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
The status of the claims as of the response filed 01/22/2026 is as follows:
Claims 1-2 and 8-12 are pending.
Claims 3-7 are canceled.
Claims 1, 8 and 9 were amended.
All pending claims have been considered below.
Respond to Arguments:
35 USC 101 Rejection
Applicants’ arguments filed 01/22/2026 Page 8-9 with respect to amended claims 1, 8, and 9 have been fully considered and are not persuasive. The rejection under 35 U.S.C. § 101 is maintained for the reasons below:
Applicant argues that Claims 1, 8, and 9, particularly the exercise guidance to the user for correcting the cause of the problem ... corresponding to each most divergent index that most effectively improves the user’s running motion, integrate the recited abstract idea into a practical application because the specification states By searching such a hierarchical structure of indexes from an upper layer toward a lower layer ... a problem index that should be preferentially improved can be identified, a problem index that leads to effective improvement of exercise motions can be identified, and the guidance information can be generated, and With such a learning model, an index that leads to effective improvement of exercise motions can be preferentially set, based on the actual performance of the exercise motions.
The Examiner respectfully disagrees because, under proper BRI, Claims 1, 8, and 9 still use generic server, memory, processor, measurement device, network, and a functionally defined trained model to receive running data, compare it to reference data, select a prioritized index, and present guidance, rather than to improve the functioning of the computer, sensor, network, or model itself. MPEP 2106 explains that a claim integrates a judicial exception into a practical application only when the additional elements impose a meaningful limit on the exception, such as an improvement to computer functionality or other technology, and not when the claim merely links the exception to a technological environment or uses generic technology as a tool;
Thus, the applicant’s position is not persuasive because the amended most effectively improves and based on a degree of improvement language narrows the abstract evaluative logic, but still culminates in recommending guidance to the user instead of applying the exception in a manner that changes or improves technology itself. For the same reason, the additional elements do not amount to significantly more at Step 2B, and Claim 2 merely adds generating the guidance from the selected index while Claims 10-12 only further specify the hierarchical indexes. Therefore, the rejection is maintained.
Rejection Under 35 U.S.C. § 103
Applicant’s arguments, filed 01/22/2026 page 10-11, with respect to amended claims 1, 8, and 9 have been fully considered and are not persuasive. The rejections under 35 U.S.C. § 103 are maintained.
Applicant argues that Claim 1, "search the hierarchical indexes from an uppermost layer of the hierarchical indexes to a lowermost layer" and "search the hierarchical indexes from the lowermost layer towards the uppermost layer for a next most divergent index... excluding any previously determined most divergent index, based on a degree of improvement", is not disclosed, suggested, or rendered obvious by the combination of Katis, Mettler, and Verstegen.
The Examiner respectfully disagreed because under proper BRI, Claim 1, "search the hierarchical indexes from an uppermost layer of the hierarchical indexes to a lowermost layer" and "search the hierarchical indexes from the lowermost layer towards the uppermost layer for a next most divergent index... excluding any previously determined most divergent index, based on a degree of improvement" means a diagnostic logic that performs both a top-down categorization and a bottom-up traversal of a data hierarchy to iteratively identify correctable secondary outliers by filtering out previously identified primary errors and ranking remaining issues based on measured progress. Thus, the applicant’s position is not persuasive because the updated rejection maps these amended limitations to the prior art combination, demonstrating that Verstegen teaches an iterative closed-loop system to continuously adjust performance toward a baseline, and Mettler teaches searching a classification tree upward from the lowest level synergies to prioritize features based on whether they are changeable, which functionally filters out and excludes previously determined or unchangeable indexes to isolate the specific next target for improvement. Therefore, the rejection is maintained. Refer to below 35 U.S.C 103 for further details about 35 U.S.C 103 rejection.
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, 2, and 8–12 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1:
Step 1 determines whether the claims fall within one of the four statutory categories under 35 U.S.C. § 101: process, machine, manufacture, or composition of matter.
The claims encompass three statutory categories: process, machine, and manufacture.
Machine (Claim 1-2, 10): The language "a running improvement guidance device, comprising: a server comprising: a memory... and a processor..." recites a machine because it describes a concrete thing consisting of parts, such as memory and processor devices combined to perform functions.
Process (Claim 8, 11): The language "a running improvement guidance method, comprising: storing on a server a trained model... providing the measurement data... searching within each layer..." recites a process because it describes a series of acts, such as storing, providing, searching, and providing guidance.
Manufacture (Claim 9, 12): The language "a non-transitory computer-readable medium storing a running improvement guidance program causing a computer to implement..." recites a manufacture because it describes a tangible article, the medium, given a new form or property through man-made means to store and execute instructions.
Having confirmed the claims are directed to statutory subject matter, the analysis proceeds to Step 2A.
Step 2A Prong One:
Prong One evaluates whether the claim recites a judicial exception, such as an abstract idea, by identifying limitations that fall within enumerated groupings like mental processes, mathematical concepts, or certain methods of organizing human activity.
Recitation Independent Claim 1:
A running improvement guidance device, comprising:
a server comprising:
a memory that stores a trained model configured to receive over a network from a measurement device input of measurement data of a user during running and to output over the network to a display device an index among hierarchized indexes for improvement of at least one of a form of a running motion of the user and a force of the running motion of the user;
and a processor that provides the measurement data of the user during running as the input to the trained model and outputs to the display device exercise guidance to the user for correcting a cause of a problem of the at least one of the forms of the running motion of the user and the force of the running motion of the user based on the index output by the trained model,
wherein the trained model is configured to iteratively, until the measurement data of the user does not diverge from reference measurement data representing a normal value of users during running:
search the hierarchical indexes from an uppermost layer of the hierarchical indexes to a lowermost layer of the hierarchical indexes via an intermediate layer of the hierarchical indexes, the uppermost layer of the hierarchical indexes including major indexes used to evaluate running exercise motions, the intermediate layer of the hierarchical indexes including intermediate indexes corresponding to problems with the at least one of the form of the running motion and the force of the running motion within the major indexes, the lowermost layer of the hierarchical indexes including the index output by the trained model corresponding to causes of the problems with the form of the at least one of the running motion and the force of the running motion within the intermediate indexes,
search within each layer of the hierarchical indexes for a most divergent index among a plurality of indexes in the layer for which the measurement data of the user most diverges from the reference measurement data representing a normal value of users during running, corresponding to the plurality of indexes in the layer
and sequentially search between each layer of the hierarchical indexes from the most divergent index in each layer of the hierarchical indexes towards the lowermost layer of the hierarchical indexes,
determine a degree of improvement of the at least one of the running motion and
the force of the running motion corresponding to the most divergent index in the lowermost
layer,
search the hierarchical indexes from the lowermost layer towards the uppermost layer for a next most divergent index among the plurality of indexes in the layer for which the measurement data of the user most diverges from the reference measurement data representing the normal value of users during running corresponding to the plurality of indexes in the layer, excluding any previously determined most divergent index, based on a degree of improvement of the at least one of the running motion and the force of the running motion corresponding to the plurality of indexes in each layer of the hierarchical indexes,
and the processor is configured to output to the display device the exercise guidance to the user for correcting the cause of the problem with the form of the at least one of the running motion and the force of the running motion corresponding to each most divergent index that most effectively improves the form of the at least one of the running motion and the force of the running motion of the user.
Note: Bold element, are additional elements further evaluated under prong two and step 2B, non-bold are the identify abstract idea.
Claims Categorized Rational:
Under the broadest reasonable interpretation, claim 1 recites a trained model configured to receive ... measurement data of a user during running, output ... an index among hierarchized indexes, search the hierarchical indexes from an uppermost layer ... to a lowermost layer, search within each layer ... for a most divergent index, determine a degree of improvement, search ... for a next most divergent index ... excluding any previously determined most divergent index, and output ... exercise guidance to the user. Functionally, claims 1, 8, and 9 recite a process that receives running data, compares that data to reference normal data, prioritizes which running problem is most significant within a hierarchy, then reevaluates progress based on improvement, and selects and outputs the next corrective guidance most likely to improve the user’s motion.
These amended limitations still recite a mental process because they amount to observing running information, evaluating that information against a baseline, judging which index is most divergent, assessing improvement, excluding a previously chosen focus, and deciding what guidance should be given next. Those are the kinds of observation, evaluation, judgment, and opinion identified in MPEP 2106 as mental processes, and they can practically be performed by a human running coach, at least with pen and paper, by watching a runner, comparing the runner to normal form, organizing potential faults into a hierarchy, picking the largest problem first, reassessing improvement, and then choosing the next corrective focus. The claims also reasonably recite a certain method of organizing human activity because the final step is not merely diagnosis, but output ... exercise guidance to the user for correcting the cause of the problem and outputting ... the exercise guidance ... corresponding to each most divergent index that most effectively improves the running motion, which directs how the user is to behave during running and thus fits the managing personal behavior and following rules or instructions sub-grouping. Accordingly, the amendments add further prioritization logic, but they do not change the Step 2A, Prong One result. Claims 1, 8, and 9 still recite a judicial exception in the form of a mental process and additionally a certain method of organizing human activity.
Dependent claim 2 does not add limitations that move the claim outside of the abstract idea category. It recites generating the exercise guidance based on the index, which is the logical conclusion of the mental evaluation. Generating guidance is what a human coach would do after identifying the problem and does not add a patent-eligible concept beyond the underlying abstract evaluation.
Claim 10-12 just describe what are the differences hierarchy indexes, that it is for example analogous to a coach include that recited specify indexes in their mental process of identifying a problem and correcting the cause. Therefore, claims 10-12 do not overcome prong one because they still recite an abstract idea—a categorized evaluation.
The claims recite a judicial exception because they set forth the abstract idea of performing a mental process of categorized evaluation. The analysis therefore proceeds to Prong Two.
Step 2A Prong Two
The claims as a whole do not integrate the judicial exception into a practical application but are instead directed to the abstract idea itself.
The additional elements recited in the claims do not integrate the abstract idea into a practical application; they merely provide a generic technological environment for it.
The claims recite performing the abstract evaluation using generic computer components (server, processor, memory, Measurement Device) and a generic network to receive data and transmit results. This is an example of merely "linking the use of a judicial exception to a particular technological environment," which is not a practical application (MPEP § 2106.05(h)).
The claims do not specify any improvement in the functioning of the computer, network, or sensor technology itself. Instead, they use these components as mere tools to automate the abstract process of providing exercise advice (MPEP 2106.05(f)). The specification confirms the generic nature of these components, listing a "wearable device, such as a smartwatch, or a smartphone" (para. [0015]) as the measurement device and a "smartphone, tablet, or personal computer" (para. [0021-0022]) as the display device. These limitations simply provide a generic technological environment for the abstract idea and do not add an inventive concept.
The trained model itself, as claimed, does not confer eligibility at this step. It is defined functionally by the abstract steps it performs (searching, comparing, identifying), not by a specific technical structure or process that improves how the computer operates. The applicant's specification discusses that a "learning model... may be generated by machine learning" (para. [0023]), but the claims do not recite a specific AI architecture or training process that overcomes a technical problem in computing. The claims simply take the abstract idea of a coach's evaluation process and state to "apply it" using a trained model on a computer, which is insufficient to establish a practical application (MPEP § 2106.05(f)).
The dependent claims 2 limitation of "generates the exercise guidance" adds no further integration. It is simply another data processing step that is part of the abstract idea itself and does not transform the claim into a practical application of the exception. The generation of advice is the inherent goal of the abstract evaluation, not a separate, integrating application. Viewed in combination, these additional elements do not integrate the abstract idea; they collectively describe a generic system for automating a mental process, which does not impose a "meaningful limit on the judicial exception" (MPEP § 2106.04(d)).
The dependent claims 10-12 limitation does not recite new additional element just further describe the abstract idea of independent claims as describe above in prong one.
The claims as a whole do not integrate the judicial exception into a practical application. Therefore, the analysis proceeds to Step 2B.
Step 2B:
The claims, evaluated individually and as a whole, fail to provide an inventive concept because the additional elements do not amount to “significantly more” than the judicial exception. In Step 2B, the analysis determines whether a claim directed to a judicial exception recites additional elements that "transform the nature of the claim" into a patent-eligible application. An inventive concept "cannot be furnished by the unpatentable... abstract idea... itself" (MPEP § 2106.05). This rationale will demonstrate that the limitations here do not supply that inventive concept.
Independent Claims Analysis
Data Processing System (Server, Memory, Processor) (Claims 1, 8, 9) The claims recite a server, memory, and processor that do not provide an inventive concept. The specification describes these in generic terms, stating the "exercise improvement guidance device 100 is configured on a server capable of communicating with the measurement device 20 and the display device 30" (para. [0022]) and that the functions "may be implemented as application software running on" an information processing device like a "smartphone, tablet, or personal computer" (para. [0022]). These descriptions portray the components as generic hardware performing their most basic functions of storing data and executing instructions. MPEP § 2106.05(f) clarifies that claim amounting to "mere instructions to implement an abstract idea on a computer" lack an inventive concept. The specification provides no details on any specialized configuration or unconventional operation of these components that would constitute an improvement to computer technology itself. Therefore, the data processing system does not provide "significantly more" than the abstract idea. It is an example of an attempt to "monopolize the [judicial exception]" by "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use" (MPEP § 2106.05(h), quoting Diamond v. Diehr).
Network and Peripheral Devices (Measurement Device, Display Device) (Claims 1, 8, 9) The claims recite a network and peripheral devices that do not provide an inventive concept. The claims recite receiving measurement data "over a network from a measurement device" and outputting guidance "over the network to a display device". These are examples of "insignificant extra-solution activity" (MPEP § 2106.05(g)). Specifically, receiving data from a sensor is mere data gathering, a pre-solution activity. Outputting guidance to a display is a post-solution activity. The specification confirms the generic nature of these components, listing a "wearable device, such as a smartwatch, or a smartphone" (para. [0015]) as the measurement device and a "smartphone, tablet, or personal computer" (para. [0021-0022]) as the display device. These limitations simply provide a generic technological environment for the abstract idea and do not add an inventive concept.
Trained Model (Claims 1, 8, 9) The claims recite a "trained model" that does not provide an inventive concept. The specification suggests that "a learning model... may be generated by machine learning" (para. [0023]) and that the hierarchical structure can be automatically generated by "autonomously analyzing a term used in each index and its biomechanical meaning" (para. [0023]). However, the claims themselves do not recite any specific, unconventional AI technique, training method, algorithm, or data structure that improves AI technology itself. The model is defined purely by its function—automating the abstract steps of hierarchical comparison. This is an attempt to claim the idea of a solution, rather than a particular technical means for achieving it. The trained model limitation does not amount to significantly more than the abstract idea it implements.
Claim "As a Whole" Analysis Viewed in combination, the additional elements fail to supply an inventive concept. The claim, taken as a whole, simply automates the abstract idea of providing personalized exercise coaching using a collection of generic computer components for their intended and conventional purposes. The combination of these elements does not result in an improvement to the functioning of the computer itself or to any other technology, nor does it effect a transformation of an article into a different state or thing. The elements together provide a mere technological setting for the abstract idea, which does not amount to "significantly more."
Dependent Claims Step 2B Evaluation
Claim 2 does not add an inventive concept. Claim 2 adds the limitation that the processor "generates the exercise guidance based on the index". The specification describes this step in general terms: "The guidance information thus generated is transmitted to the display device 30 and displayed on the display screen thereof" (para. [0031]). This limitation is an inherent part of the abstract idea itself—the goal of evaluating performance is to generate guidance. It is an insignificant post-solution activity that does not add an inventive concept to the combination. The dependent claims 10-12 limitation does not recite new additional element just further describe the abstract idea of independent claims as describe above in prong one.
The limitations added by the independent and dependent claims, both individually and in combination, fail to provide an inventive concept. They merely narrow the field of use to exercise guidance and add insignificant, pre- and post-solution activities (data gathering, displaying results) to the core abstract idea of categorized evaluation. As such, the claims as a whole do not amount to significantly more than the judicial exception and are therefore ineligible under 35 U.S.C. § 101.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-2, 8-9, are rejected under 35 U.S.C. § 103 as being unpatentable over Katis US 2018/0104541 in combination with Mettler (US 2017/0061817) and in view of Verstegen (US 2009/0269728).
Claim 1:
Katis teaches A running improvement guidance device, comprising:
a server comprising: (Katis, Abstract, para. 0018, 0021-0022, 0030).
Katis describes an automated system that provides suggestions for improving the athlete's performance for movements including distance running, which is semantically and functionally equivalent to a running improvement guidance device. This system utilizes cloud resources 150 to house its main components like the ANN 120 and coaching engine 140, which is functionally a server that provides services to a user's mobile device 160.
a memory that stores a trained model configured to receive over a network from a measurement device input of measurement data of a user during running and to output over the network to a display device an index among hierarchized indexes for improvement of at least one of a form of a running motion of the user and a force of the running motion of the user; (Katis, 0003, 0018, 0019, 0020, 0022-0023, 0027-0031, 0036).
Katis describes a coaching system that functions as a running improvement guidance device, where cloud resources 150 including a computer server act as the server. This server contains a reference database 122 (a memory) that holds data used to train an artificial neural network (ANN) 120 (a trained model). The system is configured to receive over a network (WiFi, Bluetooth) data from a scanner 130 (a measurement device). This data is input of measurement data (position data) captured from a user during a movement (which under BRI includes running supported by paragraph 0030). The system is further configured to output over the network to a display device (mobile device 160) information such as rank data or a score ("91.50/100"), which constitutes an index among hierarchized indexes. This output is provided for improvement of the user's technique, as Katis explicitly mentions providing correction to form of a movement, and for improvement of the user's power, as the system analyzes acceleration, which is directly related to force.
and a processor that provides the measurement data of the user during running as the input to the trained model and outputs to the display device exercise guidance to the user for correcting a cause of a problem of the at least one of the form of the running motion of the user and the force of the running motion of the user based on the index output by the trained model, (Katis, 0004, 0020-0022, 0028-0031).
Katis discloses a system where its processor (the ANN 120 and coaching engine 140) provides the measurement data (position data 132) as the input to the trained model (ANN 120). The processor then outputs to the display device (mobile device 160) exercise guidance (coaching information) for improving the form of the running motion (correction to form) and force (derived from acceleration data), and this guidance is generated based on the index output by the trained model (rank data). Katis discloses that the guidance is for correcting a problem with the user's form. The coaching information explicitly includes instructions to correct a deficiency in the performance and suggestions for modifying the user's form. This is functionally identical to guidance for "correcting a cause of a problem... of the form of the running motion."
wherein the trained model is configured to (Katis, paragraph 0020-0022, 0031, 0033, fig.5, 0045)
Katis reads on measuring user data against a reference representing a normal value because it identifies a specific flaw based on how the user's movement diverges from an expected baseline ideal model.
the intermediate layer of the hierarchical indexes including intermediate indexes corresponding to problems with the at least one of the form of the running motion and the force of the running motion within the major indexes, the lowermost layer of the hierarchical indexes including the index output by the trained model corresponding to causes of the problems with the form of the at least one of the running motion and the force of the running motion within the intermediate indexes, (Katis, paragraphs 0021, 0022, 0027, 0029, 0033).
Katis describes a system where an ANN 120 (trained model) processes user data. Katis’s system of a movement category with associated grades and flaws constitutes a set of hierarchical indexes, and the ANN's processing of data to identify a specific flaw for correction represents a systematic examination. This process flows from a general movement category to a specific flaw, which is a functional search from an uppermost layer... to a lowermost layer. The final output is specific coaching instructions based on an identified flaw, which represents an index output from the lowermost layer.
Katis discloses an intermediate layer corresponding to problems. Katis's reference data set 300 contains rank data that includes identification of one or more flaws in the movement, such as a deficiency in technique. This "flaw" is a "problem with the form" and functions as an "intermediate index" within the "major index" (the movement category).
The ANN 120 outputs rank data (the "index output by the trained model"). This rank data (index) is used by the coaching engine to look up coaching instructions (para. 0029). These instructions, such as C1... suggestions for modifying the user's form, are the "causes" or, more accurately, the corrections for the "problems" (the "flaws" from the intermediate layer). The rank value itself is the "index" in the "lowermost layer" that correlates to and corresponds to the "cause" (the corrective instruction).
search within each layer of the hierarchical indexes for a (Katis, abstract, paragraphs 0003, 0005, 0021-0022, 0026-0028, 0034).
Katis uses an “artificial neural network (ANN) trained via a reference data set”, which meets the interpretation of a trained model. This system organizes movements into a general “movement category” and then identifies a specific “class or subset” or “flaw”, which functionally creates a set of hierarchical indexes. The process starts by applying the user's overall “position data” to the ANN trained for that “movement category” and drills down to identify a specific “flaw”, which constitutes a search from an uppermost to a lowermost layer. Katis compares the user's performance to a “reference data set” to find “similarities” and identify a “distinction” from a higher-ranked movement.
and sequentially search between each layer of the hierarchical indexes from (Katis, paragraphs 0003, 0005, 0022, 0027, 0034).
Katis employs a trained model (an artificial neural network) to first determine a high-level assessment, the “rank value” for the performance. The system then uses this high-level assessment as a basis for a sequential search between each layer by drilling down into a lower layer, as it is “based on the rank value” that the system determines specific “coaching information”. This coaching can include notifying the user of a specific “distinction” (a lower-level problem) and providing “instructions for traversing the distinction”.
determine a degree of improvement of the at least one of the running motion and the force of the running motion corresponding to theKatis, par. 0031, 0021, 0006)
Katis describe the use of rank data and positional movement data to compare a user's current physical performance against their own previously recorded performances to evaluate how much the movement has progressed.
search, The coaching engine 140... may be compared against data regarding previous performances by the user 110 to determine whether how the user's performance has changed or improved... identify one or more flaws in the movement, such as a deficiency in technique... The rank data can indicate... a best match for the performed movement, where each class may correspond to a given characteristic, performance level, or flaw. (Katis, Paras. [0021], [0022], [0031]).
Katis read comparing current performance against previous performances to determine improvement and identifying one or more flaws in technique which represents a divergence from a best match or reference data.
and the processor is configured to output to the display device the exercise guidance to the user for correcting the cause of the problem with the form of the at least one of the running motion and the force of the running motion corresponding to . Based on the rank data, the athlete can be presented with an evaluation of the performance, instructions for subsequent movements and suggestions for improving the athlete's performance... the mobile device 160 may display results 162 of the performance... and instructions for a next movement 164 (e.g., instructions to correct a deficiency in the performance). (Katis, Paras. [0001], [0021-0022], 0034, ).
Katis read on a processor (mobile device 160) configured to output to the display device (display results 162) exercise guidance (instructions for a next movement) for correcting the cause of the problem (correct a deficiency in the performance) regarding running motion (athletic performance).
Obvious Rationales:
Katis does not teach a diagnostic search for the greatest point of error.
Mettler discloses a system that first decomposes a movement into a hierarchy of movement units and movement phases (Mettler, para. 0018), which is functionally equivalent to the claimed hierarchical indexes. The core of Mettler's analysis is to then identify and focus on the most critical error. Mettler's feedback synthesis process explicitly includes steps to 1) Determine the movement features... that the user has the ability to change... 2) Determine movement features that display a significant impact on an outcome... [and] 3) Prioritize movement features that are both changeable and have the most significant impact on outcomes (Mettler, para. 0289-0290). This act of prioritizing the feature with the greatest impact on outcome is functionally the same as searching for the most divergent index, and the process of analyzing decomposed phases and synergies represents a sequential search from the most divergent index... towards the lowermost layer.
It would have been obvious to a person of ordinary art skill to combine Katis's teachings with Mettler because both references are in the same field of art and are directed toward solving the same problem: providing effective, automated feedback to improve a user's movement skills.
Katis aims to determine coaching information for the user (Katis, para. 0022), but its feedback is based on a general "rank." Mettler explicitly seeks to improve upon such general feedback by providing a method to identify specifically where change is required in order to provide actionable information that can be leveraged directly for training (Mettler, paras. 0012, 0065).
Mettler explicitly teaches that its method of prioritizing feedback allows a user to close the training or rehabilitation loop and run it as an iterative scheme (Mettler, para. 0065). A POSITA would recognize this as a significant improvement over the more general feedback provided by Katis. Integrating Mettler's analytical software logic into Katis's established client-server hardware architecture would be a predictable combination of known elements, providing a reasonable expectation that the resulting system would successfully deliver more targeted and effective coaching.
Katis teaches a server comprising a trained model that evaluates running motions, disclosing that it “outputs rank data”, 0021 which indicates a relationship between the performance and recorded movements (Katis, para. [0003]). Katis teaches that this output is used to determine instructions, identifying “one or more flaws in the movement, such as a deficiency in technique” (Katis, para. [0027]), which corresponds to providing an index for improvement.
However, Katis utilizes an Artificial Neural Network (ANN) to process position data directly into rank data (Katis, para. [0003]). Katis fails to explicitly disclose searching the hierarchical indexes from an uppermost layer... via an intermediate layer... to a lowermost layer;
Mettler teaches the specific hierarchical indexes structure required by the claim, explicitly decomposing movement into “three primary levels of movement organization” (Mettler, para. [0112]): (i) “Movement Repertoire” and “movement profiles” (Uppermost/Major, paras. [0170], [0197]); (ii) “Movement Phases” which identify internal structure and technique flaws (Intermediate/Problems, para. [0114]); and (iii) “Movement Synergies” which correspond to neuromuscular implementation (Lowermost/Causes, para. [0115]). Verstegen teaches the search mechanism to navigate such a structure. Once the motion problem is identified via the screen, the system analyzes why the problem exists by mimicking the cognitive analysis of a human expert. The system uses “business logic integrating the various data points collected” (Verstegen, para. [0053]) to analyze the results of the movement screen against the athlete's history. To break down the problem (e.g., determining if a squat failure is due to ankle mobility vs. hip strength), the system employs an “expert System” that “could be Something as simple as a decision tree that is based on an expert trainer's responses to a series of questions” (Verstegen, para. [0024]). The breakdown includes analyzing “current state of an injury” and “injury history” (Verstegen, para. [0024]) to ensure the solution is safe.
While Katis provides the automated coaching platform, the combination with Mettler and Verstegen makes obvious the specific method of diagnosing the error via hierarchical search. A PHOSITA would interpret the claimed “hierarchical indexes” as reading on Mettler’s “hierarchical organization” (para. [0163, 0170]) and would recognize that Katis’s system is improved by integrating Verstegen’s logic to navigate it. By applying Verstegen’s “decision tree” logic (para. [0024]), the system searches through Mettler’s layers: starting at the “Movement Repertoire” (Uppermost), analyzing deviations in the “Movement Phases” (Intermediate), and determining the root cause in the “Movement Synergies” (Lowermost). This integration replaces the opaque nature of Katis’s ANN with the transparent, logical breakdown of Verstegen, allowing the system to “identify specifically where change is required” (Mettler, para. [0065]) using “business logic” (Verstegen, para. [0053]).
It would have been obvious to combine Katis with Mettler and Verstegen because all three references reside in the same field of automated athletic training and seek to solve the shared problem of providing accurate, actionable feedback to a user. Katis aims to provide instructions for correction to form (para. [0022]). Mettler provides the motivation to use a hierarchical structure because “complex movements are obtained by combination of motion segments or phases” (para. [0082]), making decomposition necessary for precise analysis. Verstegen provides the motivation to use decision-tree search logic to “mimic the thought processes employed by a professional trainer” (para. [0011, 0024, 0054]), ensuring the automated advice is as logical and reliable as a human expert's.
A person of ordinary skill in the art would have been motivated to integrate Mettler’s hierarchical data structure and Verstegen’s decision-tree search into Katis’s system to achieve the benefit of more granular and understandable diagnostics. Mettler explicitly teaches that this hierarchical decomposition “provides a sparse description of the high-dimensional movement data” and “plays a central role in functional analysis” (para. [0020]). Furthermore, Verstegen teaches that using such logic allows the system to “mimic the thought processes employed by a professional trainer” (para. [0024]). There is a reasonable expectation of success because hierarchical data structures (Mettler) and decision tree logic (Verstegen) are well-known, compatible computational techniques for diagnostic systems.
Rationale for missing elements per applicant amended limitations in Claim 1:
Claim 1 recites wherein the trained model is configured to .
Katis teaches identifying flaws by comparing user data to a reference, as shown by identify parallels between the position data 132 and the reference data set 300... identify one or more flaws in the movement, such as a deficiency in technique (Katis, Paras. [0026-0028]). This reads on comparing measurement data to a reference because Katis uses an ANN to find "distinctions" or "flaws" against a recorded reference set. However, Katis does not teach the specific iterative loop that continues until a state of non-divergence (parity) is met. Verstegen teaches that missing feature, as shown by modify the control signal so that the selected value... will cause the metabolic parameter of the athlete to tend to the preselected value... repeat these actions until the athlete's metabolic rate is stable within a target range (Verstegen, Paras. [0010], [0038]), which reads on an iterative convergence loop because it utilizes a closed-loop system to repeatedly adjust performance until a specific baseline value is achieved.
A person of ordinary skill in the art would have combined Katis with Verstegen to automate the correction of running form without constant manual intervention, by integrating Verstegen’s closed loop control system (Para. [0038]) into Katis’s artificial neural network (Para. [0021]) to create a self-correcting training model. This combination is motivated by the desire to automate a substantial portion of the athletic training process (Verstegen, Para. [0007]) and ensure the athlete engages in optimal techniques (Katis, Para. [0073]) by iteratively refining the motion until it tends toward a preselected (Verstegen, Para. [0038]) expected baseline ideal model (Katis mapping logic). Doing so would have predictably resulted in a system that automatically and precisely converges a user's running metrics toward a desired normative value through repetitive, data-driven adjustment.
Claim 1 recites search.
Under the broadest reasonable interpretation, this limitation requires a diagnostic logic that performs a bottom-up traversal of a data hierarchy to identify secondary outliers (flaws) by mathematically filtering out previously identified primary errors and ranking remaining issues based on their measured progress or potential for improvement. Katis teaches comparing performance against previous sessions to identify technical flaws, as shown by compared against data regarding previous performances... to determine whether how the user's performance has changed or improved... identify one or more flaws in the movement, such as a deficiency in technique (Katis, Paras. [0022], [0027]). This reads on searching for divergent indexes based on improvement because Katis tracks how specific form deficiencies change over time relative to a reference model. However, Katis does not teach the specific bottom-up search (lowermost toward uppermost) nor the exclusionary logic required to find a next most divergent index by specifically excluding previously identified ones.
Mettler teaches searching a hierarchical data structure to identify and prioritize specific movement flaws, as shown by navigating the classification tree starting “at the lowest level, the movement synergies” upward to overall stroke outcomes, and establishing a “hierarchical prioritization... based on whether a feature of the movement unit is changeable” (Mettler, Paras. [0021], [0027], [0432]). A person of ordinary skill in the art would understand that Mettler’s framework functionally reads on searching from the lowermost layer toward the uppermost layer for a next most divergent index because the system systematically evaluates base-level mechanical deviations (synergies/phases) to correct higher-level movement errors (stroke outcomes). Furthermore, prioritizing only the actionable changeable features filter out and exclude any previously determined (e.g., unchangeable or already resolved) divergent indexes, thereby isolating the specific next target for improvement.
A person of ordinary skill in the art would have combined before the effective filing date Katis with Mettler to effectively correct a deficiency in the technique (Katis, Para. [0027]) and accelerate acquisition of complex movement skills (Mettler, Para. [0014]), by configuring Katis's flaw-identification neural network to process performance data using the bottom-up hierarchical classification tree of Mettler. Because applying Mettler's hierarchical prioritization... based on whether a feature... will impact a quality or outcome (Mettler, Para. [0027]) to Katis's multiple identified flaws in the movement (Katis, Para. [0022]) ensures the system can identify specifically where change is required (Mettler, Para. [0065]) rather than presenting the user with an overwhelming, unranked list of errors. Doing so would have predictably resulted in an optimized, iterative coaching system that systematically isolates the next most critical correctable athletic flaw while efficiently excluding previously resolved or non-actionable mechanical errors.
Claim 1 recites and the processor is configured to output to the display device the exercise guidance to the user for correcting the cause of the problem with the form of the at least one of the running motion and the force of the running motion corresponding to
Katis relies on general rank data and does not explicitly teach that this output specifically corresponds to an index evaluated to "most effectively" improve the form. Mettler teaches this functional requirement, as shown by prioritized based on whether a feature of the movement unit is changeable and will impact a quality or outcome... Prioritize movement features that are both changeable and have the most significant impact on outcomes (Mettler, Paras. [0027], [0290]), which reads on guidance corresponding to an index that most effectively improves the form because it explicitly selects and outputs corrections based on their maximized ("most significant") impact on the user's overall performance.
A person of ordinary skill in the art would have combined before the effective filing date Katis with Mettler to accelerate the acquisition of complex movement skills (Mettler, Para. [0014]), by integrating Mettler’s impact-based prioritization logic (Para. [0290]) into Katis’s coaching output engine (Para. [0022]). This combination is legally justified because, as established in the preceding limitations, applying Mettler's hierarchical search isolates the most critical error; consequently, outputting Katis's instructions to correct a deficiency (Katis, Para. [0022]) based on Mettler's most significant impact on outcomes (Mettler, Para. [0290]) ensures the user is not wasting time on minor flaws. Doing so would have predictably resulted in a coaching system that outputs guidance precisely targeted at the specific mechanical root cause that yields the maximum effective improvement in the user's running form.
Claim 2:
Katis in combination with Mettler and Verstegen teaches, The running improvement guidance device according to claim 1, wherein the processor generates the exercise guidance based on the index output by the trained model. (Katis, paragraphs 0003, 0021-0022, 0028, claim 1).
Katis discloses a “coaching engine”, which is a processor, that is responsible for “determining instructions for a second movement”, which is functionally how the processor generates the exercise guidance. This generation of instructions is explicitly performed “based on the rank value”, which is the index output by the trained model (“the ANN”). The “rank value” is the metric determined by the ANN that “indicates a relationship between the performance of the first movement and a subset of the plurality of recorded movements”.
Note: Claims 8-9 are rejected with the same analysis above for being very similar to Claim 1-2.
Claims 10-12 are rejected under 35 U.S.C. § 103 as being unpatentable over Katis (US 2018/0104541) in view of Mettler (US 2017/0061817), and in view of Verstegen (US 2009/0269728), and further in view of Souza (Reference U), and Teng (Reference V). , refer to PTO-892 for NPLs.
Claim 10.
Katis in combination with Mettler and Verstegen teaches, The running improvement guidance device according to claim 1, wherein the major indexes include
Katis in view of Verstegen teaches the system of claim 1, including a processor and trained model that performs a hierarchical search by analyzing a particular movement category (e.g., a box jump) as an uppermost layer, identifying one or more flaws in the movement, such as a deficiency in technique as an intermediate layer, and mapping those flaws via rank data to coaching instructions as a lowermost layer as described in claim 1 mapping above.
However, Katis fails to disclose the specific hierarchical indexes for running, namely wherein the major indexes include landing impact, vertical motion, and braking force, and the lowermost indexes include ground contact position and ground contact angle.
Souza (Reference U) teaches the missing indexes in bold, describing a systematic video-based running biomechanics analysis that measures Foot strike patterns (mapping to ground contact position), Foot inclination angle (mapping to ground contact angle), and identifies higher peak vertical ground reaction force (mapping to landing impact/vertical motion) and braking impulse (mapping to braking force) as key factors in running injury (Souza, Abstract and Figures 2-3).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to combine the teachings of Katis with Souza because both references share the purpose of tracking and evaluating athletic performance to prevent injury (Katis, par. 0034, 0030; Souza, Introduction). Specifically, a POSITA developing Katis's distance running mode would look to Souza to supply the validated Evidence-Based biomechanical parameters required to populate the hierarchical layers of the trained model with meaningful data rather than arbitrary inputs (Souza, Title).
A person of ordinary skill in the art would have been motivated to integrate the ground contact and impact indexes from Souza into the system of Katis to achieve the benefit of reducing injury risk, as Souza teaches that Running biomechanics play an important role in the development of injuries and identifying these variables allows for the development of treatment strategies (Souza, summary).
A PHOSITA would have had a reasonable expectation of success in combining the references because the modification required only ordinary skill and routine experimentation. Souza demonstrates that these variables are essentially measuring 2-dimensional (2D) video surrogates (Souza, Summary), which is perfectly compatible with Katis's method of generating position data based on image data.
The combination of Katis, Mettler, Verstegen, and Souza teaches the running improvement guidance device with hierarchical indexes for the lower extremities (landing impact, braking force, ground contact). However, this combination fails to disclose wherein the lowermost indexes of the lowermost layer include trunk angle.
Teng (Reference V) teaches the Missing Element in bold, describing a study where The trunk angle was calculated as the orientation of the trunk segment relative to the global coordinate system (Teng, Methods) and identifying sagittal plane trunk posture as a critical variable (Teng, Abstract).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to combine the teachings of Katis + Souza with Teng because both references operate in the same field of running biomechanics and injury prevention.
A POSITA would recognize that Souza's lower-body analysis is complemented by Teng's upper-body analysis, creating a comprehensive "whole-body" index list for the processor, as Teng explicitly links trunk mechanics to knee stress which is a common running injury (Teng, Abstract).
A person of ordinary skill in the art would have been motivated to integrate the trunk angle from Teng into the system of Katis to achieve the benefit of reducing joint stress, as Teng teaches that Incorporation of a forward trunk lean may be an effective strategy to reduce PFJ Patellofemoral Joint stress during running (Teng, Conclusion).
A PHOSITA would have had a reasonable expectation of success in combining the references because the modification required only ordinary skill and routine experimentation. Integrating a defined kinematic vector like trunk angle (from Teng) into a processor capable of tracking athlete position data (as taught by Katis ) is a routine task in computer vision and biomechanics analysis.
Note: Claims 8-9, 11-12 are rejected by the same analysis above as being very similar.
Relevant Prior Arts:
US20220001236 figure 10, describe a process where apparatus, players swing parameter information is obtained, analyses, focused, to provide a feedback, then player executes practice swing, measured, analyses and provided fast feedback and include a conditional whether coaching parameter are being correctly identified if not repeat the loop.
US20160271452 refer to fig. 1 that present a iterative loop of improvement through collection of know exercise data, computed data analytic, compare actual performance to ideal metrics, instructions and feedback to exerciser and also poste exercise performance summary.
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 JOSHUA DAMIAN RUIZ whose telephone number is (571)272-0409. The examiner can normally be reached 0800-1800.
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, Shahid Merchant can be reached at (571) 270-1360. 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.
/JOSHUA DAMIAN RUIZ/Examiner, Art Unit 3684
/KAREN A HRANEK/Primary Examiner, Art Unit 3684
Prosecution Insights