matusik teaches scanning the object as it is being manufactured and adjusting parameters based on the measurements in Matusik [0105-0106] "In FIG. 3, the use of both a machine learning based predistorter G 215 and a machine learning based simulator Ĥ 216 may improve the fidelity of manufactured objects. In some examples, to train the predistorter G 215 and the simulator Ĥ 216, training data is collected from the printing of previous objects and the collected training data is used to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216.").
Claim 11:
Matusik in view of Susnjara teaches “The method according to claim 10, characterized in that the software based on the at least one artificial intelligence algorithm, trained on the basis of the data detected during step (ii), is configured to implement the dedicated software of the numerical control machine which consequently adjusts operation of the extruder device, of the movement means and of the automatic detection device during the manufacture of the three-dimensional article.” (Matusik teaches measuring characteristics of the printer including temperature and jet characteristics, and that the changes may be captured by updating the parameters ΘĤ and ΘG in Matusik [0093] "In some implementations, the characteristics of the printer H depend on various environmental or state variables. Examples may include ambient temperature, individual jet characteristics (e.g., degree of clogging) etc. In some implementations, changes in such variables may be captured by updating the parameters ΘĤ and ΘG. In other implementations they are explicitly measured or estimated, and may be represented as mn or {circumflex over (m)}n, and provided as a further input to Ĥ and G."; Matusik teaches scanning the object as it is being manufactured and adjusting parameters based on the measurements in Matusik [0105-0106] "In FIG. 3, the use of both a machine learning based predistorter G 215 and a machine learning based simulator Ĥ 216 may improve the fidelity of manufactured objects. In some examples, to train the predistorter G 215 and the simulator Ĥ 216, training data is collected from the printing of previous objects and the collected training data is used to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216."; Matusik teaches that fabrications plans include the movements of the positioning system in Matusik [0199] "In some examples, the fabrication plans described above refer to the sequence of operations required to be executed by the hardware in order to manufacture an object or a set of objects. It should include how the movements of the positioning system, the deposition timing and amount of deposition from the deposition device(s) (e.g., extruders, dispensers, nozzles). Other information might include: curing/sintering sequence and settings (e.g., UV light, IR, heat); movement and settings of supporting devices (e.g., rollers, scrapers), temperature/pressure settings, etc.").
Claim 19:
Matusik teaches “A plant for manufacturing three-dimensional articles by deposition of a plurality of overlapping or adjoining layers of a material for additive manufacturing, said plant comprising: a processing unit having an installed software;” (Matusik teaches that the approaches described may be implemented using a processor with software in Matusik [0218] "More generally, the approaches described above can be implemented, for example, using a programmable computing system executing suitable software instructions or it can be implemented in suitable hardware such as a field-programmable gate array (FPGA) or in some hybrid form. For example, in a programmed approach the software may include procedures in one or more computer programs that execute on one or more programmed or programmable computing system (which may be of various architectures such as distributed, client/server, or grid) each including at least one processor, at least one data storage system (including volatile and/or non-volatile memory and/or storage elements), at least one user interface (for receiving input using at least one input device or port, and for providing output using at least one output device or port)."),
“a machine for deposition of the layers of material for additive manufacturing comprising: at least one extruder device for the deposition of the layers of material;” (Matusik teaches an extruder for deposition in Matusik [0199] "In some examples, the fabrication plans described above refer to the sequence of operations required to be executed by the hardware in order to manufacture an object or a set of objects. It should include how the movements of the positioning system, the deposition timing and amount of deposition from the deposition device(s) (e.g., extruders, dispensers, nozzles)."),
“means for the automatic detection of data regarding the process parameters relating to one or more of the deposition of the layers of material and data regarding the geometric and/or dimensional and qualitative and structural characteristics of the deposited layers of material for additive manufacturing;” (Matusik teaches measuring characteristics of the printer which would affect the characteristics of the material in Matusik [0093] "In some implementations, the characteristics of the printer H depend on various environmental or state variables. Examples may include ambient temperature, individual jet characteristics (e.g., degree of clogging) etc. In some implementations, changes in such variables may be captured by updating the parameters ΘĤ and ΘG. In other implementations they are explicitly measured or estimated, and may be represented as mn or {circumflex over (m)}n, and provided as a further input to Ĥ and G."; Matusik teaches scanning the object as it is being manufactured and adjusting parameters based on the measurements in Matusik [0105-0106] "In FIG. 3, the use of both a machine learning based predistorter G 215 and a machine learning based simulator Ĥ 216 may improve the fidelity of manufactured objects. In some examples, to train the predistorter G 215 and the simulator Ĥ 216, training data is collected from the printing of previous objects and the collected training data is used to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216."),
“a computerized numerical control system having dedicated software for controlling said at least one extruder device and said movement means;” (Matusik teaches the fabrication plans include the movements of a positioning system and amount of deposition from an extruder in Matusik [0199] "In some examples, the fabrication plans described above refer to the sequence of operations required to be executed by the hardware in order to manufacture an object or a set of objects. It should include how the movements of the positioning system, the deposition timing and amount of deposition from the deposition device(s) (e.g., extruders, dispensers, nozzles)."),
“said processing unit being configured to receive the data from said automatic detection means, to process said data and to obtain optimized reference values of the process parameters from said data;” (Matusik teaches that training data including scans are processed in a training procedure to determine updated predistorter parameters in Matusik [0106-0107] "For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. Referring to FIG. 4, in one example of such a training procedure, a combined loss 430 is determined as a sum (or weighted sum) of a simulation loss, an inverse design loss, a cycle input 3D model loss, and a cycle manufactured 3D model loss. That combined loss is provided to parameter optimization module 431 that determines the updated predistorter parameters ΘG and the updated simulation parameters ΘĤ."), and
“wherein the software of said processing unit is based on at least one artificial intelligence algorithm and is configured to adjust and implement the software of the computerized numerical control system.” (Matusik teaches that the model parameters may be updated using an artificical neural network in Matusik [0106-108] "For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. Referring to FIG. 4, in one example of such a training procedure, a combined loss 430 is determined as a sum (or weighted sum) of a simulation loss, an inverse design loss, a cycle input 3D model loss, and a cycle manufactured 3D model loss. That combined loss is provided to parameter optimization module 431 that determines the updated predistorter parameters ΘG and the updated simulation parameters ΘĤ. In some examples, the parameter optimization module 431 determines the parameters ΘG and ΘĤ by using an optimization algorithm to iteratively update the parameters ΘG and ΘĤ until some stopping criteria (e.g., number of iterations, convergence). One common way of updating model parameters involves computing/estimating gradients of the combined loss function with respect to the model parameters. For example, a backpropagation algorithm for an artificial neural networks uses such gradient estimates. Of course, other parameter update schemes can be used."; Matusik teaches that the parameters provided to the predistorer generates a modified input specification for the printer in Matusik [0097] "In the approach shown in FIG. 2, an input specification 105 of the object to be fabricated is provided to a machine learning based predistorter G 215 that is configured according to parameters ΘG. The predistorter G 215 processes the input specification 105 to generate a corresponding modified input specification 120. The modified input specification 120 is provided to a printer H 125 which prints a fabricated part 175 according to the modified input specification 120.").
Matusik does not appear to explicitly teach “means for the movement of the at least one extruder device;” or “a table for supporting the three-dimensional articles;” However, Susnjara does teach these claim limitations.
Susnjara teaches “means for the movement of the at least one extruder device;” (Susnjara teaches that the carriage 24 which includes the extruder 60 is displaceable along guide rails by a servomotor in Susnjara [0042] "Machine 1 includes a bed 20 provided with a pair of transversely spaced side walls 21 and 22, a gantry 23 supported on side walls 21 and 22, carriage 24 mounted on gantry 23, a carrier 25 mounted on carriage 24, an extruder 60, and an applicator assembly 26 mounted on carrier 25. Supported on bed 20 between side walls 21 and 22 is a worktable 27 provided with a support surface disposed in an x-y plane, which may be fixed or displaceable along an x-axis. In the displaceable version, the worktable 27 may be displaceable along a set of rails mounted on the bed 20 by means of servomotors and rails 28 and 29 mounted on the bed 20 and operatively connected to the worktable 27. Gantry 23 is disposed along a y-axis, supported at the ends thereof on end walls 21 and 22, either fixedly or displaceably along an x-axis on a set of guide rails 28 and 29 provided on the upper ends of side walls 21 and 22. In the displaceable version, the gantry 23 may be displaceable by a set of servomotors mounted on the gantry 23 and operatively connected to tracks provided on the side walls 21 and 22 of the bed 20. Carriage 24 is supported on gantry 23 and is provided with a support member 30 mounted on and displaceable along one or more guide rails 31, 32 and 33 provided on the gantry 23. Carriage 24 may be displaceable along a y-axis on one or more guide rails 31, 32 and 33 by a servomotor mounted on the gantry 23 and operatively connected to support member 30. Carrier 25 is mounted on a set of spaced, vertically disposed guide rails 34 and 35 supported on the carriage 24 for displacement of the carrier 25 relative to carriage 24 along a z-axis. Carrier 25 may be displaceable along the z-axis by a servomotor mounted on carriage 24 and operatively connected to carrier 25."), and
“a table for supporting the three-dimensional articles;” (Susnjara teaches a worktable 27 in Susnjara [0042] "Machine 1 includes a bed 20 provided with a pair of transversely spaced side walls 21 and 22, a gantry 23 supported on side walls 21 and 22, carriage 24 mounted on gantry 23, a carrier 25 mounted on carriage 24, an extruder 60, and an applicator assembly 26 mounted on carrier 25. Supported on bed 20 between side walls 21 and 22 is a worktable 27 provided with a support surface disposed in an x-y plane, which may be fixed or displaceable along an x-axis. In the displaceable version, the worktable 27 may be displaceable along a set of rails mounted on the bed 20 by means of servomotors and rails 28 and 29 mounted on the bed 20 and operatively connected to the worktable 27.").
Matusik and Susnjara are analogous art because they are from the same field of endeavor of additive manufacturing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having teachings of Matusik and Susnjara before him/her, to modify the teachings of an Intelligent additive manufacturing method of Matusik to include the pump and extruder including a heavy duty screw of Susnjara because adding the Methods and apparatus for processing and dispensing material during additive manufacturing of Susnjara would allow for displacing the carrier in order to appropriately deliver material and provide the desired end product as described in Susnjara [0049] “In the course of fabricating a component, pursuant to the methods described herein, the control system of the machine 1, in executing the inputted program, may control the several servomotors described above to displace the gantry 23 along the x-axis, displace the carriage 24 along the y-axis, displace the carrier 25 along a z-axis, pivot lower applicator segment 42 about an axis disposed in an x-y plane and rotate bracket 47 about a z-axis thereof, in accordance with the inputted program, to appropriately deliver material 53 and provide the desired end product or a near duplicate thereof. The control system of machine 1A may control the several servomotors to display gantry 23, carriage 24, and carrier 25A in a similar manner to appropriate deliver material 53.”
Claim 20:
Matusik in view of Susnjara teaches “The plant according to claim 19, characterized in that the automatic detection means comprise one or more of (Matusik teaches a scanner which determines scan data representing the surface of the object in Matusik [0082] "In other embodiments, as illustrated n FIG. 1 (with dotted lines representing optional paths associated with feedback), a scanner 150 is schematically illustrated in a configuration that is used to determine scan data 155 representing the result of the fabrication of the object, with this data representing the geometry and/or material of the surface or near-surface section (e.g., material composition or density as a function of depth). In an example of such a feedback arrangement, the scanner 150 determines scan data 155 representing the surface 171 of a previously fabricated part 170 of the object.").
Claim 21:
Matusik in view of Susnjara teaches “The plant (1) according to claim 19, characterized in that said machine comprises a local control unit connected to said at least one extruder device and one or more of said automatic detection means and said movement means, said local control unit being connected to said processing unit.” (Matusik teaches a scanner which determines scan data representing the surface of the object in Matusik [0082] "In other embodiments, as illustrated n FIG. 1 (with dotted lines representing optional paths associated with feedback), a scanner 150 is schematically illustrated in a configuration that is used to determine scan data 155 representing the result of the fabrication of the object, with this data representing the geometry and/or material of the surface or near-surface section (e.g., material composition or density as a function of depth). In an example of such a feedback arrangement, the scanner 150 determines scan data 155 representing the surface 171 of a previously fabricated part 170 of the object."; Matusik teaches the fabrication plans include the movements of a positioning system and amount of deposition from an extruder in Matusik [0199] "In some examples, the fabrication plans described above refer to the sequence of operations required to be executed by the hardware in order to manufacture an object or a set of objects. It should include how the movements of the positioning system, the deposition timing and amount of deposition from the deposition device(s) (e.g., extruders, dispensers, nozzles)."; Matusik teaches that the approaches described may be implemented using a processor with software i.e. the processor may be a local control unit which is connected to the extruder, detection means, and movement means in Matusik [0218] "More generally, the approaches described above can be implemented, for example, using a programmable computing system executing suitable software instructions or it can be implemented in suitable hardware such as a field-programmable gate array (FPGA) or in some hybrid form. For example, in a programmed approach the software may include procedures in one or more computer programs that execute on one or more programmed or programmable computing system (which may be of various architectures such as distributed, client/server, or grid) each including at least one processor, at least one data storage system (including volatile and/or non-volatile memory and/or storage elements), at least one user interface (for receiving input using at least one input device or port, and for providing output using at least one output device or port).").
Claim 22:
Matusik in view of Susnjara teaches “The plant according to claim 19, characterized in that the information and the instructions for the computerized numerical control system are contained in a file obtained by means CAM software based on a model of the article obtained by means of 3D or solid modelling CAD software,” (Matusik teaches that the specifications of objects can be represented as a solid model in Matusik [0193] "In some examples, the specifications of objects (or parts of objects) can be represented as any one of a mesh: e.g., a list of vertices and triangles; a triangle soup, a boundary representation (brep), a solid model: a voxel grid (uniform and adaptive), a tetrahedral model, a solid model with material assignment, a procedural model, and implicit model."; Matusik teaches that the predistorter 215 provides a modified specification such as printer commands i.e. operations to be performed for manufacture of the three-dimensional article in Matusik [0081] "Continuing to refer to FIG. 1, the printing system 100 is characterized by a parameterized predistorer 115, which takes as input the specification 105 and produces a modified specification 120 (or other suitable input to the printer 125, such as printer commands or some sort of printing plan). The terms “predistorter” or “distort” do not connote particular types of operations, and should not be viewed as limiting (i.e., operations other than geometric modification are encompassed). In some or all embodiments, the predistorter is configured to essentially form an inverse (a “pre-inverse”) of a printing process, such that when the modified specification 120 is provided to a printer 125 and fabricated to form a corresponding fabricated part 175 of the object, that part 175 matches the originally desired specification 105 that is provided as input."), and
“the CAD and CAM software being interfaced with said software based on at least one artificial intelligence algorithm so as to be trained in such a way as to optimize the procedures for analysis and decomposition of the article and for obtaining instructions for the manufacture of the article.” (Matusik teaches that the model parameters may be updated using an artificical neural network in Matusik [0106-108] "For example, training data includes a number of corresponding pairs (yi,zi), where zi is the difference between successive scans (e.g., represented as Z1−Zi-1) representing the incremental result of printing according to the input yi, of modified input specifications 120 and scans 155 are stored during one or more previous printing passes. That training data is then processed in a batch training procedure to determine predistorter parameters ΘG for the predistorter G 215 and simulation parameters ΘĤ for the simulator Ĥ 216. Referring to FIG. 4, in one example of such a training procedure, a combined loss 430 is determined as a sum (or weighted sum) of a simulation loss, an inverse design loss, a cycle input 3D model loss, and a cycle manufactured 3D model loss. That combined loss is provided to parameter optimization module 431 that determines the updated predistorter parameters ΘG and the updated simulation parameters ΘĤ. In some examples, the parameter optimization module 431 determines the parameters ΘG and ΘĤ by using an optimization algorithm to iteratively update the parameters ΘG and ΘĤ until some stopping criteria (e.g., number of iterations, convergence). One common way of updating model parameters involves computing/estimating gradients of the combined loss function with respect to the model parameters. For example, a backpropagation algorithm for an artificial neural networks uses such gradient estimates. Of course, other parameter update schemes can be used."; Matusik Fig. 3 [As shown above in claim 6] teaches that the updated parameter ΘG is used to determine the predistorter settings which affects the input specifications 120 supplied to the printer.
Claim 23:
The limitations of claim 23 are substantially the same as claim 17 and it is rejected for the same reasons.
Claims 12 and 24 are rejected under 35 U.S.C. 103 as being unpatentable over Matusik et al. (US20200143006A1, in view of Susnjara et al. (US20190322044A1), further in view of Moler et al. (US20210276263A1).
Claim 12:
Matusik teaches “The method according to claim 7” as described above. Matusik teaches “…the process parameters detected in said step (ii) comprise… the flowrate of the thermoplastic material,” (Matusik teaches measuring jet characteristics such as a degree of clogging which would affect the flowrate in Matusik [0093] "In some implementations, the characteristics of the printer H depend on various environmental or state variables. Examples may include ambient temperature, individual jet characteristics (e.g., degree of clogging) etc. In some implementations, changes in such variables may be captured by updating the parameters ΘĤ and ΘG. In other implementations they are explicitly measured or estimated, and may be represented as mn or {circumflex over (m)}n, and provided as a further input to Ĥ and G.").
Matusik does not appear to explicitly teach “…the process parameters detected in said step (ii) comprise… and the pressure of the thermoplastic material upstream and/or downstream of the pump.” However, Susnjara does teach this claim limitation (Susnjara teaches monitoring the pressure at the inlet of the gear pump 66 in Susnjara [0052] "As the feed rate of the CNC machine changes, representative servo feedback signals from the moving axes are processed in the machine control computer 81 to control the speed of output pump 66, and correspondingly, the speed of extruder screw 63. Stated differently, machine control computer 81 serves to increase and/or decrease the speeds of extruder screw 63 and gear pump 66 based on increases/decreases in movement of CNC machine 1 during a 3D printing manufacturing process. In embodiments where sensor 72 is a pressure sensor, sensor 72 may monitor the pressure at the inlet of gear pump 66, outputting an analog signal into servo controller 79 and/or machine control computer 81, which in turn, influences the servo loop controlling the extruder screw 63 to bias, adjust, or otherwise fine tune the synchronized speed between extruder screw 63 and gear pump 66, in order to compensate for pressure changes at the inlet of gear pump 66.").
Matusik and Susnjara are analogous art because they are from the same field of endeavor of additive manufacturing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having teachings of Matusik and Susnjara before him/her, to modify the teachings of an Intelligent additive manufacturing method of Matusik to include the pressure measurement of Susnjara because adding the Methods and apparatus for processing and dispensing material during additive manufacturing of Susnjara would allow for maintaining an optimal pressure within the extruder without a breaker plate or screen as described in Susnjara [0058-0059] “The ability to generate the required pressure may be accomplished with a lower-cost system that reduces mechanical complexity without the need for a breaker plate or a screen (such as a filter) between an end of the screw 63 and gear pump 66, as shown in FIG. 4A, for example. Control of gear pump 66 may be performed without unduly restricting throughput, resulting in higher flow rates for extruder 60. In an exemplary configuration, nozzle 51 may have an open round shape (FIG. 6) which offers little resistance to material flow. Gear pump 66 may restrict flow to the nozzle 51, thereby avoiding the need to provide a nozzle having significant resistance to material flow. A desired or optimal pressure within extruder 60 may be created and maintained by controlling the relative speeds of the extruder 60 and gear pump 66.” Neither Matusik or Susnjara appear to explicitly teach “The method according to claim 7, characterized in that the process parameters detected in said step (ii) comprise the temperature of the thermoplastic material,” However, Moler does teach this claim limitation (Moler teaches measuring the temperature of polymer material i.e. thermoplastic material in Moler [0050] "In order to especially reliably be able to prevent the polymer material melt from solidifying in the area of the cooling means 4 and/or in the area of the 3D-print head 5, the 3D-printer 1 has several temperature sensors 12 to 14, which can be arranged in the area of the cooling means 4 and in the area of the layered manufacturing means 5 in such a way that the temperature of the polymer material melt can be measured or determined. The temperature sensors 12 to 14 are communicatively connected with the regulating unit 11. The regulating unit 11 can instruct the temperature sensors to record measured temperature values. Alternatively, the temperature sensors 12 to 14 can also be configured to independently perform the temperature measurements. The temperature measurements take place during the cooling, conveying and additive manufacturing process, preferably continuously.").
Matusik, Susnjara, and Moler are analogous art because they are from the same field of endeavor of additive manufacturing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having teachings of Matusik, Susnjara, and Moler before him/her, to modify the teachings of an Intelligent additive manufacturing method of Matusik modified to include the pressure measurement of Susnjara to include the temperature measurement of Moler because adding the Method and device for producing a three-dimensional object of Moler would prevent polymer material melt from solidfying as described in Moler [0050] "In order to especially reliably be able to prevent the polymer material melt from solidifying in the area of the cooling means 4 and/or in the area of the 3D-print head 5, the 3D-printer 1 has several temperature sensors 12 to 14, which can be arranged in the area of the cooling means 4 and in the area of the layered manufacturing means 5 in such a way that the temperature of the polymer material melt can be measured or determined. The temperature sensors 12 to 14 are communicatively connected with the regulating unit 11. The regulating unit 11 can instruct the temperature sensors to record measured temperature values. Alternatively, the temperature sensors 12 to 14 can also be configured to independently perform the temperature measurements. The temperature measurements take place during the cooling, conveying and additive manufacturing process, preferably continuously."
Claim 24:
The limitations of claim 24 are substantially the same as claim 12 and is rejected for the same reasons.
Claim 13 is rejected under 35 U.S.C. 103 as being unpatentable over Matusik et al. (US20200143006A1, in view of Susnjara et al. (US20190322044A1), further in view of Indyk et al. (US20220314537A1).
Claim 13:
Matusik in view of Susnjara teaches “The method according to claim 9, characterized in that the process parameters detected in said step (ii) comprise the speed of displacement of the at least one extruder device,” (Susnjara teaches a flow sensor which would monitor the speed of displacement in Susnjara [0051] "The speed of extruder screw 63 also may be modified by inputs from one or more sensors 72 (e.g., a pressure sensor or a flow sensor) operably coupled to the extruder."), and
“the process parameters detected in said step (ii) comprise the… speed of rotation of the pump” (Susnjara Fig. 4A teaches the servo controller receiving pump feedback, which may include the pump speed and that the speed of the gear pump may be determined as a function of the speed of the extruder screw and vice versa in Susnjara [0051] "A stable flow rate into conduit 52 and through application nozzle 51 may be regulated by providing servo control of the speed of gear pump 66, through an exemplary controller formed by the machine's control computer 81 and servo control system, based on the speed of the CNC machine's moving axes. The speed of extruder screw 63 likewise may be regulated in proportion with the speed of gear pump 66 by a servo control loop. A signal from the gear pump servo loop is processed to control the output of the extruder servo drive in proportion with that of gear pump 66, thus synchronizing the speed of the extruder with that of the gear pump by a predetermined proportion. In other words, the operation speed of gear pump 66 and extruder screw 63 may be dependent on one another. That is, the speed of extruder screw 63 may be determined as a function of the speed of gear pump 66, and vice versa. The speed of extruder screw 63 also may be modified by inputs from one or more sensors 72 (e.g., a pressure sensor or a flow sensor) operably coupled to the extruder."
[AltContent: rect]
PNG
media_image5.png
719
572
media_image5.png
Greyscale
), and
“the process parameters detected in said step (ii) comprise the… speed of rotation of the screw of the extruder.” (Susnjara Fig. 4A teaches the servo controller receiving screw feedback, which may include the screw speed and that the speed of the gear pump may be determined as a function of the speed of the extruder screw and vice versa in Susnjara [0051] "A stable flow rate into conduit 52 and through application nozzle 51 may be regulated by providing servo control of the speed of gear pump 66, through an exemplary controller formed by the machine's control computer 81 and servo control system, based on the speed of the CNC machine's moving axes. The speed of extruder screw 63 likewise may be regulated in proportion with the speed of gear pump 66 by a servo control loop. A signal from the gear pump servo loop is processed to control the output of the extruder servo drive in proportion with that of gear pump 66, thus synchronizing the speed of the extruder with that of the gear pump by a predetermined proportion. In other words, the operation speed of gear pump 66 and extruder screw 63 may be dependent on one another. That is, the speed of extruder screw 63 may be determined as a function of the speed of gear pump 66, and vice versa. The speed of extruder screw 63 also may be modified by inputs from one or more sensors 72 (e.g., a pressure sensor or a flow sensor) operably coupled to the extruder.").
Neither Matusik or Susnjara appear to explicitly teach “the process parameters detected in said step (ii) comprise the… inclination of the nozzle,” or “the process parameters detected in said step (ii) comprise the… distance of the nozzle from a table supporting the three-dimensional article or from a surface of the three-dimensional article,” However, Indyk does teach these claim limitations.
Indyk teaches “the process parameters detected in said step (ii) comprise the… inclination of the nozzle,” (Indyk teaches a positioning system which can detect the printing angle in Indyk [0056] "Positioning system 160 can also include one or more current position sensors 168, which can operate to detect the position of the printing head, the movement of the printing head, the printing angle, and/or the curing angle, among other system operational aspects. Current position sensor(s) can be end or continuous sensors and can be any suitable type of sensor(s), such as, for example, electric, magnetic, mechanical, ultrasonic, laser, and the like."), and
“the distance of the nozzle from a table supporting the three-dimensional article or from a surface of the three-dimensional article,” (Indyk teaches a positioning system which can detect the position of the printing head which would determine the distance of the nozzle from a table in Indyk [0056] "Positioning system 160 can also include one or more current position sensors 168, which can operate to detect the position of the printing head, the movement of the printing head, the printing angle, and/or the curing angle, among other system operational aspects. Current position sensor(s) can be end or continuous sensors and can be any suitable type of sensor(s), such as, for example, electric, magnetic, mechanical, ultrasonic, laser, and the like.").
Matusik, Susnjara, and Indyk are analogous art because they are from the same field of endeavor of additive manufacturing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having teachings of Matusik, Susnjara, and Indyk before him/her, to modify the teachings of an Intelligent additive manufacturing method of Matusik modified to include the pump and extruder comprising a screw of Susnjara to include the measurement of printing angle and position of the printing head of Indyk because adding the Three-dimensional printing head with adjustable printing angle of Indyk would avoid or reduce warpage in printed materials as described in Indyk [0029] " The many advantages provided by the disclosed embodiments enable unique manufacturing methods of 3D printing to produce building structures from floors to ceilings, to print around structural frames of buildings, to print complex infill portions of building structures or building components, and to print lightweight and durable complex structures, such as hyperboloid structures. 3D printing at an adjusted angle also avoids or significantly reduces warpage in the printed materials and allows for more robust and stronger products in the case of curved or complex geometries. These and other advantages provide significant improvements over traditional 3D printing processes of buildings and the parts of thereof in terms of efficiency, labor cost reduction, high quality and preciseness in automation levels, and the variety of the designs may be printed utilizing the disclosed technologies.”
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Gmeiner et al. (US20210187859A1) teaches an encoder that senses the position of the print head relative to the building platform in Gmeiner [0092] “In order to achieve desired accuracy of the printing process, the feeding rate of the pattern data feeder is synchronized with the movement speed of the print head 3. To this end a linear encoder 11 may be provided on the print head 3 that is moved along a stationary linear encoder bar 12 so as to sense the position and/or the velocity of the print head 3 relative to the building platform 1."
Pinskiy et al. (US20200247063A1) teaches a 3d printer which includes an image analyzer to improve upon a design in Pinskiy [0093] "In some embodiments, image analyzer 180 can be configured to detect a correlation between one or more print parameters and fewer anomalies in a printed layer and/or a completed printed object. In further embodiments, image analyzer 180 can be configured to detect a correlation between one or more print parameters and the measured mechanical, optical and/or electrical properties of a completed printed object. In response to detecting one or more such correlations, image analyze can provide information, data, and/or instructions which alter the manner in which one or more layers of an object being printed or one or more objects to be printed in the future are printed. In some embodiments, image analyzer 180 can provide information, data, and/or instructions, for example to a three-dimensional modeling software, to improve upon a production design."; Pinskiy teaches a pre-classifier that can be trained using a convolutional neural network in Pinskiy [0083] "In some embodiments, the pre-classifier can have any suitable topology. For example, in some embodiments, the pre-classifier can be a CNN. As a more particular example, in some embodiments, the pre-classifier can be a CNN having the same topology as the failure classifier described above in connection with 910. In some embodiments, the pre-classifier can be trained in any suitable manner, such as by training the pre-classifier using any suitable training set of training images. In some embodiments, training images included in the training set can be obtained by capturing images of different layers of an object during printing with errors corresponding to different scaled extrusion commands injected into the print at the different layers, such that errors corresponding to different extrusion values are included in the training set."; Pinskiy teaches that an extrusion quality score can be used to modify a printing parameter in Pinskiy [0089] "Note that, in some embodiments, the extrusion quality score determined by the extrusion classifier can be used for any suitable purpose. For example, in some embodiments, the extrusion quality score can be used to modify a printing parameter, such as an extrusion velocity, an extrusion volume, a rate of motion of an extruder head, a temperature of an extruder nozzle, and/or any other suitable parameter. In some embodiments, the extrusion quality score can be provided as an input to a reinforcement learning agent that modifies a printing parameter during printing of an object, such as shown in and described below in connection with FIGS. 10 and 11."
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Zachary A Cain whose telephone number is (571)272-4503. The examiner can normally be reached Mon-Fri 7:00-3:30 CST.
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, Kenneth M Lo can be reached at (571) 272-9774. 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.
/Z.A.C./ Examiner, Art Unit 2116
/KENNETH M LO/ Supervisory Patent Examiner, Art Unit 2116