DETAILED ACTION
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 .
This action is in response to the communication filed 10/14/2025.
No prior art is applied to Claim 1 for the reasons noted in the Office Action of 4/14/2025.
Response to Arguments
Applicant's arguments filed 10/14/22025 have been fully considered but they are not persuasive.
The declaration filed 10/14/2025 is acknowledged.
With respect to the declaration and pages 8-10 of the amendment,
First, the Examiner acknowledges applicant’s experience regarding machine learning models. However, the Examiner nevertheless notes that the instant application is directed towards a metal detector, with the primary issue being the manner in which various signals from different metal detector sensors are implemented. Applicant has, respectfully, not demonstrated any experience with regard to metal detectors, metal detector sensors, or any related field. One of the primary issues here is not whether information could be obtained from any particular generic model or algorithm, but rather what applicant does with specific metal detector sensor signals in order to detect a metallic object, and whether such a feature was conventional such that a further explanation was unnecessary. To that point, the Examiner respectfully disagrees and notes that the Mr. Filios has not reasonably demonstrated that he is a person of ordinary skill in the metal detecting art, but where such skill is necessary in order to reasonably explain both the manner in which a controller can use various signals from metal detector sensors to identify a metallic object, whether a person of ordinary skill in the art would have found such a use as being conventional, and whether such a use would require undue experimentation.
Second, the primary issue in Claim 1 with regard to written description is the manner in which applicant implements “a controller that is configured to … (b) detect the metallic object based on the metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal, and (c) output to a user interface, information to the user indicating the detection of the metallic object” in the last paragraph. Mr. Filios’s declaration merely argues what could be done with various aspects of the data using various algorithms. However, none of this explanation reasonably addresses, within the four corners of the application, what the controller does with all of the above signals to indicate the detection of a metallic object. Meaning, from a written description perspective, applicant must reasonably convey the manner in which the controller uses the above collective set of signals to indicate the detection of metal. Here the test is not whether undisclosed methods could have been used, but rather, the test is whether the original disclosure itself reasonably demonstrates the manner that applicant is implementing such a feature. Based upon the evidence of record, it was not well-known to collectively use these signals together to identify or otherwise detect metal, especially by way of a neural network or other similar machine learning model, and thus an explanation reasonably establishing the manner in which the controller implements such a feature is necessary. However, the original disclosure does not reasonably provide this explanation. The declaration also does not reasonably establish what applicant is doing with the above signals to detect a metallic object, nor does it reasonably establish that applicant ever intended to include any of the various algorithms mentioned in the declaration.
Furthermore, the declaration, respectfully, generically mentions various undisclosed algorithms, but it does not reasonably establish the manner in which any of those algorithms were actually used or could be used with all of the above signals to detect a metal object. Meaning, the declaration merely states that various approaches could be used, such that how data can be optimized or what a model would most likely use, but it never explains the manner in which any of these approaches would use the above signals, as explained, to detect a metallic object. Meaning, it does not reasonably explain what the controller does with any of the signals it receives such that it can detect a metallic object as claimed. This declaration, respectfully, is essentially devoid of any detail explaining the manner in which the signals are used for the claimed purpose, as it does not reasonably explain the manner in which any of the approaches would yield the detection of metal or that any of the disclosed approaches must have been found in the original disclosure. Instead, the declaration amounts to opinion evidence, in which it mentions various algorithms or models, but never provides any details of the manner or how these models or algorithms would use the claimed signals to detect a metallic object. As explained in MPEP 716.01(c)(I), “TO BE OF PROBATIVE VALUE, ANY OBJECTIVE EVIDENCE SHOULD BE SUPPORTED BY ACTUAL PROOF,” and in MPEP 716.01(c)(III), “In assessing the probative value of an expert opinion, the examiner must consider the nature of the matter sought to be established, the strength of any opposing evidence, the interest of the expert in the outcome of the case, and the presence or absence of factual support for the expert’s opinion … Although an affidavit or declaration which states only conclusions may have some probative value, such an affidavit or declaration may have little weight when considered in light of all the evidence of record in the application.”
Here, the declaration does not reasonably provide any explanation demonstrating that the original disclosure would necessarily have included any of the various algorithms and models, or provide a single complete explanation based upon any of these algorithms or models as to the manner in which the controller uses these algorithms or models to detect a metallic object using the signals it is claimed to use. In the rejections found below, the Examiner has specifically identified claim features that raise an issue of written description, and has provide a detailed explanation as to why these features are not reasonably disclosed such that a person of ordinary skill in the art would recognize that applicant had possession of the claim features. The Examiner respectfully notes that applicant’s arguments and presented declaration do not reasonably address these specific arguments or specifically explain why applicant does have support for these claim features. As such, the declaration does not reasonably establish proper written description.
Third, while enablement more broadly addresses whether a person of ordinary skill in the art would be able to figure any reasonable way of implementing the claim feature without undue experimentation, even if that way was not applicant’s way, the declaration does not reasonably establish that it was well-known to use any of the algorithms or models to implement the above claim feature, or that such an implementation would not require undue experimentation. The declaration does not provide a single complete example demonstrating how any of the algorithms or models could be used to implement the claim feature, and instead merely generically asserts that things like classification “is a simple SoftMax layer at the end of the neural network” that “would most likely use a cross entropy loss or connectionist temporal transition.” However, such a statement 1) does not reasonably establish how any of these features would enable a controller to use the signals to detect a metallic object, and 2) stating what is a disclose would “most likely use” does not reasonably establish what would actually be used or known to a person in the metal detecting arts to enable the controller to implement the claim feature.
No single example is presented that reasonably establishes the manner in which the controller would detect metal in the claimed manner, or that any particular use of any of the algorithms or models were well-known or sufficiently well known to enable the controller to implement the claimed detection of metal without undue experimentation.
Lastly, the Examiner respectfully notes that the declaration does not reasonably identify any particular claim to reasonably establish which specific claim features it was intended to address, and as best understood, not all rejected claim limitations are reasonably addressed.
The Examiner notes that applicant has bolded a request for an interview; however, the Examiner respectfully notes that applicant has filed a written amendment with the Office, and as explained in 37 C.F.R 1.2, “All business with the Patent and Trademark Office should be transacted in writing.” As such, the following Office Action is in response to the amendment filed. Should applicant desire an interview, applicant is invited to contact the Examiner to discuss setting up such an interview.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-14 and 16-20 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
MPEP 2163.02 “An applicant shows that the inventor was in possession of the claimed invention by describing the claimed invention with all of its limitations using such descriptive means as words, structures, figures, diagrams, and formulas that fully set forth the claimed invention. Lockwood v. Am. Airlines, Inc., 107 F.3d 1565, 1572, 41 USPQ2d 1961, 1966 (Fed. Cir. 1997).”
MPEP 2163.03(V) “While there is a presumption that an adequate written description of the claimed invention is present in the specification as filed. In re Wertheim, 541 F.2d 257, 262, 191 USPQ 90, 96 (CCPA 1976), a question as to whether a specification provides an adequate written description may arise in the context of an original claim. An original claim may lack written description support when (1) the claim defines the invention in functional language specifying a desired result but the disclosure fails to sufficiently identify how the function is performed or the result is achieved or (2) a broad genus claim is presented but the disclosure only describes a narrow species with no evidence that the genus is contemplated. See Ariad Pharms., Inc. v. Eli Lilly & Co., 598 F.3d 1336, 1349-50 (Fed. Cir. 2010) (en banc). The written description requirement is not necessarily met when the claim language appears in ipsis verbis in the specification. "Even if a claim is supported by the specification, the language of the specification, to the extent possible, must describe the claimed invention so that one skilled in the art can recognize what is claimed. The appearance of mere indistinct words in a specification or a claim, even an original claim, does not necessarily satisfy that requirement."Enzo Biochem, Inc. v. Gen-Probe, Inc., 323 F.3d 956, 968, 63 USPQ2d 1609, 1616 (Fed. Cir. 2002).”
As to Claim 1,
The phrase “a controller that is configured to … (b) detect the metallic object based on the metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal, and (c) output to a user interface, information to the user indicating the detection of the metallic object” in the last paragraph lacks proper written description.
The original disclosure does not reasonably explain the manner in which all of the above signals, collectively, are used to detect a metallic object. The original disclosure does not reasonably disclose a complete flow chart, formula, or other reasonable explanation to demonstrate the manner in which applicant is implementing the claim features to reasonably demonstrate possession. While applicant does disclose the use of a neural network, the mere disclosure of a neural network does not reasonably demonstrate proper written description because applicant does not disclose the specific neural network that applicant is relying upon, and applicant is not, as best understood, relying upon any “off the shelf” neural networks currently available and that are well known. Instead, as best understood, applicant has designed a specific neural network with a specific number of nodes and trained in a specific manner in order to achieve the claimed features. However, applicant does not reasonably provide or disclosure such a network. Merely stating that a neural network, trained learning model, or other similar level of disclosure does not reasonably demonstrate proper written description because these are essentially concepts and not are not specific enough in a disclosed implementation to reasonably apprise a person of ordinary skill in the art of the manner in which applicant is implementing these features.
First, while the model is not claimed in this claim, as best understood, it is the intended manner in which applicant identifies a metallic object from all of the above signals noted above. To that extent, with regard to the model itself, the Examiner acknowledges that applicant discloses one example of the model can be a neural network. However, merely stating that a neural network can be used as the model does not reasonably and sufficiently provide proper written description because such a disclosure is not sufficiently demonstrating the manner in which the model or neural network is implemented. The Examiner acknowledges that applicant is not required to disclose that which is well known, and applicant can rely upon that which is well known. However, merely disclosing that a model such as a neural network can be used is not sufficient to demonstrate the manner in which the network is implemented.
As a simple example, if applicant was implementing a claim feature using a well-known equation, such as a Fourier series, than the mere statement that such a series was used may reasonably demonstrate proper written description because such a formula is well known in the art. A Fourier series is a well-defined equation, and when relying upon the well-known version of the Fourier series, all Fourier series implementations are essentially the same, differing only in the actual variables input into the equation. Here, no additional specific information would be required by applicant as applicant is relying upon a well-known and specific enough equation that a person of ordinary skill in the art would reasonably be able to implement.
However, the same cannot be said for a neural network. No two neural networks are the same as each neural network is tailored to a specific need, and where the number of neurons, hidden layers, training variables such as the weights, and input variables all differ from one neural network to the other. Meaning, unlike the use of a well-known equation such as the above noted Fourier series, the actual implementation of each neural network goes beyond merely having different variables input into the work. The actual construct of the neural network itself, such as the number of neurons, hidden layers, weights, and the manner in which the network is trained, are all decided by the individual creating the neural network. It is this construct that is equivalent to the Fourier series, but it is for this reason that applicant must provide a sufficient amount of information about the manner in which the neural network is constructed in order to establish proper written description. This is because no single off the shelf neural network can be taken and immediately used for the above claimed purpose. As such, no single neural network, well known or otherwise, could be used for the above claimed purpose without being structurally modified to suit the need of the instant application.
Even the manner in which the neural network is trained matters because, for example, if two individuals were both given the same basic neural network, the manner in which it is trained, including the weights used, can cause these two individuals to arrive at two different neural networks with differencing levels of accuracy.
As such, the actual construct of the neural network is a critical feature that must reasonably be disclosed to demonstrate possession and proper written description. However, applicant does not reasonably provide such information, and instead merely states that a neural network is used and provides what appears to be an unlabeled network (115) in Figure 4 with no further explanation as to what the various layers or neurons represent or the manner in which this neural network functions.
Furthermore, the Examiner additionally notes that no neural network has been uncovered from any prior art reference that is designed to combine and/or interpret the above metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal, and to subsequently have a configuration that allows the neural network or model to correlate these signals to a plurality of metallic objects. As such, performing this feature was not well known in the art, and applicant must therefore reasonably provide an explanation as to the manner in which applicant implements such a feature to demonstrate possession.
Applicant however does not reasonably provide such an explanation. The application is essentially silent as to the manner in which the networks uses all of the above signals to arrive at an ability to correlate these signals to different metallic objects. Applicant does not reasonably explain, for example, the manner in which the work uses or treats the signals from the coils and the signals from the grid of 3-axis magnetic sensors to allow it to identify an object. Applicant does not explain how the model or neural network treats the orientation signal or movement sensor signal to assist in making such an identification of different metallic objects. The Examiner acknowledges that movement signals and orientation signals can be used to help compensate for features such as unwanted movement or inconsistent movement, but applicant must reasonably explain the manner in which these signals are being used, and explain the manner in which the model or neural networks uses these signals for the desired purpose of identifying different metallic objects. However, the original disclosure does not reasonably provide such information.
While the Examiner has largely addressed the concept of a neural network as applicant has disclosed this as one example of the trained learning model, the Examiner further notes that no other model or any other feature is provided or disclosed that would reasonably perform or allow for the above claim features. Applicant has neither identified any well-known model that could be used in the well-known manner, such as a Fourier series, without any substantive structural changes, or provided the actual model itself that would be reasonably capable of performing this feature. Applicant has also not identified any other feature in the disclosure that can reasonably be used to implement the claim features, as claimed. In short, it is not sufficient to create a box and label it “neural network” or “trained machine learning model” without sufficient detail to demonstrate what these models are to demonstrate proper written description when applicant is not relying upon a specific well-known implementation of any of these particular models. Applicant, as best understood, cannot simply taken any network already designed and merely change the inputs, similar to implementing a Fourier series. Instead, applicant would reasonably need to redesign any well-known network or create one from scratch such that the network is not a well-known network but rather one designed by applicant. However, applicant does snot reasonably provide such a network, thereby not reasonably establishing proper written description. This is because a person of ordinary skill in the art would not reasonably recognizes the manner in which applicant is implementing the network to reasonably demonstrate possession of the claim feature.
As such, the above claim phrase lacks proper written description, because applicant does not reasonably demonstrate possession of the trained machine learning model that is usable by the controller to interpret the above signals and correlate these signals to different types of metallic objects as claimed. The claim also lacks proper written description because applicant does not reasonably disclose the manner in which all of the above signals are used together to detect the metallic object as claimed. Applicant does not reasonably explain the manner in which the controller uses these signals, collectively, in any reasonably specific manner, in order to demonstrate possession. As such, the above phrase lacks proper written description.
As to Claim 2,
The phrase “the controller is further configured to perform operations further comprising: … identifying, based on the collective magnetic image signal of the local magnetic field, the metallic object and based on: determining a material of the metallic object; determining a size of the metallic object; determining a depth of the metallic object; and labeling the metallic object based on the material, size, and depth” on lines three to the end. However, such a feature requires, as best understood, that the trained machine learning model be used to correlate to and identify a particular metallic object using the collective magnetic image signal. For the same reasons explained above in the rejection of Claim 1, applicant does not reasonably demonstrate the manner in which such a model or neural network is implemented to allow it to identify or correlate the signals to any particular metallic object. Applicant does not reasonably disclose the manner in which the controller is able to identify, based on the collective magnetic image signal, the metallic object as claimed, or the manner in which any of the determining and labeling features are implemented. The original disclosure does not reasonably demonstrate proper written description for these features as it does not reasonably provide a sufficient explanation to demonstrate the manner in which applicant is implementing these features such that a person of ordinary skill in the art would reasonably recognize that applicant had possession of these claim features and recognize the manner in which applicant implements these features.
As to Claims 3, 11, 12, and 14,
The phrase “wherein the controller is configured to detect the metallic object using a trained machine learning model, wherein the trained machine learning model is configured to correlate the metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal to a plurality of different types of metallic objects” on lines 1-5 lacks proper written description.
The original disclosure does not reasonably explain the manner in which all of the above signals, collectively, are used to detect a metallic object. The original disclosure does not reasonably disclose a complete flow chart, formula, or other reasonable explanation to demonstrate the manner in which applicant is implementing the claim features to reasonably demonstrate possession. While applicant does disclose the use of a neural network, the mere disclosure of a neural network does not reasonably demonstrate proper written description because applicant does not disclose the specific neural network that applicant is relying upon, and applicant is not, as best understood, relying upon any “off the shelf” neural networks currently available and that are well known. Instead, as best understood, applicant has designed a specific neural network with a specific number of nodes and trained in a specific manner in order to achieve the claimed features. However, applicant does not reasonably provide or disclosure such a network. Merely stating that a neural network, trained learning model, or other similar level of disclosure does not reasonably demonstrate proper written description because these are essentially concepts and not are not specific enough in a disclosed implementation to reasonably apprise a person of ordinary skill in the art of the manner in which applicant is implementing these features.
First, while the model is not claimed in this claim, as best understood, it is the intended manner in which applicant identifies a metallic object from all of the above signals noted above. To that extent, with regard to the model itself, the Examiner acknowledges that applicant discloses one example of the model can be a neural network. However, merely stating that a neural network can be used as the model does not reasonably and sufficiently provide proper written description because such a disclosure is not sufficiently demonstrating the manner in which the model or neural network is implemented. The Examiner acknowledges that applicant is not required to disclose that which is well known, and applicant can rely upon that which is well known. However, merely disclosing that a model such as a neural network can be used is not sufficient to demonstrate the manner in which the network is implemented.
As a simple example, if applicant was implementing a claim feature using a well-known equation, such as a Fourier series, than the mere statement that such a series was used may reasonably demonstrate proper written description because such a formula is well known in the art. A Fourier series is a well-defined equation, and when relying upon the well-known version of the Fourier series, all Fourier series implementation are essentially the same, differing only in the actual variables input into the equation.
However, the same cannot be said for a neural network. No two neural networks are the same as each neural network is tailored to a specific need, and where the number of neurons, hidden layers, training variables such as the weights, and input variables all differ from one neural network to the other. Meaning, unlike the use of a well-known equation such as the above noted Fourier series, the actual implementation of each neural network goes beyond merely having different variables input into the work. The actual construct of the neural network itself, such as the number of neurons, hidden layers, weights, and the manner in which the network is trained, are all decided by the individual creating the neural network. It is this construct that is equivalent to the Fourier series, but it is for this reason that applicant must provide a sufficient amount of information about the manner in which the neural network is constructed in order to establish proper written description. This is because no single off the shelf neural network can be taken and immediately used for the above claimed purpose. As such, no single neural network, well known or otherwise, could be used for the above claimed purpose without being structurally modified to suit the need of the instant application.
Even the manner in which the neural network is trained matters because, for example, if two individuals were both given the same basic neural network, the manner in which it is trained, including the weights used, can cause these two individuals to arrive at two different neural networks with differencing levels of accuracy.
As such, the actual construct of the neural network is a critical feature that must reasonably be disclosed to demonstrate possession and proper written description. However, applicant does not reasonably provide such information, and instead merely states that a neural network is used and provides what appears to be an unlabeled network (115) in Figure 4 with no further explanation as to what the various layers or neurons represent or the manner in which this neural network functions.
Furthermore, the Examiner additionally notes that no neural network has been uncovered from any prior art reference that is designed to combine and/or interpret the above metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal, and to subsequently have a configuration that allows the neural network or model to correlate these signals to a plurality of metallic objects. As such, performing this feature was not well known in the art, and applicant must therefore reasonably provide an explanation as to the manner in which applicant implements such a feature to demonstrate possession.
Applicant however does not reasonably provide such an explanation. The application is essentially silent as to the manner in which the networks uses all of the above signals to arrive at an ability to correlate these signals to different metallic objects. Applicant does not reasonably explain, for example, the manner in which the work uses or treats the signals from the coils and the signals from the grid of 3-axis magnetic sensors to allow it to identify an object. Applicant does not explain how the model or neural network treats the orientation signal or movement sensor signal to assist in making such an identification of different metallic objects. The Examiner acknowledges that movement signals and orientation signals can be used to help compensate for features such as unwanted movement or inconsistent movement, but applicant must reasonably explain the manner in which these signals are being used, and explain the manner in which the model or neural networks uses these signals for the desired purpose of identifying different metallic objects. However, the original disclosure does not reasonably provide such information.
While the Examiner has largely addressed the concept of a neural network as applicant has disclosed this as one example of the trained learning model, the Examiner further notes that no other model or any other feature is provided or disclosed that would reasonably perform or allow for the above claim features. Applicant has neither identified any well-known model that could be used in the well-known manner, such as a Fourier series, without any substantive structural changes, or provided the actual model itself that would be reasonably capable of performing this feature. Applicant has also not identified any other feature in the disclosure that can reasonably be used to implement the claim features, as claimed.
As such, the above claim phrase lacks proper written description, because applicant does not reasonably demonstrate possession of the trained machine learning model that is usable by the controller to interpret the above signals and correlate these signals to different types of metallic objects as claimed. The claim also lacks proper written description because applicant does not reasonably disclose the manner in which all of the above signals are used together to detect the metallic object as claimed. Applicant does not reasonably explain the manner in which the controller uses these signals, collectively, in any reasonably specific manner, in order to demonstrate possession. As such, the above phrase lacks proper written description.
As to Claim 3,
The phrase “labeling the metallic object comprises comparing one or more of the material, size, and depth of the metallic object to machine learning models of labeled objects stored in a database” on lines 5-7 lacks proper written description. Similar to the issues noted in the above rejection of Claim 1, which is incorporated herein, applicant does not reasonably disclose the manner in which applicant labels any of the above material, size, and depth, and applicant does not reasonably disclose any of the above machine learning models as claimed. Merely comparing does not reasonably provide a labeling, and additional steps beyond the comparison are required in order to label the metallic object. Furthermore, what is contained in these claimed machine learning models to reasonably apprise a person of ordinary skill in the art of the manner in which the comparisons are implemented.
As to Claim 6,
The phrase “the controller is further configured to additionally receive the temperature signal and detect the metallic object additionally based on the temperature signal” on lines 4-6 lacks proper written description.
Similar to the above explanation in above Claim 1, this claim fails to provide proper written description because it does not reasonably disclose the manner in which temperature, in combination with all previously claimed signals, are collectively used to detect the metallic object as claimed. For brevity, the above Claim 1 rejection is incorporated herein, and this claim stands rejected for similar reasons. The original disclosure does not reasonably disclose flow charts, formulas, or other reasonable explanation to sufficiently explain the manner in which temperature is used, either by itself or in conjunction with any or all of the previously claimed signals, to detect a metallic object. Applicant does not reasonably explain or demonstrate the manner in which the neural network, model, or any other feature is reasonably established to otherwise implemented to demonstrate possession of the claim feature. As such, this phrase lacks proper written description.
As to Claim 7,
The phrase “the controller is further configured to additionally receive the humidity level signal and detect the metallic object additionally based on the humidity level signal” on lines 4-6 lacks proper written description.
Similar to the above explanation in above Claim 1, this claim fails to provide proper written description because it does not reasonably disclose the manner in which humidity level, in combination with all previously claimed signals, are collectively used to detect the metallic object as claimed. For brevity, the above Claim 1 rejection is incorporated herein, and this claim stands rejected for similar reasons. The original disclosure does not reasonably disclose flow charts, formulas, or other reasonable explanation to sufficiently explain the manner in which humidity level is used, either by itself or in conjunction with any or all of the previously claimed signals, to detect a metallic object. Applicant does not reasonably explain or demonstrate the manner in which the neural network, model, or any other feature is reasonably established to otherwise implemented to demonstrate possession of the claim feature. As such, this phrase lacks proper written description.
As to Claim 8,
The phrase “the controller is further configured to additionally receive the pressure level signal and detect the metallic object additionally based on the pressure level signal” on lines 4-6 lacks proper written description.
Similar to the above explanation in above Claim 1, this claim fails to provide proper written description because it does not reasonably disclose the manner in which pressure level, in combination with all previously claimed signals, are collectively used to detect the metallic object as claimed. For brevity, the above Claim 1 rejection is incorporated herein, and this claim stands rejected for similar reasons. The original disclosure does not reasonably disclose flow charts, formulas, or other reasonable explanation to sufficiently explain the manner in which pressure level is used, either by itself or in conjunction with any or all of the previously claimed signals, to detect a metallic object. Applicant does not reasonably explain or demonstrate the manner in which the neural network, model, or any other feature is reasonably established to otherwise implemented to demonstrate possession of the claim feature. As such, this phrase lacks proper written description.
As to Claim 11,
The phrase “wherein the controller is further configured to perform operations further comprising: classifying one or more of the metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal of the metallic object, iterative training the machine learning model of the metallic object based on the one or more classified signals ” on lines 5-11 introduces new matter. At issue here is that applicant does not originally disclose classifying one signal or anything less than all of the signals, and similarly, applicant does not originally disclose the iterative training a machine learning model using less than all of the classified signals. See for example original Claim 11 which requires the use of the classified “signals,” and not merely any one signal. This phrase therefore introduces new matter.
As to Claim 12,
The phrase “wherein the controller is configured to perform operations further comprising: receiving user input of a frequency of the signal from the metal detecting device; and classifying the signal from the metal detecting device based on the user input and the trained machine learning model” on lines 5-9 lacks proper written description.
Similar to the issues noted in the above rejection of Claim 1, which is incorporated herein, applicant does not reasonably disclose the manner in which applicant classifies the signal as claimed by using either the user input and/or the trained machine learning model. Merely disclosing a use of a trained model without any reasonable explanation as to the manner in which this model is implemented and used in the claimed manner does not reasonably apprise a person of ordinary skill in the art of the manner in which applicant is implementing this claim feature or demonstrate possession of the claim feature. While applicant is not required to provide every detail or that which is well-known in the art, applicant must still provide a sufficient explanation in the form of flow charts, formulas, or other reasonably explanation as set forth in Lockwood v. Am. Airlines, Inc., 107 F.3d 1565, 1572, 41 USPQ2d 1961, 1966 (Fed. Cir. 1997) (see MPEP 2163.02) in order to demonstrate possession. This phrase therefore lacks proper written description.
As to Claim 13,
The phrase “the controller is configured to perform operations further comprising outputting a classification of the metallic object to the user interface” on lines 1-3 lacks proper written description.
To the extent that applicant includes a classification of the metallic object and classified combined signals as claimed, these phrases lacks proper written description. Similar to the issues noted in the above rejection of Claim 1, which is incorporated herein, applicant does not reasonably disclose the manner in which applicant classifies the signal as claimed by using either the user input and/or the trained machine learning model. Merely disclosing a use of a trained model without any reasonable explanation as to the manner in which this model is implemented and used in the claimed manner does not reasonably apprise a person of ordinary skill in the art of the manner in which applicant is implementing this claim feature or demonstrate possession of the claim feature. While applicant is not required to provide every detail or that which is well-known in the art, applicant must still provide a sufficient explanation in the form of flow charts, formulas, or other reasonably explanation as set forth in Lockwood v. Am. Airlines, Inc., 107 F.3d 1565, 1572, 41 USPQ2d 1961, 1966 (Fed. Cir. 1997) (see MPEP 2163.02) in order to demonstrate possession. This phrase therefore lacks proper written description.
As to Claim 14,
The phrase “the trained machine learning model is a deep neural network” on lines 6-7 lacks proper written description. Applicant does not provide any reasonable guidance as to the manner in which such a network is implemented. There is no evidence on the record that applicant is relying upon any specific well-known deep neural network, and instead applicant, as best understood, is merely disclosing a concept of a network without any reasonably explanation as to the manner in which such a network is implemented in the context of the disclosure. Applicant, as best understood, cannot simply taken any network already designed and merely change the inputs, similar to implementing a Fourier series. Instead, applicant would reasonably need to redesign any well-known network or create one from scratch such that the network is not a well-known network but rather one designed by applicant. However, applicant does snot reasonably provide such a network, thereby not reasonably establishing proper written description. This is because a person of ordinary skill in the art would not reasonably recognizes the manner in which applicant is implementing the network to reasonably demonstrate possession of the claim feature. As such, this phrase lacks proper written description.
As to Claim 16,
The phrase “controller is configured to perform operations further comprising classifying metallic objects based on one or more of the metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal from the metal detecting device of the metallic object in a plurality of different surrounding environments” on lines 1-5 lacks proper written description and introduces new matter.
1) Similar to the issues noted in the above rejection of Claim 1, which is incorporated herein, applicant does not reasonably disclose the manner in which applicant classifies the signals as claimed by using either the user input and/or the trained machine learning model. Merely disclosing a use of a trained model without any reasonable explanation as to the manner in which this model is implemented and used in the claimed manner does not reasonably apprise a person of ordinary skill in the art of the manner in which applicant is implementing this claim feature or demonstrate possession of the claim feature. While applicant is not required to provide every detail or that which is well-known in the art, applicant must still provide a sufficient explanation in the form of flow charts, formulas, or other reasonably explanation as set forth in Lockwood v. Am. Airlines, Inc., 107 F.3d 1565, 1572, 41 USPQ2d 1961, 1966 (Fed. Cir. 1997) (see MPEP 2163.02) in order to demonstrate possession. This phrase therefore lacks proper written description.
2) At issue here is that applicant does not originally disclose classifying one signal or anything less than all of the signals. See for example original Claim 11 which requires the use of the classified “signals,” and not merely any one signal. This phrase therefore introduces new matter.
As to Claim 17,
The phrase “the controller is configured to perform operations further comprising classifying the metallic object based on user input, wherein the user input includes at least one of a current location of the metal detecting device or an identification label of at least one metallic object” on lines 1-4 lacks proper written description.
Similar to the issues noted in the above rejection of Claim 1, which is incorporated herein, applicant does not reasonably disclose the manner in which applicant classifies the signal as claimed by using either the user input and/or the trained machine learning model. Merely disclosing a use of a trained model without any reasonable explanation as to the manner in which this model is implemented and used in the claimed manner does not reasonably apprise a person of ordinary skill in the art of the manner in which applicant is implementing this claim feature or demonstrate possession of the claim feature. While applicant is not required to provide every detail or that which is well-known in the art, applicant must still provide a sufficient explanation in the form of flow charts, formulas, or other reasonably explanation as set forth in Lockwood v. Am. Airlines, Inc., 107 F.3d 1565, 1572, 41 USPQ2d 1961, 1966 (Fed. Cir. 1997) (see MPEP 2163.02) in order to demonstrate possession. This phrase therefore lacks proper written description.
As to Claims 2-14 and 16-20,
These claims stand rejected for incorporating and reciting the above rejected subject matter of their respective parent claim(s) and therefore stand rejected for the same reasons.
Claims 1-14 and 16-20 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, because the specification does not reasonably provide enablement for 1) “a controller that is configured to … (b) detect the metallic object based on the metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal, and (c) output to a user interface, information to the user indicating the detection of the metallic object” in the last paragraph of Claim 1; 2) “the controller is further configured to perform operations further comprising: … identifying, based on the collective magnetic image signal of the local magnetic field, the metallic object and based on: determining a material of the metallic object; determining a size of the metallic object; determining a depth of the metallic object; and labeling the metallic object based on the material, size, and depth” on lines three to the end of Claim 2; 3) “wherein the controller is configured to detect the metallic object using a trained machine learning model, wherein the trained machine learning model is configured to correlate the metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal to a plurality of different types of metallic objects” on lines 1-5 of Claims 3, 11, 12, and 14; 4) “labeling the metallic object comprises comparing one or more of the material, size, and depth of the metallic object to machine learning models of labeled objects stored in a database” on lines 5-7 of Claim 3; 5) “the controller is further configured to additionally receive the temperature signal and detect the metallic object additionally based on the temperature signal” on lines 4-6 of Claim 6; 6) “the controller is further configured to additionally receive the humidity level signal and detect the metallic object additionally based on the humidity level signal” on lines 4-6 of Claim 7; 7) “the controller is further configured to additionally receive the pressure level signal and detect the metallic object additionally based on the pressure level signal” on lines 4-6 of Claim 8; 8) “wherein the controller is configured to perform operations further comprising: receiving user input of a frequency of the signal from the metal detecting device; and classifying the signal from the metal detecting device based on the user input and the trained machine learning model” on lines 5-9 of Claim 12; 9) “the controller is configured to perform operations further comprising outputting a classification of the metallic object to the user interface” on lines 1-3 of Claim 13; 10) “the trained machine learning model is a deep neural network” on lines 5-6 of Claim 14; 11) “controller is configured to perform operations further comprising classifying metallic objects based on one or more of the metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal from the metal detecting device of the metallic object in a plurality of different surrounding environments” on lines 1-5 of Claim 16; and 13) “the controller is configured to perform operations further comprising classifying the metallic object based on user input, wherein the user input includes at least one of a current location of the metal detecting device or an identification label of at least one metallic object” on lines 1-4 of Claim 17. Specifically, the Examiner does not find any explanation as to how applicant is implementing the above detecting of the metallic object based on the metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal, and subsequent configuration of the controller to label the metallic object, detect the metallic object additional from temperature, humidity, and pressure signals, detect the above metallic object using a trained machine learning model that is configured to correlate the above metal detecting signal, the collective magnetic image signal, the movement signal, and the orientation signal to a plurality of different types of metallic objects, and classify the metal object as claimed. This is a scope of enablement rejection because the specification does not enable one of ordinary skill to use the invention commensurate with the scope of the claims without undue experimentation.
There