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 .
Status of the Claims
Claims 1-18 were previously pending and subject to a non-final office action mailed 12/17/2024. Claims 1 and 10 were amended; claims 9 and 18 were cancelled, and no claim was added in a reply filed 05/19/2025. Therefore claims 1-8 and 10-17 are currently pending and subject to the final office action below.
Response to Arguments
Applicant's arguments filed 05/19/2025 in regards to section 101 rejection have been fully considered but they are not persuasive.
Applicant argues that the claims are not directed to methods of organizing human activity and mental processes because the claims do not fall within the definitions of each of the enumerated sub groupings of fundamental economic principles or practices, commercial or legal interactions, managing personal behavior, and relationships or interactions between people as provided by MPEP 2106.04(a)(2). Applicant further argues that the claims are rather directed at a technological improvement in freight and supply chain management by integrating large language models with data preprocessing to enhance accuracy in maritime shipping predictions.
Claim 1 addresses a technical problem rather than a business or logistical concern. Freight management systems can be complex due to the number of variables involved in just one shipment. Slight changes in one stage of shipment can result in costly delays in downstream shipments and transits and affect all parties involved. Claim 1, as amended, recites many features that are directed towards improving freight management and maritime shipping using a novel approach that integrates large language models into the management process. Specifically, claim 1 processes historical shipment data in a manner that large language models can understand, and using large language models trained to accurately predict shipment progress date sequences for new shipments based on how various factors affect the progress of historical shipments. By “training a large language model” using “generated shipping profiles” where “each of the plurality of shipping profiles indicates an association between each of the plurality of categories of shipment data of a historical shipment and each date of the shipment progress date sequence of the historical shipment,” claim 1 provides that large language models can be utilized to predict a “shipment progress date sequence” that “indicates a sequence of forecasted dates on which a series of maritime shipment progression milestones will occur.”
Incorporating large language models for predicting shipment progress date sequence is not currently present in maritime freight management and supply chain logistics. Large language models are typically not able to be used in this context due to their inability to understand the shipping industry as well as the meaning of events taking place in a shipment. Many current methods utilize other types of models that may be more computationally intensive to train and less accessible than large language models. By standardizing shipment data and generating shipping profiles in a manner that large language models can understand, claim 1 can improve upon current freight management methods by providing accurate predictions of shipment progress date sequence using large language models. Therefore, Applicant respectfully submits that claim 1 as amended provides more than just “certain methods of organizing human activity…Further, Applicant submits that claim 1 recites limitations such as “generating, a plurality of shipping profiles…,” and “training a large language model using the plurality of generated shipping profiles…” that do not fall into the category of mental processes that can be performed by a human” (remarks p. 3-5). Examiner respectfully disagrees.
Examiner respectfully argues that the claims are not directed towards an improvement in technology but rather an abstract idea that is performed in a computer environment in order to automate the process and any increase in efficiency is due to the computers themselves and not the claim limitations. A machine learning model is a mathematical model that is performed in a computer in order to automate the process. The fact that a machine learning model itself is used does not provide an improvement to a technical field. The machine learning model needs to provide an improvement to the computer functionality or to a technical field. However, using a generic machine learning model to perform functions that are usually performed does not provide an improvement but simply applies an abstract process in a computer in order to take advantage of the inherent efficiency of computer to perform the calculations faster. Training a large language model on data is a generic function of an LLM, the training itself on new data does not improve the LLM model, the computer or a technical field, it simply takes advantage of a generic function of the LLM in order to automate the process. Therefore, the claims are not directed towards an improvement and not patent eligible.
Applicant further argues “similar to the Synopsys. example, claim 1 does not recite a mental process because the human mind is not equipped to perform such a process. Claim 1 utilizes shipment data from a plurality of historical shipments, and generates shipping profiles based on the shipment data where it determines the association between each piece of shipment data for a shipment and each date of the shipment progress date sequence of the historical shipment. The generated shipping profiles are then used to train large language models in a manner that enables the large language models to capture statistical correlations between specific pieces of data in shipping profiles and date sequences. Claim 1 is similar to the Synopsis. example in that claim 1 also recites a multiple-step manipulation of data for a specific training scheme that trains large language models to generate shipment progress predictions. The relevant computation resources necessary to perform these steps are not things that can be performed simply by the human mind, which makes claim 1 patent eligible.“ (remarks p. 6). Examiner respectfully disagrees.
Examiner respectfully argues that the current claims are actually analogous to Synopsys and not patent eligible. In Synopsys, the claims dealt with translating functional description into hardware component description. The current claims deal with translating historical shipping data into predictive shipping profiles. Both patents fundamentally involve taking one form of data representation and converting it into another representation and converting it into another representation. This is the same abstract concept of data transformation. In Synopsys, the circuit translation could be done “mentally or by pencil and paper”, the current claims a shipping expert could conceptually review historical shipping patterns, identify trends and associations, make predictions about the future shipments and adjust forecasts based on deviations. While the scale differs, the fundamental mental process of pattern recognition and prediction from historical data is the same. In Synopsys, claims didn’t improve computer technology itself, similarly, the current claims do not improve how computers process data or enhance computer functionality, they merely use computers as tools to perform conventional data analysis and prediction. Essentially, both use computers to automate what humans traditionally did, without improving the computer’s capabilities. Therefore, like Synopsys, the current claims are not patent eligible because they do not provide an improvement to computers or technical field.
Applicant argues that Claim 1 as amended provides for the use of a specific technique to improve the functioning of large language models to solve a technological problem arising in the field of maritime freight management because:
Systems and methods in accordance with embodiments of the invention can remedy the above issues by utilizing a data processing pipeline that trains an LLM using historical shipping data that is preprocessed to impose structure to the data that allows the LLM to learn patterns that enable the forecasting of the date on which a series of shipment progression milestones will occur (e.g. departure dates, and/or arrival dates).
Under MPEP § 2106.04(d)(1), an invention is not abstract if it is integrated into a practical application. The present claims satisfy this requirement because they use real historical shipment data and process them in a manner that can be understood by large language models, such that large language models can be trained using these historical shipment data to predict shipment progress date sequences. The claimed invention is not merely an abstract idea performed on a generic computer but instead relies on specialized data integration and large language models to solve a real-world problem in the maritime shipping industry. Additionally, MPEP § 2106.04(d) states that an invention that improves a technological field is not abstract. The claimed invention clearly falls into this category as it enhances shipment forecasting accuracy by using large language models which are uncommon in the shipment industry, providing a solution beyond mere automation of a human process (remarks p. 8-9). Examiner respectfully disagrees.
Despite applicant’s arguments the core claim is still directed to the abstract idea of data pattern recognition and prediction. The same type of mental process that humans have performed for centuries in logistics. Courts have consistently held that merely applying abstract ideas to a specific field and on a computer doesn’t make them patent eligible. Similarly, applying pattern recognition to “maritime freight management” is just applying an abstract concept to a particular industry.
Although the current claims describe training a machine learning model, the ML model is a standard machine learning model that makes predictions. Training neural networks on datasets is well-known, conventional computer activity.
Furthermore, Applicant’s reliance on MPEP 2106.04(d) is misapplied and misplaced. The current claims do not improve how computers function or process data, using LLMs for pattern recognition is their intended, conventional application and this is automation of existing business processes and not technological advancement. Unlike patent eligible computer improvements, the current claim lacks novel data structures, improved algorithms, enhanced computer performance, new computational techniques or specific technological solutions beyond applying known ML to shipping.
Therefore, the current claims are directed towards an abstract idea that is not integrated into a practical application.
Applicant argues that “that claim 1 provides non-routine or unconventional features that are not well understood in the art. Specifically, claim 1 recites “generating a plurality of shipping profiles…,” and “training a large language model using the plurality of generated shipping profiles…” Claim 1 is not similar to existing technologies, which do not provide a way to train a large language model such that it can be used to generate “shipment progress date sequence” predictions. Claim 1 processes data from “historical shipments” to generate “shipping profiles” in a manner that can be used to train large language models. According to MPEP § 2106.05(d), a claim includes an inventive concept if it applies a specific technical improvement beyond what is well understood, routine, or conventional in the field. Large language models may be more accessible and easier to use, but they have yet to be implemented in the shipment industry for freight management and shipment predictions. Thus, unlike conventional shipment prediction systems, claim 1 provides a specific technical improvement beyond what is well understood, routine, or conventional in the field. “(remarks p. 9-10). Applicant respectfully disagrees.
First, Applicant’s assertion that training LLMs for shipping predictions is “unconventional” ignores certain realities such as training a machine learning model on any data is a generic function of LLMs, predictive analytics for supply chain management has been commercial deployed for years and time series forecasting using historical data is fundamental to operation research.
Despite claims of “specific technical improvement”, the patent merely describes standard data preprocessing, conventional ML training, routine pattern recognition and basic prediction algorithm. None of these improve how computers functions, they just use computers conventionally.
Furthermore, the core innovation of generating “shipping profiles” that associate shipment categories with data is simply creating data structures that represent relationships. This is abstract information organization regardless of the domain which is a mental process. Adding “maritime shipping” context doesn’t transform this mental process into patent eligible subject matter.
Applicant’s claim that this doesn’t “monopolize all training methods” misses the point that section 101 does not require complete monopolization to render claims ineligible. The claims still cover fundamental data analysis and prediction processes and alternative implementations do not cure abstract idea problems if the core concept remains abstract.
Therefore, the claims do not provide an inventive concept because the claimed “data processing pipeline” represents a collection of conventional systems such as routine integration of historical data collection, data standardization, pattern recognition and prediction generation which alone or in combination to not improve the computer or technical field. Therefore, the claims are not patent eligible.
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 1-8 and 10-17 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.
Claim 1 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim recites “receiving input data comprising a plurality of categories of shipment data for a plurality of historical shipments, wherein the shipment data comprises a shipment progress date sequence for each of the plurality of historical shipments; standardizing each of the plurality of categories of shipment data for the plurality of historical shipments; generating a plurality of shipping profiles based on the standardized shipment data, wherein each of the plurality of shipping profiles indicates an association between each of the plurality of categories of shipment data of a historical shipment and each date of the shipment progress date sequence of the historical shipment; and wherein the trained large language model is capable of generating a new shipping profile, wherein the new shipping profile comprises a predicted shipment progress date sequence associated with an upcoming shipment, wherein the predicted shipment progress date sequence indicates a sequence of forecasted dates on which a series of maritime shipment progression milestones will occur, wherein the trained large language model is capable of adjusting the predicted shipment progress date sequence when an actual shipment progress deviates from the predicted shipment progress date sequence.”
The limitations above are a process that, under its broadest reasonable interpretation, covers a method of predicting shipping progress which is a method of organizing a human activity, mental processes and mathematical concepts. That is, the method allows for commercial interactions (including sales activities or behaviors; business relations) and managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions) and allows for concepts that can be performed in the human mind and ones that contain mathematical relationships.
This judicial exception is not integrated into a practical application. In particular, the claim recites “training a large language model using the plurality of generated shipping profiles”. This additional element is recited at a high level of generality and amounts to apply it instructions. Accordingly, this additional element, alone or in combination, does not integrate the abstract idea into a practical application. The claim is directed to an abstract idea.
The claim does not include additional elements, alone or in combination, that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element does not integrate the abstract idea into a practical application or provide significantly more limitations.
Dependent claim 2 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 1 without successfully integrating the exception into a practical application or providing significantly more limitations.
Dependent claim 3 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 1 without successfully integrating the exception into a practical application or providing significantly more limitations.
Dependent claim 4 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 1 without successfully integrating the exception into a practical application or providing significantly more limitations.
Dependent claim 5 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 1 without successfully integrating the exception into a practical application (enterprise resource planning (ERP) systems and freight management software is recited at a high level of generality and amounts to apply it instructions) or providing significantly more limitations.
Dependent claim 6 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 1 without successfully integrating the exception into a practical application or providing significantly more limitations.
Dependent claim 7 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 1 without successfully integrating the exception into a practical application or providing significantly more limitations.
Dependent claim 8 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 1 without successfully integrating the exception into a practical application or providing significantly more limitations.
Claim 10 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim recites “receiving input data comprising a plurality of categories of shipment data for a plurality of historical shipments, wherein the shipment data comprises a shipment progress date sequence for each of the plurality of historical shipments; standardizing each of the plurality of categories of shipment data for the plurality of historical shipments; generating a plurality of shipping profiles based on the standardized shipment data, wherein each of the plurality of shipping profiles indicates an association between each of the plurality of categories of shipment data of a historical shipment and each date of the shipment progress date sequence of the historical shipment; and wherein the trained large language model is capable of generating a new shipping profile, wherein the new shipping profile comprises a predicted shipment progress date sequence associated with an upcoming shipment, wherein the predicted shipment progress date sequence indicates a sequence of forecasted dates on which a series of maritime shipment progression milestones will occur, wherein the trained large language model is capable of adjusting the predicted shipment progress date sequence when an actual shipment progress deviates from the predicted shipment progress date sequence.”
The limitations above are a process that, under its broadest reasonable interpretation, covers a method of predicting shipping progress which is a method of organizing a human activity, mental processes and mathematical concepts. That is, the method allows for commercial interactions (including sales activities or behaviors; business relations) and managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions) and allows for concepts that can be performed in the human mind and ones that contain mathematical relationships.
This judicial exception is not integrated into a practical application. In particular, the claim recites “training a large language model using the plurality of generated shipping profiles”, processor and non-transitory machine readable medium. This additional element is recited at a high level of generality and amounts to apply it instructions. Accordingly, this additional element, alone or in combination, does not integrate the abstract idea into a practical application. The claim is directed to an abstract idea.
The claim does not include additional elements, alone or in combination, that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element does not integrate the abstract idea into a practical application or provide significantly more limitations.
Dependent claim 11 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 10 without successfully integrating the exception into a practical application or providing significantly more limitations.
Dependent claim 12 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 10 without successfully integrating the exception into a practical application or providing significantly more limitations.
Dependent claim 13 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 10 without successfully integrating the exception into a practical application or providing significantly more limitations.
Dependent claim 14 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 10 without successfully integrating the exception into a practical application (enterprise resource planning (ERP) systems, freight management software is recited at a high level of generality and amounts to apply it instructions) or providing significantly more limitations.
Dependent claim 15 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 10 without successfully integrating the exception into a practical application or providing significantly more limitations.
Dependent claim 16 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 10 without successfully integrating the exception into a practical application or providing significantly more limitations.
Dependent claim 17 is also directed to an abstract idea without significantly more because it further narrows the abstract idea described in relation to claim 10 without successfully integrating the exception into a practical application or providing significantly more limitations.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claim(s) 1-3 and 10-12 is/are rejected under 35 U.S.C. 103 as being unpatentable over Atwood (US 11501245) in view of Xue, “PROMPTCAST: A NEW PROMPT-BASED LEARNING PARADIGM FOR TIME SERIES FORECASTING”, published by openreview.net on Sept 22, 2022, hereinafter Xue.
As per claim 1/10, Atwood discloses a method for training a ML model to generate accurate shipping forecasts comprising:
receiving input data comprising a plurality of categories of shipment data for a plurality of historical shipments, wherein the shipment data comprises a shipment progress date sequence for each of the plurality of historical shipments (4:21-34, “(93) The training data 162 can include, for example, historical data that indicates the historical outcomes of various previous shipments (e.g., which may take the form of shipping logs). In some implementations, the training data 162 can include a plurality of training example pairs, where each training example pair provides: (1) a set of data (e.g., incorrect and/or incomplete data); and (2) a ground truth label associated with such set of data, where the ground truth label provides a “correct” prediction for the set of data. For example, the training example can include: (1) itinerary and/or vehicle event data that was collected prior or during a shipment; and (2) a ground truth label derived from historical data that indicates what the actual itinerary or events were for such shipment.”, 19:59-67, “As another example, a training example pair can include: (1) data that describes an original itinerary or event set and an updated itinerary or event set (e.g., including predicted new arrival or departure events, predicted new discharge dates, etc.); and (2) automatic supply chain adjustments (e.g., modification of requested delivery dates or labor fulfillment dates) that should be performed based on change from the original itinerary to the updated itinerary.”)
generating a plurality of shipping profiles based on the standardized shipment data, wherein each of the plurality of shipping profiles indicates an association between each of the plurality of categories of shipment data of a historical shipment and each date of the shipment progress date sequence of the historical shipment (19:35-65: “(94) As one example, a training example pair can include: (1) vehicle tracking entries; and (2) ground truth labels for whether each vehicle tracking entry corresponds to a stopped vehicle or a vehicle in transit. As another example, a training example pair can include: (1) vehicle tracking entries (e.g., filtered entries); and (2) ground truth labels for whether each vehicle tracking entry corresponds to a vehicle at a transportation location or a vehicle at a waiting area. As another example, a training example pair can include: (1) vehicle tracking entries (e.g., filtered entries); and (2) ground truth clusters that correspond to transportation locations at which the vehicles are stationed, docked, parked, etc. As another example, a training example pair can include: (1) vehicle tracking entries (e.g., filtered entries); and (2) ground truth geofences that correspond to transportation locations at which the vehicles are stationed, docked, parked, etc. As another example, a training example pair can include: (1) vehicle tracking entries (e.g., filtered entries); and (2) ground truth labels that correspond to whether each vehicle tracking entry correspond to a newly achieved milestone. As yet another example, a training example pair can include: (1) an original set of itinerary or event data and changes to such itinerary or event data (e.g., addition of milestone events); and (2) additional updates to the itinerary (e.g., predicted new discharge dates, etc.). As another example, a training example pair can include: (1) data that describes an original itinerary or event set and an updated itinerary or event set (e.g., including predicted new arrival or departure events, predicted new discharge dates, etc.); and (2) automatic supply chain adjustments (e.g., modification of requested delivery dates or labor fulfillment dates) that should be performed based on change from the original itinerary to the updated itinerary”, the itineraries that associate shipment categories 9arrival, departure, discharge) with specific milestone dates); and
training a ML model using the plurality of generated shipping profiles, wherein the trained ML model is capable of generating a new shipping profile, wherein the new shipping profile comprises a predicted shipment progress date sequence associated with an upcoming shipment, wherein the predicted shipment progress date sequence indicates a sequence of forecasted dates on which a series of maritime shipment progression milestones will occur (19:8-20, “91) In some implementations, the model trainer 160 can perform supervised training techniques using a set of labeled training data. In other implementations, the model trainer 160 can perform unsupervised training techniques using a set of unlabeled training data. The model trainer 160 can perform a number of generalization techniques to improve the generalization capability of the models being trained. Generalization techniques include weight decays, dropouts, or other techniques. (92) In particular, the model trainer 160 can train a machine-learned model 110 and/or 140 based on a set of training data 162. The model trainer 160 can be implemented in hardware, software, firmware, or combinations thereof.”, 26:23-29, “(131) Based on the set of data 602, the machine-learned model 110 can produce a model prediction 606. As examples, the model prediction 606 can include a prediction of the ground truth label 604. Thus, as one example, if the ground truth label 604 provides one or more actual milestone location(s) and/or date(s), then the model prediction 606 can correspond to predicted milestone location(s) and/or date(s).” the system generates predictions of milestone dates, which constitute a new shipping profile with forecasted progress date sequence.”, 26:55-61, “36) In response to the set of data 702, the machine-learned model 110 can produce a model prediction 706. The model prediction 706 can be of any of the different types of data discussed with respect to 604 or 606 of FIG. 6, or other forms of data. As one example, the model prediction 706 can include one or more predicted milestone location(s) and/or date(s).”)
wherein the trained large language model is capable of adjusting the predicted shipment progress date sequence when an actual shipment progress deviates from the predicted shipment progress date sequence (16:11-16, “(74) The supply chain adjustment system 106 can perform automatic supply chain adjustments based on the imputed milestones. As examples, the supply chain adjustment system 106 can proactively detect, share, and resolve shipment issues.”, 5:48-64, “(14) Furthermore, in some implementations, the supply chain management computing system can also predict or otherwise update additional information regarding the item of cargo based on the imputed shipment milestone(s), including, as examples, a current location of the item of cargo, a predicted date of discharge for the item of cargo at the destination, predicted identities of the shipment vehicle(s), and/or the like. As such, the imputed shipment milestone(s) can be used to predict and improve the accuracy of data that is downstream or otherwise reliant upon the accuracy of the shipment event data. Such improved downstream data can then be used to make adjustments to various aspects of the supply chain as a whole (e.g., in an automated and/or intelligent fashion). For example, other alerts and tasks can be automatically performed as well, such as, for example, automated billing operations based on the imputed shipment milestone(s).”).
However, Atwood does not disclose but Xue discloses standardizing each of the plurality of categories of shipment data for the plurality of historical shipments (abstract, “the numerical input and output are transformed into prompts”, page 1, “Numerical forecasting methods always take numerical values as input an generate numerical values as the prediction for the next time step. Instead, the input and output of the proposed prompt-based forecasting (Figure 1 (b)) are natural language sentences. This paradigm change enables the utilization of language generation models for forecasting”, page 3, “This data-to-text transformation is referred as a prompting process in this work (the details of prompting are presented in the next section). Specifically, the input numerical sequence xm t1: tobs is turned into input prompts and the forecasting target value xmtobs+1 is transformed as the output prompt. Consequently, the time series forecasting can be addressed through a natural language generation paradigm, and language foundation models can be adopted as the core forecasting models in PromptCast task…In this section, we demonstrate the design and construction of the proposed PISA dataset. The overall designing guideline is: (1) to preprocess original data given in the numerical format (raw data) for the forecasting task setting (Sec. 3.1) and (2) to transform the numerical data to natural language input/output formats with prompts (Sec. 3.2). We also describe the features and statistics (Sec. 3.3)”) and training a large language model using the plurality of generated shipping profiles (page 3, 7-8).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of the invention to include the limitations above as taught by Xue in the teaching of Atwood, in order to e to directly apply language models for forecasting purposes (please see Xue abstract).
As per claim 2/11, Atwood discloses wherein the plurality of categories of shipment data comprises at least one source of shipment data selected from the group consisting of: freight logistic data, importer supply chain data, and external factor data (10:1-10, “(98) As one example, vehicle tracking data can include automatic identification system (AIS) data. The AIS is an automatic tracking system that uses transponders on ships to identify the current locations (e.g., with some margin of error) of the ships. Information provided by AIS equipment, such as unique identification, position, course, and speed, can be obtained by the carrier computing systems 60 and/or a third-party system (not shown) and supplied (e.g., via an API) to the supply chain management computing system 102.”).
As per claim 3/12, Atwood discloses wherein the freight logistic data comprises at least one source of freight logistic data selected from the group consisting of: vessel traffic data, vessel schedules, ocean carrier service itineraries, lists of actively registered International Maritime Organization (IMO) vessels, and ocean container vessel data (10:1-10, “(98) As one example, vehicle tracking data can include automatic identification system (AIS) data. The AIS is an automatic tracking system that uses transponders on ships to identify the current locations (e.g., with some margin of error) of the ships. Information provided by AIS equipment, such as unique identification, position, course, and speed, can be obtained by the carrier computing systems 60 and/or a third-party system (not shown) and supplied (e.g., via an API) to the supply chain management computing system 102.”).
Claim(s) 4-6 and 13-15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Atwood (US 11501245) in view of Xue, as disclosed in the rejection of claim 2, in further view of McAlister (US 20230259874).
As per claim 4/13, Atwood does not disclose but McAlister discloses wherein the external factor data comprises at least one source of external factor data selected from the group consisting of: inclement weather conditions, port congestion conditions, and current market conditions (paragraph 40, 83, “[0040] In variants, transit data can be augmented with additional data (e.g., from one or more third party data sources). This additional data can include weather data, traffic data, news data, social media data, holiday schedule, and/or any other relevant data.”)
Therefore, it would have been obvious to one of ordinary skill in the art at the time of the invention to include the limitations above as taught by McAlister in the teaching of Atwood, in order to determine, select, and/or train one or more models to predict package transit (McAlister, abstract)
As per claim 5/14, Atwood does not disclose but McAlister discloses wherein the importer supply chain data comprises at least one source of importer supply chain data selected from the group consisting of: Packing Lists (PL), enterprise resource planning (ERP) systems, freight management software, manual data entries, and custom-house data (paragraph 37, ““[0037] Transit data for one or more packages (e.g., a single package, each package in a set, an aggregated set of packages, a shipment, etc.) can include shipment data, tracking data, transit time (e.g., number of days in transit, number of business days in transit, time in transit, arrival time, etc.), transit predictions (e.g., where the prediction can be for any transit data), and/or any other data. Shipment data can include: delivery data, package shipment creation date, shipment receipt date (e.g., date the package was received at the shipping facility), shipment delivery date (e.g., including whether the package is still in transit), a shipping carrier identifier, the carrier facilities, throughput and/or processing times of shipping carrier facilities and/or fulfillment facilities (e.g., including delay estimates, actual processing times, etc.), carrier and/or facility statistics (e.g., the average, median, variance, or other statistical measure of transit time, or residency time, etc.), shipper identifier, recipient identifier, contextual parameters (e.g., day of the week, day of the year, week of the year, month, weather, logistic parameters, dock throughput, shipping carrier information, etc.), shipping carrier, shipping carrier service, shipping lane, shipping lane direction, shipping origin (e.g., location, region, address, zip code, etc.), shipping destination, zone, route (e.g., including origin, destination, one or more legs, a set of geolocations along the route path, etc.), package size and/or dimensions, package weight, package value, package insurance value, and/or any other package feature and/or shipment parameter. Tracking data (e.g., tracking details) can include and/or be determined based on: a shipping status (e.g., in transit, delivered, picked up, received, etc.); a message from the carrier (e.g., containing scan event information); a location (e.g., city, state, country, and zip information about the location); a shipping carrier identifier; a facility identifier (e.g., carrier facility, fulfillment facility, etc.), a service type; a container type; an estimated delivery date; an estimated delivery time; a date and/or time a shipment was picked up for delivery; a date and/or time a shipment was delivered to a final destination; a date and/or time a shipment arrived at an intermediate facility and/or delivery vehicle; a date and/or time a shipment departed an intermediate facility and/or delivery vehicle; shipment handoff information (e.g., confirmations); anomalous events (e.g., received from the carrier or a third-party data source), origin address information; destination address information; a shipment identifier; scan event information (e.g., description, location, time, source, scan code, barcode, message associated with a scan result of one or more scan events, associated with a shipping label printed for the package, etc.); a total time in transit (e.g., duration between creation and delivery, duration between departure from a facility and delivery, etc.); sequences of shipment patterns in historical transit data (e.g., movement of a shipment between warehouses, fulfillment centers, carriers, delivery destination, etc.); and/or any other suitable data (e.g., metadata associated with a shipment). The transit time can include: a total shipment transit time, constituent shipment transit times (e.g., wherein each constituent shipment transit time is associated with a leg of a multi-leg shipment), active shipment transit times (e.g., wherein the package is being actively moved), passive shipment transit times (e.g., wherein the package is sitting in a warehouse), a duration between any two tracking details associated with a package (e.g., time between receipt of tracking detail N and tracking detail N−1 for the shipment), and/or any other suitable data related to package shipment time. Anomalous events can include package delays, loss, damage, theft, and/or any other uncommon event (e.g., that occurs below a threshold frequency for all shipments, all shipments associated with a specific set of metadata, etc.).”)(please see claim 5 rejection for combination rationale)
As per claim 6/15, Atwood does not disclose but McAlister discloses wherein the large language model is trained using a training corpus comprising a data frame built by combining freight logistic data, importer supply chain data, external factor data, and shipment outcomes (paragraph 36-38, 41-44, 48-51, 78-80, the model are trained by combining various data such as package data (logistic data), carrier data (importer supply chain data), weather data (external factor data) and date and time when the packages were delivered (shipment outcomes))(please see claim 4 rejection for combination rationale).
Claim(s) 7 and 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Atwood (US 11501245) in view of Xue, as disclosed in the rejection of claim 1, in further view of Ashby (US 11315207).
As per claim 7/16, Atwood does not disclose but Ashby discloses wherein the input data comprises Master Bill of Lading (MBL) numbers identifying a cargo booking (col. 5:31-41, “The system where the shipping container capacity data further includes at least one of the group including of: consignee city, state, carrier name, carrier code, container size, arrival date, port of arrival, bill of lading number, master bill of lading number, container type and container number. The system where the shipping container capacity data is drawn from at least one of public cargo data and private cargo data.”).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of the invention to include the limitations above as taught by Ashby in the teaching of Atwood, in order to identify and utilize otherwise unused, deadheading shipping containers (Ashby, abstract).
Claim(s) 8 and 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Atwood (US 11501245) in view of Xue, as disclosed in the rejection of claim 1, in further view of Skaaskrud (WO 2008088817).
As per claim 8/17, Atwood does not disclose but Skaaskrud discloses wherein input data is processed using optical character recognition to extract text from the input data (page 8, ln 30-38, “Another object of the present invention is to provide such a global Web-based shipping, tracking, and delivery network, wherein each pickup/delivery vehicle is equipped with a mobile digital image capture and processing (MICAP) system that is used to (i) capture and process digital images of single-sheet shipping documents (i.e. manifests, air bills, bills of lading, etc) at the point of pickup, (ii) automatically decode embedded bar codes and perform optical character recognition (OCR) on text presented therein, (iii) format the captured digital images with such decoded/recognized shipping information, and (iv) transmit the formatted image files to a data collection and processing center in the network, for automatic recognition processing and loading of the shipping information into the RDBMS maintained by the network.).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of the invention to include the limitation above as taught by Skaaskrud in the teaching of Atwood, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to OMAR ZEROUAL whose telephone number is (571)272-7255. The examiner can normally be reached Flex schedule.
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, Resha Desai can be reached at (571) 270-7792. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
OMAR . ZEROUAL
Examiner
Art Unit 3628
/OMAR ZEROUAL/Primary Examiner, Art Unit 3628