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 .
DETAILED ACTION
Claims 1-20 are pending. Claims 1, 8 and 15 are independent.
This Application was published as U.S. 20260038481.
Apparent priority: 24 July 2021.
This Application is a parent of application no.19/039946 which is a CIP of the instant application.
The instant Application U.S. 18/788,568 and the U.S. Patent No. 12190525 have an inventor in common but are not assigned to the same assignee.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more. The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Step 1: The independent Claims are directed to statutory categories:
Claim 1 is a system claim and directed to the machine or manufacture category of patentable subject matter.
Claim 8 is a method claim and directed to the process category of patentable subject matter.
Claim 15 is a computer-readable-storage device claim and is directed to the machine or manufacture category of patentable subject matter.
Step 2A, Prong One: Does the Claim recite a Judicially Recognized Exception? Abstract Idea? Are these Claims nevertheless considered Abstract as a Mathematical Concept (mathematical relationships, mathematical formulas or equations, mathematical calculations), Mental Process (concepts performed in the human mind (including an observation, evaluation, judgment, opinion), or Certain Methods of Organizing Human Activity (1-fundamental economic principles or practices (including hedging, insurance, mitigating risk), 2-commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations), 3- managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions) and fall under the judicial exception to patentable subject matter?)
The rejected Claims recite Mental Processes or Methods of Organizing Human Activity.
Step 2A, Prong Two: Additional Elements that Integrate the Judicial Exception into a Practical Application? Identifying whether there are any additional elements recited in the claim beyond the judicial exception(s), and evaluating those additional elements to determine whether they integrate the exception into a practical application of the exception. “Integration into a practical application” requires an additional element(s) or a combination of additional elements in the claim to apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception, such that the claim is more than a drafting effort designed to monopolize the exception. Uses the considerations laid out by the Supreme Court and the Federal Circuit to evaluate whether the judicial exception is integrated into a practical application.
The rejected Claims do not include additional limitations that point to integration of the abstract idea into a practical application and are therefore directed to the abstract idea.
Claim 1 is a generic automation of a communication between a flying plane and the air traffic controllers and the air traffic controller can tell whether the plane is at risk or not.
The Claim includes the use of a VAE but as an incidental element without providing a nexus between this element and the rest of the Claim. Accordingly, the VAE is considered a well-understood, routine, and conventional machine components that are being used for their well-understood, routine, conventional and rather generic functions.
1. An aviation anomaly detection system comprising:
an interface configured to receive audio communications between an air traffic control station and a plurality of aircraft; [Jane is sitting in the control tower and listening to the conversation of the pilots and air traffic controllers.]
a speech-to-text converter configured to convert the received audio communications from the interface to text data; and [Jane transcribes the conversations.]
a processor configured to
determine at least one aviation anomaly from the text data with a variational autoencoder (VAE) deep learning model, and [Jane can determine whether one of the planes is in distress from the conversation.] (The VAE is sitting in this Claim in a disconnected fashion and lacks nexus. If the invention is in the use of the VAE, then the particulars of the VAE, perhaps its training or manner of use need to be claimed with particularity in order to create an integration of the abstract idea into a practical application or to cause the claim as a whole to amount to significantly more than the underlying abstract idea.]
generate an alert based upon the at least one aviation anomaly. [Jane can issue a warning to the authorities.]
Step 2B: Search for Inventive Concept: Additional Elements Do not amount to Significantly More: The limitations of display, processors, memory, programs, interface, STT in Claim 1 and counterpart Claims 8 and 15 are well-understood, routine, and conventional machine components that are being used for their well-understood, routine, conventional and rather generic functions. Additionally, these limitations are expressed parenthetically and lack nexus to the Claim language and as such are a separable and divisible mention to a machine. Accordingly, they are not sufficient to cause the Claim as a whole to amount to significantly more than the underlying abstract idea.
The Dependent Claims do not add limitations that could integrate the abstract idea into a practical technological application or could help the Claim as a whole to amount to significantly more than the Abstract idea identified for the Independent Claim:
2. The aviation anomaly detection system of claim 1 wherein the VAE deep learning model comprises a plurality of VAE deep learning models including at least some of an Adaptive Moment Estimation (ADAM) deep learning VAE model, a Stochastic Gradient Descent with Momentum (SGDM) deep learning VAE model, and a root mean square propagation (RMSProp) deep learning VAE model. [This Claim merely lists types of VAEs and still does not establish a nexus with how the VAE is being used in the particular context of the Claim.]
3. The aviation anomaly detection system of claim 2 wherein the processor is configured to select a given VAE deep learning model from among the plurality thereof based upon a game theory reward matrix. [All of the technological features recited in this Claim are expressed generically and without particulars that establish a nexus with the problem at hand and could potentially provide a technological solution (integration into a practical application). How are these techniques used in the particular context of this Application? Include in the Claim.]
4. The aviation anomaly detection system of claim 1 wherein the at least one aviation anomaly comprises at least one of a pilot readback error and a pilot deviation error. [This Claim provides two choices for anomaly: two types of pilot error. Not technological integration. A person can detect pilot error.]
5. The aviation anomaly detection system of claim 1 wherein the processor is further configured to determine aircraft locations from the text data, and determine the at least one aviation anomaly based upon relative positions of determined aircraft locations. [Jane can tell that the plane is not where the pilot thinks it is.]
6. The aviation anomaly detection system of claim 1 wherein the VAE deep learning model is trained based upon a plurality of air traffic communications messages generated from a machine learning (ML) large language model (LLM). [Training is generically based on training samples. In this case the training data are generated by an LLM and are not real data from the field. Particulars of the use of the LLM for generating the samples are needed. Otherwise two generic components each engaged in its generic job.]
7. The aviation anomaly detection system of claim 1 wherein the interface is configured to receive aircraft ground control audio communications and air route control audio communications. [ This is the definition of the communication radio between a pilot and air traffic control and not sufficient to cause the claim to amount to significantly more.]
With respect to Independent Claim 8 and independent Claim 15, which have limitations similar to the limitations of Claim 1 and some other generic components. Accordingly, they do not include additional limitations that can 1) integrate the Abstract Idea into a practical application or 2) cause the Claim as a whole to amount to more than the underlying abstract idea.
Claim 8 is a method claim with limitations corresponding to the limitations of Claim 1 and is rejected under similar rationale.
Claim 9 is a method claim with limitations corresponding to the limitations of Claim 2 and is rejected under similar rationale.
Claim 10 is a method claim with limitations corresponding to the limitations of Claim 3 and is rejected under similar rationale.
Claim 11 is a method claim with limitations corresponding to the limitations of Claim 4 and is rejected under similar rationale.
Claim 12 is a method claim with limitations corresponding to the limitations of Claim 5 and is rejected under similar rationale.
Claim 13 is a method claim with limitations corresponding to the limitations of Claim 6 and is rejected under similar rationale.
Claim 14 is a method claim with limitations corresponding to the limitations of Claim 7 and is rejected under similar rationale.
Claim 15 is a computer program product system claim with limitations corresponding to the limitations of method Claim 1 and is rejected under similar rationale.
Claim 16 is a computer program product system claim with limitations corresponding to the limitations of method Claim 2 and is rejected under similar rationale.
Claim 17 is a computer program product system claim with limitations corresponding to the limitations of method Claim 3 and is rejected under similar rationale.
Claim 18 is a computer program product system claim with limitations corresponding to the limitations of method Claim 4 and is rejected under similar rationale.
Claim 19 is a computer program product system claim with limitations corresponding to the limitations of method Claim 5 and is rejected under similar rationale.
Claim 20 is a computer program product system claim with limitations corresponding to the limitations of method Claim 6 and is rejected under similar rationale.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1-5, 7-12, and 14-19 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Rahmes (U.S. 12,190,525) in view of Tkaczyk (U.S. 20220383866).
Regarding Claim 1, Rahmes teaches:
1. An aviation anomaly detection system comprising: [Rahmes: “Systems And Methods Of Aviation Data Communication Anomaly Detection, As In Air Traffic Control Surveillance Systems.” Title.]
an interface configured to receive audio communications between an air traffic control station and a plurality of aircraft; [Rahmes, “Systems and methods for detecting anomalies in aviation data communication systems (e.g., air traffic control surveillance systems), include a processor receiving device status information….” Abstract. “The invention is directed to aviation data communication anomaly detection. More specifically, the invention is directed to systems and methods for efficiently and effectively monitoring air traffic control surveillance systems, such as Automatic Dependent Surveillance-Contract (ADS-C) and Automatic Dependent Surveillance-Broadcast (ADS-B) systems for anomalies by means of a variational autoencoder.” 1:10-18.]
a speech-to-text converter configured to convert the received audio communications from the interface to text data; and
a processor configured to determine at least one aviation anomaly from the text data with a variational autoencoder (VAE) deep learning model, and [Rahmes, “… A variational autoencoder receives and optimizes the device status information and determines whether it qualifies as an anomaly.”] generate an alert based upon the at least one aviation anomaly. [Rahmes, “Data Deviation—Any deviation from data (anomalies in data). Any deviation from the baseline traffic would indicate an alert and flag the data as suspicious;” 14:23-25.]
Rahmes does not teach STT on audio communications.
Tkaczyk teaches:
a speech-to-text converter configured to convert the received audio communications from the interface to text data; and [Tkaczyk Figure 2, “speech-to-text 204.”]
Rahmes and Tkaczyk pertain to the detection of anomaly in flights from ATC communications and it would have been obvious to use the transcript of the communications from Tkaczyk for analysis of Rahmes as part of the data to be analyzed. This combination falls under combining prior art elements according to known methods to yield predictable results or use of known technique to improve similar devices (methods, or products) in the same way. See MPEP 2141, KSR, 550 U.S. at 418, 82 USPQ2d at 1396.
Regarding Claim 2, Rahmes teaches:
2. The aviation anomaly detection system of claim 1 wherein the VAE deep learning model comprises a plurality of VAE deep learning models including at least some of an Adaptive Moment Estimation (ADAM) deep learning VAE model, a Stochastic Gradient Descent with Momentum (SGDM) deep learning VAE model, and a root mean square propagation (RMSProp) deep learning VAE model. [Rahmes, “In an embodiment, the processor optimizes the device status information by generating a plurality of probabilistic models of the device status information and determining which of the plurality of models is optimal. In an embodiment, the processor determines which of the plurality of models is optimal by applying a game theoretic optimization to the plurality of models and selecting which of the plurality of models to use to generate the n-D point scatter plot in latent space. In an embodiment, the plurality of models includes at least two of Adam, a replacement optimization algorithm for stochastic gradient descent for training deep learning models, stochastic gradient descent with momentum (SGDM), or root mean square propagation (RMSProp).” 2:53-65.]
Regarding Claim 3, Rahmes teaches:
3. The aviation anomaly detection system of claim 2 wherein the processor is configured to select a given VAE deep learning model from among the plurality thereof based upon a game theory reward matrix. [Rahmes, “… A game theoretic optimization is applied to the plurality of models, and the best model is used to generate the n-D point scatter plot in latent space….” Abstract. “Optimization utilizes a linear program to optimally choose which deep learning model to use per data point. A reward matrix, A, is created with data image loss values for different solvers.” 7:18-21.]
Regarding Claim 4, Rahmes does not list the pilot error types.
Tkaczyk teaches:
4. The aviation anomaly detection system of claim 1 wherein the at least one aviation anomaly comprises at least one of a pilot readback error and a pilot deviation error. [Tkaczyk includes an example of a readback error: “[0034] In one example, an air traffic controller could instruct a pilot to land an aircraft at a specific runway (e.g., Runway 9B) but instead of confirming the same runway, the pilot confirms the instruction to land at the wrong runway (e.g., Runway 9E). In this example, due to the similar sounding runway identifiers 9B and 9E, both the air traffic controller and the pilot might consider their communications to be correctly understood. While this is a simple miscommunication, if this potential anomaly is left uncorrected, it could result in an anomaly occurring that can have significant negative outcomes.”]
Rahmes and Tkaczyk pertain to the detection of anomaly in flights from ATC communications and it would have been obvious to use the particular error type taught by Tkaczyk as anomalies detected by the system of Rahmes.
Regarding Claim 5, Rahmes teaches:
5. The aviation anomaly detection system of claim 1 wherein the processor is further configured to determine aircraft locations from the text data, and determine the at least one aviation anomaly based upon relative positions of determined aircraft locations. [Rahmes: “An anomaly in the aviation data communications dataset would represent any unusual combination of this data. Some examples might include: an aircraft reporting a location not consistent with which ground station it is in contact with (e.g., aircraft reports it is over Alaska but in contact with a ground station in Guam); an unusual/erroneous latitude or longitude; an unusual bearing for this location on this flight path; unusual arrival airport for the given lat/lon/altitude (e.g., flying very low over NY but neither departure nor arrival are NY).” 1:56-67.]
Regarding Claim 7, Rahmes teaches:
7. The aviation anomaly detection system of claim 1 wherein the interface is configured to receive aircraft ground control audio communications and air route control audio communications. [Rahmes: “An anomaly in the aviation data communications dataset would represent any unusual combination of this data. Some examples might include: an aircraft reporting a location not consistent with which ground station it is in contact with (e.g., aircraft reports it is over Alaska but in contact with a ground station in Guam); an unusual/erroneous latitude or longitude; an unusual bearing for this location on this flight path; unusual arrival airport for the given lat/lon/altitude (e.g., flying very low over NY but neither departure nor arrival are NY).” 1:56-67.]
Claim 8 is a method claim with limitations corresponding to the limitations of Claim 1 and is rejected under similar rationale.
Claim 9 is a method claim with limitations corresponding to the limitations of Claim 2 and is rejected under similar rationale.
Claim 10 is a method claim with limitations corresponding to the limitations of Claim 3 and is rejected under similar rationale.
Claim 11 is a method claim with limitations corresponding to the limitations of Claim 4 and is rejected under similar rationale.
Claim 12 is a method claim with limitations corresponding to the limitations of Claim 5 and is rejected under similar rationale.
Claim 14 is a method claim with limitations corresponding to the limitations of Claim 7 and is rejected under similar rationale.
Claim 15 is a computer program product system claim with limitations corresponding to the limitations of method Claim 1 and is rejected under similar rationale.
Claim 16 is a computer program product system claim with limitations corresponding to the limitations of method Claim 2 and is rejected under similar rationale.
Claim 17 is a computer program product system claim with limitations corresponding to the limitations of method Claim 3 and is rejected under similar rationale.
Claim 18 is a computer program product system claim with limitations corresponding to the limitations of method Claim 4 and is rejected under similar rationale.
Claim 19 is a computer program product system claim with limitations corresponding to the limitations of method Claim 5 and is rejected under similar rationale.
Claims 6, 13, and 20 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Rahmes and Tkaczyk in view of Schuetz (U.S. 20250238707).
Regarding Claim 6, Rahmes does not teach the use of LLM to generate samples for the training of its VAE. Neither does Tkaczyk.
Schuetz teaches:
6. The aviation anomaly detection system of claim 1 wherein the VAE deep learning model is trained based upon a plurality of air traffic communications messages generated from a machine learning (ML) large language model (LLM). [Schuetz teaches that a “text classification machine learning model” that may be implemented as a VAE is trained by “expressions” / “messages” generated by an LLM. “[0155] … In the examples of FIG. 5 and FIG. 6, the digital assistant executes an intent classification model in the example form of a text classification machine learning model that was trained on a diverse range of sample expressions generated by an LLM (e.g., the LLM 118).” “[0202] Some of the techniques that may be used in generative AI are: … [0206] Variational autoencoders (VAEs): VAEs may encode input data into a latent space (e.g., a compressed representation) and then decode it back into output data. The latent space can be manipulated to generate new variations of the output data. VAEs may use self-attention mechanisms to process input data, allowing them to handle long text sequences and capture complex dependencies.”]
Rahmes/Tkaczyk and Schuetz pertain to the use of machine learning including VAEs and it would have been obvious to use the method of Schuetz that uses training samples generated by an LLM for the training of the VAE of the combination. This combination falls under combining prior art elements according to known methods to yield predictable results or use of known technique to improve similar devices (methods, or products) in the same way. See MPEP 2141, KSR, 550 U.S. at 418, 82 USPQ2d at 1396.
Claim 13 is a method claim with limitations corresponding to the limitations of Claim 6 and is rejected under similar rationale.
Claim 20 is a computer program product system claim with limitations corresponding to the limitations of method Claim 6 and is rejected under similar rationale.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 4-5, 7-8, 11-12, 14-15, and 18-19 are rejected under 35 U.S.C. 103 as being unpatentable over Tkaczyk (U.S. 20220383866) in view of Rangan (U.S. 20230356842) and further in view of Galvin (U.S. 20250378308).
Regarding Claim 1, Tkaczyk teaches:
1. An aviation anomaly detection system comprising:
an interface configured to receive audio communications between an air traffic control station and a plurality of aircraft; [Tkaczyk, Figure 1, “communications 102” between “aircraft 104” and “air traffic controllers 108.” “[0039] In embodiments, a processor in (e.g., a processor comprised within ATC communication system 100) may monitor ATC communications. In these embodiments, while monitoring, a processor may be configured to record ATC utterances, record timestamps associated with the ATC utterances, capture source data (e.g., if the source of the communication is ATC or non-ATC), and determine other communication information, such as the communication channel, relevant airports, and controller identity information….” ATC = Air Traffic Control. Figure 2, “monitor for ATC communication utterance 202.”]
a speech-to-text converter configured to convert the received audio communications from the interface to text data; and [Tkaczyk, Figure 1, “communications 102” leading to “transcripts …110.” “[0039] … In embodiments, a processor may convert some or all ATC utterances from speech to text using NLP service 112 to generate transcripts of some or all of the ATC utterances….” Figure 2, “speech-to-text 204.”]
a processor configured to determine at least one aviation anomaly from the text data with a variational autoencoder (VAE) deep learning model, and generate an alert based upon the at least one aviation anomaly. [Tkaczyk, Figure 1, “transcripts …110” generating the “alert 122” back to the “air traffic controllers 108” by use of the “natural language processing service 112.” Figure 2 shows the NLP service as e.g. Watson (see Figure 5 and [0065]). “[0041] In some embodiments, a processor may be configured to receive information regarding a particular aircraft (e.g., the aircraft of the pilot communicating with an air traffic controller).… In embodiments, ATC communication system 100 may be configured to use this information to detect anomalies in ATC communications and execute a response to mitigate or eliminate the negative outcomes potentially associated with the detected anomaly. In some embodiments a processor may execute a response that creates an alert 122 to air traffic controllers, aircraft, or other communication participants. Examples of such response alerts are provided in FIG. 13.” Figure 2, “Anomaly Detection 210” and “Response Execution 212” which is based on the “Alert Rules.”]
Tkaczyk does not teach the use of deep learning or machine learning and uses an NLP service 112
Rangan teaches:
a processor configured to determine at least one aviation anomaly from the text data with a variational autoencoder (VAE) deep learning model, and generate an alert based upon the at least one aviation anomaly. [Rangan teaches monitoring the communications of pilots in the cockpit (in-flight risk management IRM) and using a machine learning model to detect errors and anomalies and determine that the pilot is fatigued because he is making mistakes and provide alerts and warnings. See Figure 2, the pilots talking and the “Risk Management Application 144” monitoring them. “[0116] In some implementations, flight risk mitigation interventions provided in process 300 include providing a warning to the flight crew member….” “[0107] … Machine learning models used to process the measured fatigue indicator values to generate updated fatigue indicator profiles can include Bayesian optimization, deep neural networks, or Markov chain methods….” “[0053] In response to detecting that a standard protocol command litany has been initiated, the trained machine learning module analyzes, in real time, the subsequent audio data collected by the audio sensors 214 to determine whether the appropriate response(s) to the initial command have been provided by the flight crew members. In some implementations, if it is detected, based on real time analysis of the audio data, that a flight crew member delays in providing or fails to provide the appropriate response to the detected command, this can be indicative that the flight crew member is experiencing increased fatigue. In some implementations, the machine learning module analyzing the data provided by the audio sensors 214 is trained to differentiate between standard protocol commands and other conversations between the flight crew members. As such, the trained machine learning model can be used to detect and analyze whether a flight crew members accurately performs a standard protocol command litany. In some implementations, the IRM computing devices can be triggered to begin listening for a command litany based a pre-defined action associated with the command litany. Table 1 below provides examples of standard protocol commands and, in some cases, actions associated with a command litany.” “1…. continuously monitoring cockpit conversations; in response to detecting a command litany trigger: identifying, based on the command litany trigger, a particular command litany that a flight crew has initiated, and monitoring for an expected sequence of commands and responses associated with the particular command litany; determining a timing between responses and commands spoken during the particular command litany and an accuracy of the responses and commands spoken during the particular command litany; and generating, based on the timing and the accuracy measured fatigue indicators for each flight crew member involved particular command litany.”]
Tkaczyk and Rangan pertain to safety of aviation and generating warnings when an error/anomaly is detected based on speech of the pilots and others involved and it would have been obvious to substitute the more sophisticated machine learning system of Rangan for the NLP of Tkaczyk which is simpler to have a more up to date and sophisticated method of error detection. This combination falls under simple substitution of one known element for another to obtain predictable results or use of known technique to improve similar devices (methods, or products) in the same way. See MPEP 2141, KSR, 550 U.S. at 418, 82 USPQ2d at 1396.
Rangan uses machine learning models which are defined as “[0107] … Machine learning models used to process the measured fatigue indicator values to generate updated fatigue indicator profiles can include Bayesian optimization, deep neural networks, or Markov chain methods….” It does not mention a VAE expressly.
Galvin teaches:
a processor configured to determine at least one aviation anomaly from the text data with a variational autoencoder (VAE) deep learning model, and generate an alert based upon the at least one aviation anomaly. [Galvin is directed to anomaly detection and uses VAEs to do so. See Figures 6-7. “[0084] An error handling and anomaly detection subsystem 134 identifies and handles any errors, anomalies, or unexpected patterns in the generated output. It incorporates techniques for detecting and correcting syntactic or semantic errors, identifying out-of-distribution samples, or flagging potential issues that require human intervention. This subsystem plays a critical role in maintaining the quality and reliability of the generated output by proactively identifying and addressing any problems or inconsistencies. It helps to prevent the propagation of errors downstream and ensures that the generated output is trustworthy and dependable.” “[0069] … The machine learning training subsystem 600 is responsible for training the VAE decoder to accurately reconstruct or generate data from the latent space….” “[0110] Another form of deep learning coding includes variational autoencoders (VAEs). Variational autoencoders extend the concept of autoencoders by introducing a probabilistic framework. VAEs learn to encode the input data into a probability distribution in the latent space, rather than a single point….” “[0181] … The combination of the VAE encoder, latent transformer, and VAE decoder enables the system to handle a wide range of data types and perform various tasks, such as data compression, anomaly detection, sequence prediction, and data generation….”]
Tkaczyk/Rangan and Galvin pertain to detection of an error/anomaly and it would have been obvious to replace the machine learning of the combination with VAE of Galvin which is a type of machine learning and particularly adept at anomaly detection in the input data. This combination falls under simple substitution of one known element for another to obtain predictable results or use of known technique to improve similar devices (methods, or products) in the same way. See MPEP 2141, KSR, 550 U.S. at 418, 82 USPQ2d at 1396.
Regarding Claim 4, Tkaczyk teaches:
4. The aviation anomaly detection system of claim 1 wherein the at least one aviation anomaly comprises at least one of a pilot readback error and a pilot deviation error. [Tkaczyk includes an example of a readback error: “[0034] In one example, an air traffic controller could instruct a pilot to land an aircraft at a specific runway (e.g., Runway 9B) but instead of confirming the same runway, the pilot confirms the instruction to land at the wrong runway (e.g., Runway 9E). In this example, due to the similar sounding runway identifiers 9B and 9E, both the air traffic controller and the pilot might consider their communications to be correctly understood. While this is a simple miscommunication, if this potential anomaly is left uncorrected, it could result in an anomaly occurring that can have significant negative outcomes.”]
Regarding Claim 5, Tkaczyk teaches:
5. The aviation anomaly detection system of claim 1 wherein the processor is further configured to determine aircraft locations from the text data, and determine the at least one aviation anomaly based upon relative positions of determined aircraft locations. [Tkaczyk determines aircraft locations on the ground and in the air that is relevant to the ATC communications and Figure 1 shows this information as “static and dynamic ground context 114” and “static and dynamic airspace context 116” and are used with the “transcripts … 110” to detect an anomaly and generate the “alert 122.” Figure 2, “Ground and Airspace Context (static and dynamic)” is fed to the “aircraft/engine operating parameters” which is one of the factors going into the “anomaly detection 210” which is based on the ATC utterances obtained at 202. “[0040] In embodiments, the processor may be configured to receive static and dynamic ground context information 114 and static and dynamic airspace context information 116, relevant to the ATC communications and ATC utterances. … While FIG. 1 depicts static and dynamic ground context information 114 and static and dynamic airspace context information 116 as separate entities, hereinafter for simplicity such information will be referred to as static and dynamic data sources.”]
Regarding Claim 7, Tkaczyk teaches:
7. The aviation anomaly detection system of claim 1 wherein the interface is configured to receive aircraft ground control audio communications and air route control audio communications. [Tkaczyk, Figure 2, “ground and airspace context (static and dynamic)” is being provided to the “obtain aircraft and engine operating parameters 208.” “[0033] … Due to the nature of directing multiple pieces of large equipment, even simple mistakes in the aircraft industry can be disastrous. Despite past events, anomalies (e.g., mistakes) in commands exchanged between the air traffic controllers, ground controllers, aircrafts, ground personnel, and ground vehicles, continue to contribute to forced pilot diversions, vehicle and personnel incursions, and operational incidents.”]
Claim 8 is a method claim with limitations corresponding to the limitations of Claim 1 and is rejected under similar rationale.
Claim 11 is a method claim with limitations corresponding to the limitations of Claim 4 and is rejected under similar rationale.
Claim 12 is a method claim with limitations corresponding to the limitations of Claim 5 and is rejected under similar rationale.
Claim 14 is a method claim with limitations corresponding to the limitations of Claim 7 and is rejected under similar rationale.
Claim 15 is a computer program product system claim with limitations corresponding to the limitations of method Claim 1 and is rejected under similar rationale.
Claim 18 is a computer program product system claim with limitations corresponding to the limitations of method Claim 4 and is rejected under similar rationale.
Claim 19 is a computer program product system claim with limitations corresponding to the limitations of method Claim 5 and is rejected under similar rationale.
Claims 2, 9, and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Tkaczyk, Rangan, Galvin in view of Lee (U.S. 20210110158).
Regarding Claim 2, Tkaczyk and Rangon were not cited for teaching the VAE and Galvin teaches:
2. The aviation anomaly detection system of claim 1 wherein the VAE deep learning model comprises a plurality of VAE deep learning models including at least some of an Adaptive Moment Estimation (ADAM) deep learning VAE model, a Stochastic Gradient Descent with Momentum (SGDM) deep learning VAE model, and a root mean square propagation (RMSProp) deep learning VAE model. [Galvin: “[0169] … Model parameters and hyperparameters can include, but are not limited to, bias, train-test split ratio, learning rate in optimization algorithms (e.g., gradient descent), choice of optimization algorithm (e.g., gradient descent, stochastic gradient descent, of Adam optimizer, etc.), …” ADAM is an equivalent of gradient descent method used in neural networks.]
Rationale as provided for Claim 1 because VAE was combined from Galvin and ADAM is an algorithm used in VAE.
Galvin teaches one type of VAE and the Claim asks for a plurality.
Lee teaches:
2. The aviation anomaly detection system of claim 1 wherein the VAE deep learning model comprises a plurality of VAE deep learning models including at least some of an Adaptive Moment Estimation (ADAM) deep learning VAE model, a Stochastic Gradient Descent with Momentum (SGDM) deep learning VAE model, and a root mean square propagation (RMSProp) deep learning VAE model. [Lee teaches the use of VAE and teaches the different optimization methods for the VAE as listed in the Claim. “[0017] The method may further include generating an image of a hidden product by applying a second machine learning model based on a generative model including any one of a generative adversarial network (GAN), a conditional GAN (cGAN), a deep convolution GAN (DCGAN), an auto-encoder, or a variational auto-encoder (VAE) to the captured image, and acquiring product information from the generated product when a product is hidden by another product or a surrounding environment.” “[0134] In machine learning or deep learning, learning optimization algorithms may be deployed to minimize a cost function, and examples of such learning optimization algorithms include gradient descent (GD), stochastic gradient descent (SGD), momentum, Nesterov accelerate gradient (NAG), Adagrad, AdaDelta, RMSProp, Adam, and Nadam.”]
Tkaczyk/Rangan/Galvin and Lee pertain to the use of machine learning including VAE models and it would have been obvious to use the methods listed in Lee which are equivalents of ADAM as different alternative with the system of combination to provide for more than one optimization method. This combination falls under combining prior art elements according to known methods to yield predictable results or use of known technique to improve similar devices (methods, or products) in the same way. See MPEP 2141, KSR, 550 U.S. at 418, 82 USPQ2d at 1396.
Claim 9 is a method claim with limitations corresponding to the limitations of Claim 2 and is rejected under similar rationale.
Claim 16 is a computer program product system claim with limitations corresponding to the limitations of method Claim 2 and is rejected under similar rationale.
Claims 3, 10, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Tkaczyk, Rangan, Galvin in view of Denli (U.S. 20200183047) and Mount (U.S. 20080103872).
Regarding Claim 3, Tkaczyk and Rangon were not cited for teaching the VAE and Galvin does not mention game theory.
Denli teaches:
3. The aviation anomaly detection system of claim 2 wherein the processor is configured to select a given VAE deep learning model from among the plurality thereof based upon a game theory reward matrix. [Denli teaches that VAE or VAE-combined GAN models use game theory to generate an adversarial mechanism. “[0078] …The noise distribution may also be learned as a prior distribution using a machine learning process such as a decoder … or an autoencoder, or a variational autoencoder (VAE) or VAE-combined GAN(VAEGAN) model.” “[0080] … This competition between G and D networks may converge at a local Nash equilibrium of Game Theory … and generative model G learns mapping from noise and input x providing conditions to output y, G: (x, z) [Wingdings font/0xE8]y. …”]
Tkaczyk/Rangan/Galvin and Denli pertain to the use of machine learning including VAEs and it would have been obvious to use the VAEGAN model of Denli which uses game theory in the GAN portion with the system of combination as one alternative. This combination falls under simple substitution of one known element for another to obtain predictable results or use of known technique to improve similar devices (methods, or products) in the same way. See MPEP 2141, KSR, 550 U.S. at 418, 82 USPQ2d at 1396.
Game theory includes the use of reward matrices but Denli does not state that.
Mount teaches:
wherein the processor is configured to select a given VAE deep learning model from among the plurality thereof based upon a game theory reward matrix. [Mount teaches that game theory includes the use of punishment and reward matrices: “[0078] … It is contemplated that the resulting effects of the new introduction on competitors within an industry on both the sales volume measure and exposure to competition measure can be used as method of measuring punishment and reward matrices in multiplayer games as described in numerous game theory applications….”]
Tkaczyk/Rangan/Galvin/Denli and Mount pertain to the use of machine learning including models that use game theory and would have been obvious to combine Mount with the system of combination to show that game theory includes the use of reward matrices. This combination falls under combining prior art elements according to known methods to yield predictable results or use of known technique to improve similar devices (methods, or products) in the same way. See MPEP 2141, KSR, 550 U.S. at 418, 82 USPQ2d at 1396.
Claim 10 is a method claim with limitations corresponding to the limitations of Claim 3 and is rejected under similar rationale.
Claim 17 is a computer program product system claim with limitations corresponding to the limitations of method Claim 3 and is rejected under similar rationale.
Claims 6, 13, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Tkaczyk, Rangan, Galvin in view of Schuetz (U.S. 20250238707).
Regarding Claim 3, Tkaczyk and Rangon were not cited for teaching the VAE and Galvin does not teach generating the training sample of the VAE by an LLM.
Schuetz teaches:
6. The aviation anomaly detection system of claim 1 wherein the VAE deep learning model is trained based upon a plurality of air traffic communications messages generated from a machine learning (ML) large language model (LLM). [Schuetz teaches that a “text classification machine learning model” that may be implemented as a VAE is trained by “expressions” / “messages” generated by an LLM. “[0155] … In the examples of FIG. 5 and FIG. 6, the digital assistant executes an intent classification model in the example form of a text classification machine learning model that was trained on a diverse range of sample expressions generated by an LLM (e.g., the LLM 118).” “[0202] Some of the techniques that may be used in generative AI are: … [0206] Variational autoencoders (VAEs): VAEs may encode input data into a latent space (e.g., a compressed representation) and then decode it back into output data. The latent space can be manipulated to generate new variations of the output data. VAEs may use self-attention mechanisms to process input data, allowing them to handle long text sequences and capture complex dependencies.”]
Tkaczyk/Rangan/Galvin and Schuetz pertain to the use of machine learning including VAEs and it would have been obvious to use the method of Schuetz that uses training samples generated by an LLM for the training of the VAE of the combination. This combination falls under combining prior art elements according to known methods to yield predictable results or use of known technique to improve similar devices (methods, or products) in the same way. See MPEP 2141, KSR, 550 U.S. at 418, 82 USPQ2d at 1396.
Claim 13 is a method claim with limitations corresponding to the limitations of Claim 6 and is rejected under similar rationale.
Claim 20 is a computer program product system claim with limitations corresponding to the limitations of method Claim 6 and is rejected under similar rationale.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Janjos (U.S. 20240095595) is directed to the training of VAE for detection of anomaly in patterns. “[0005] Variational autoencoders (VAE) are used as backbones in solving a plurality of technical problems. For example, a VAE may be used for detecting anomalies in sensor measurements. …” “[0008] … a computer-implemented method for training a machine learning system, wherein the machine learning system is configured for anomaly detection and/or sampling a trajectory for a traffic participant and/or sampling of sensor signals and/or for determining a value characterizing a likelihood of an input signal with respect to a training dataset. …” “[0016] … The value characterizing a likelihood may also be used for anomaly detection, e.g., by determining the input signal as anomalous if the value characterizing the likelihood falls below a predefined threshold and determining the input signal to be normal otherwise.” [0065] The modification unit (180) determines the new parameters (Φ′) based on the first loss value. In the given embodiment, this is done using a gradient descent method, preferably stochastic gradient descent, Adam, or AdamW. In further embodiments, training may also be based on an evolutionary algorithm or a second-order method for training neural networks.”
Any inquiry concerning this communication or earlier communications from the examiner should be directed to FARIBA SIRJANI whose telephone number is (571)270-1499. The examiner can normally be reached on 9 to 5, M-F.
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, Pierre Desir can be reached on 571-272-7799. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/Fariba Sirjani/
Primary Examiner, Art Unit 2659
Prosecution Insights