DETAILED ACTION
This action is responsive to applicant’s communication filed 12/15/2025.
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 .
Status of the Claims
Claims 1-20 are rejected under 35 U.S.C. 103.
Response to Arguments
Applicant's arguments filed 12/15/2025 have been fully considered but they are not persuasive.
Applicant argues on Pages 6-7 of the Remarks that Mehr does not teach providing an input comprising the spatial toolpath data to the machine learning model and obtaining a revised feed rate for the toolpath segment different from the nominal feed rate for the toolpath segment “prior to printing the part”. Applicant argues Mehr does not teach executing these steps “prior to printing the part” because Mehr teaches determining adjusted process characteristics in “real-time” and thus requires re-calculation during each print job. The examiner respectfully disagrees with applicant’s claim construction and interpretation of the teachings of Mehr.
Mehr performing a calculation (using machine learning) in real-time does not exclude the calculation from being executed “prior to printing the part”. Rather, as discussed in the rejections below and indicated by the applicant, the process taught by Mehr is iterative. See at least ¶ 32, 52, 99, 102, and 133. Before the deposition by the deposition apparatus is executed (i.e. before the printing of the part), the machine learning model is used to determine the revised feed rate (adjusted from a nominal rate) and other process characteristic necessary for printing the part. The providing and obtaining steps of the claim are therefore being executed prior to printing the part in each iteration as taught by Mehr.
The claims do not exclude or teach away from an iterative or a real-time approach. In response to applicant's argument that the references fail to show certain features of the invention, it is noted that the features upon which applicant relies (i.e., a “pre-print process, the results of which can be deployed during multiple subsequent print jobs”; see Page 7 of the Remarks) are not recited in the rejected claims. Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993).
Applicant argues on Page 7 of the Remarks that Bauer fails to include “an induced feed rate variation” in machine learning training data because Baeur is merely directed to an FFT, not machine learning. In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
In this case, the use of machine learning and training a machine learning model is taught by the primary reference, Mehr, and it is not necessary for Baur to also teach training of a machine learning model. Mehr further teaches including feed rates as part of the training data (see ¶ 113). Mehr is deficient in teaching that the training feed rates include induced feed rate variation. However, Baeur teaches a process of introducing feed rate variation and teaches an advantage of introducing feed rate variation is mitigating flow interruptions. See ¶ 5-6, 13, 45, 122, and 138. Given Baur’s teachings, it would have been obvious to a person of ordinary skill in the art for a machine learning system that determines adjusted feed rates for a printing process, such as the system taught by Mehr, to include induced feed rate variation for the purpose of training the machine learning system to determine the feed rates that would reduce flow interruptions, a benefit taught by Baeur.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-8 and 11-18 are rejected under 35 U.S.C. 103 as being unpatentable over MEHR (US 2018/0341248 A1) in view of SHIMIZU (US 2018/0210406 A1).
Regarding Claim 1, MEHR teaches a method of providing a feed rate for three-dimensional printing a part, the method comprising: (¶ 52, 72-73, 97: A system and process for “wire feed additive manufacturing… capable of producing different deposited bead widths within a layer by varying process control parameter like travel speed and wire feed rate”.)
obtaining computer readable toolpath instructions for the part, wherein the toolpath instructions specify a nominal feed rate (¶ 52, 67) for a toolpath segment and spatial toolpath data of the toolpath segment; (¶ 84, 87, 96-97, 108: “Process parameters” include a nominal feed rate and traverse speed for a toolpath and spatial toolpath data, including “the location of a deposition apparatus as a function of time (i.e., a tool path), the angle of a deposition apparatus with respect to a deposition direction, the angle of overhang in an intended geometry”. These parameters are set and are adjusted by layer during the additive manufacturing process.)
providing an input comprising the spatial toolpath data to a trained machine learning system, (¶ 111: Real-time characterization data is input into a machine learning system trained on past characterization data to correlate the process parameters with a result of a deposition process, including a specific location in the part being manufactured (¶ 178, 180), and adjust the parameters in real-time. See the summary of the process in ¶ 2-3, 133.)
prior to printing the part, (¶ 32, 99, 102: The process of determining adjusted process characteristics is performed iteratively. Therefore, before printing a part (i.e. step “d” of the process described in ¶ 133) in each iteration, the trained machine learning model is provided with the inputs necessary to output the revised feed rate data and other adjusted process characteristics.)
wherein the trained machine learning system has been trained using training data comprising: training spatial toolpath data… and training feed rate data; (¶ 113: The process parameters, such as the spatial toolpath data discussed above and “a rate of material deposition, a rate of displacement for a deposition apparatus” (i.e. feed rate), is used to “generate process characterization data”, which is training data.)
obtaining a revised feed rate for the toolpath segment different from the nominal feed rate for the toolpath segment, (The “traverse speed” and “wire feed rate” in particular are adjusted because (¶ 52) “Careful adjustment of these parameters is necessary in order to attain stable deposition on a flat surface.” See ¶ 3 and 5 for further context.)
prior to printing the part, (¶ 32, 52, 133: The process is iterative. In an iteration, before the deposition is executed using the deposition apparatus (i.e. step “d” of the process described in ¶ 133), the machine learning model is used to provide the revised feed rate and other process characteristics.)
wherein the revised feed rate is output from the trained machine learning system; (¶ 5: “the one or more process control parameters to be predicted or adjusted comprise a rate of material deposition, a rate of displacement for a deposition apparatus, a rate of acceleration for a deposition apparatus”. See the summary of the process in ¶ 2-3, 133, including inputting real-time “characterization data” (namely, training data of past feed rates and spatial toolpath data) into a machine learning model to determine adjusted “process control parameters”, including adjusted feed rates (i.e. the output of the model).)
and providing revised computer readable toolpath instructions, wherein the revised machine learning toolpath instructions comprise the revised feed rate. (¶ 133: “c) providing a predicted optimal set or sequence of one or more process control parameters for fabricating the object [i.e. revised computer readable toolpath instructions], wherein the predicted optimal set of one or more process control parameters are derived using a machine learning algorithm that has been trained using the training data set of step (b); and d) performing the deposition process, e.g., an additive manufacturing process, to fabricate the object, wherein real-time process characterization data is provided by one or more sensors as input to the machine learning algorithm to adjust one or more process control parameters in real-time.” As discussed in ¶ 3, 52, 97, 180, the adjusted process control parameters include an adjusted “wire feed rate”, “traverse speed”, or “a rate of material deposition, a rate of displacement for a deposition apparatus” (i.e. feed rate).)
MEHR does not explicitly teach that the training data comprises training closed loop gain data.
However, SHIMIZU, which is directed to machine learning for a numerical controller, teaches training data for a machine learning process for adjusting controller settings, including training closed loop gain data (¶ 29, 72, 83: Feed rate data and closed loop (namely, PID) gain data is input into a machine learning device that correlates the gain data with other training data in order to determine “adjustment of an override control setting value” for a numerical controller for a machine.)
Before the effective filing date of the invention, it would have been obvious to one
of ordinary skill in the art to modify the machine learning algorithm for adjusting process control parameters for an additive manufacturing process, including adjusted feed rates, taught by MEHR by including closed loop gain data as input into the machine learning algorithm as taught by SHIMUZU. Since the references are similarly directed to using machine learning to adjust control parameters for a manufacturing device, the combination would have yielded predictable results. Including gain data as input to the model would have been obvious since SHIMIZU (¶ 92-93) teaches that there is a correlation between the gain values and the feed rates and the accuracy of the machining process. SHIMIZU (¶ 2) further teaches an advantage of factoring gain data into the automated process parameter adjustment of MEHR is that “time and effort for adjusting the gain may be eliminated”.
Claim 11 is directed to a system but otherwise recites the same limitations as claim 1. Claim 11 is therefore rejected using the same reasoning discussed above.
Regarding Claim 2, MEHR in view of SHIMIZU further teaches further comprising printing the part using the revised computer readable toolpath instructions. (MEHR, ¶ 133: “d) performing the deposition process, e.g., an additive manufacturing process, to fabricate the object, wherein real-time process characterization data is provided by one or more sensors as input to the machine learning algorithm to adjust one or more process control parameters in real-time.” Using the revised instructions determined by the machine learning algorithm, the deposition process is performed, i.e. the part is printed. See ¶ 178.)
Claim 12 recites the same limitations as claim 2 and is rejected for the same reasoning.
Regarding Claim 3, MEHR in view of SHIMIZU further teaches wherein the input further comprises a predetermined closed loop gain value. (SHIMIZU, ¶ 72: “a current value of an override control setting value (such as each gain of PID control) according to a program, is input to a machine learning device 20 as state information.” A current gain of PID control is a predetermined closed loop gain value.)
The same reason to combine discussed in the rejection of claim 1 applies to claim 3.
Claim 13 recites the same limitations as claim 3 and is rejected for the same reasoning.
Regarding Claim 4, MEHR in view of SHIMIZU further teaches wherein the training data comprises data that is specific to the part, the method further comprising training the trained machine learning system using the training data. (MEHR, ¶ 113: Process characterization data, which is training data, includes “a chemical composition of a deposited material, a segregation state of constituents in a deposited material, a geometrical property of a deposited material (e.g., a local curvature of a printed part)”, which is data specific to a part. Also see ¶180, which discusses correlating process parameters with specific portions of a part.)
Claim 14 recites the same limitations as claim 4 and is rejected for the same reasoning.
Regarding Claim 5, MEHR in view of SHIMIZU further teaches wherein the toolpath instructions specify an anisotropic region of the part that comprises the toolpath segment. (MEHR, ¶ 102, 108: The process control parameters, which determine the toolpath instructions, specify an anisotropic region of the part, namely “the angle of overhang in an intended geometry”.)
Claim 15 recites the same limitations as claim 5 and is rejected for the same reasoning.
Regarding Claim 6, MEHR in view of SHIMIZU further teaches wherein the anisotropic region comprises at least one of: an overhang, an outer perimeter, or a thin feature. (MEHR, ¶ 102, 108: The process control parameters, which determine the toolpath instructions, specify an anisotropic region of the part, namely “the angle of overhang in an intended geometry”.)
Claim 16 recites the same limitations as claim 6 and is rejected for the same reasoning.
Regarding Claim 7, MEHR in view of SHIMIZU further teaches wherein the spatial toolpath data for the toolpath segment comprises at least one of: a distance from an edge of the part, a distance from a corner of the part, or a rounding radius of the toolpath segment. (MEHR, ¶ 96: “The main steps for generating MAT-based tool paths are: (i) computation of the medial axis; (ii) decomposition of the geometry into one or more regions or domains, where each domain is bounded by a portion of the medial axis and a boundary loop; (iii) generation of the tool path for each domain by offsetting from the medial axis loop toward the corresponding boundary loop with an appropriate step-over distance.” Toolpath data includes a distance from a boundary, i.e. a distance from an edge of the part. ¶ 95: The MAT tool paths also include “an associated radius function of maximally inscribed circles”.)
Claim 17 recites the same limitations as claim 7 and is rejected for the same reasoning.
Regarding Claim 8, MEHR in view of SHIMIZU further teaches wherein the training data further comprises training overhang angle data. (MEHR, ¶ 113: The characterization data, which is training data, includes “an angle of overhang in a deposited geometry, an angle of overhang in an intended geometry”.)
Claim 18 recites the same limitations as claim 8 and is rejected for the same reasoning.
Claims 9-10 and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over MEHR (US 2018/0341248 A1) in view of SHIMIZU (US 2018/0210406 A1) and further in view of BAUER (US 2019/0022725 A1).
Regarding Claim 9, MEHR in view of SHIMIZU teaches all the limitations of claim 1, on which claim 9 depends.
MEHR further teaches the training data of the machine learning algorithm includes the training feed rate data (¶ 113: The process parameters, such as the spatial toolpath data discussed above and “a rate of material deposition, a rate of displacement for a deposition apparatus” (i.e. feed rate), is used to “generate process characterization data”, which is training data.).
MEHR in view of SHIMIZU does not teach wherein the training feed rate data comprises an induced feed rate variation.
However, BAUER, which is directed to additive manufacturing using variable feed rates, teaches (¶ 13) a printer for fabricating a three-dimensional object based on a computerized model of geometry of the object, the object having an interior geometry and an exterior geometry, that comprises an induced feed rate variation. (¶ 6, 13, 122, 138: Feed rate profiles include varying or periodic feed rates based on the build layer and whether an interior or exterior portion of an object is being printed. The feed rates are correlated with other process parameters, including the “printing geometry”.)
Before the effective filing date of the invention, it would have been obvious to one
of ordinary skill in the art to modify the training data used to train a machine learning algorithm to adjust process parameters for an additive manufacturing process, including training feed rate data, taught by MEHR in view of SHIMIZU by including an induced feed rate variation as taught by BAUER as part of the training data since there is a correlation between variable feed rates and mitigating flow interruptions, such as clogs and clumping (¶ 43). Since the references similarly teach varying feed rates during an additive manufacturing process, the combination would have yielded predictable results. As taught by BAEUR (¶ 5), “By employing time-varying build material feed rates within the extruder, these artifacts can be mitigated and resulting flow interruptions can be avoided.”
Claim 19 recites the same limitations as claim 9 and is rejected for the same reasoning.
Regarding Claim 10, MEHR in view of SHIMIZU and BAUER further teaches wherein the induced feed rate variation comprises a periodic induced feed rate variation along an edge of a feature. (BAUER, ¶ 13: “a feed rate controller configured to vary the feed rate that the drive system feeds the build material into the nozzle according to a predetermined feed rate profile which predetermined feed rate profile can be broken up into blocks of time, some of which blocks of time relate to fabricating the interior geometry of the object and some of which blocks of time relate to fabricating the exterior geometry of the object”. The feed rate profile is periodic and includes varying the feed rate along an exterior geometry, i.e. along an edge of a feature. Also see ¶ 6-7, 122, which discusses using a periodic feed rate.)
The same motivation to combine discussed in the rejection of claim 9 applies to claim 10.
Claim 20 recites the same limitations as claim 10 and is rejected for the same reasoning.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant’s disclosure.
Paddock (US 2022/0347930 A1) teaches making adjustments to a printing process using a simulation prior to a start of printing. (¶ 88-91)
Picard (US 2023/0021335 A1) teaches determining the optimum manufacturing parameters for the overhang or overhangs of a target part before manufacturing the part. (¶ 16, 98)
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to RAMI RAFAT OKASHA whose telephone number is (571)272-0675. The examiner can normally be reached M-F 10-6 EST.
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, SCOTT BADERMAN can be reached at (571) 272-3644. 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.
/RAMI R OKASHA/Primary Examiner, Art Unit 2118
Prosecution Insights