DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Objections
2. Claim 1, 6, 11, 16, and 22 are objected to because of the following informalities: a) In claim 1 lines 2-3, please change
“obtaining a neural network, the neural network being trained to determine a state change of a physical system over time, training data of the neural network indicating states” to:
--obtaining a neural network, the neural network being trained to determine a state change of a physical system over time, and training data of the neural network indicating states--.
b) In claim 6 lines 7-10, please change
“determining a negative value of the first message feature representation as a second message feature representation from the second physical unit to the first physical unit, the first message feature representation characterizing an effect of the second physical unit on the first physical unit” to:
--determining a negative value of the first message feature representation as a second message feature representation from the second physical unit to the first physical unit, the second message feature representation characterizing an effect of the second physical unit on the first physical unit--.
*If this is indeed an inadvertent typo, appropriate corrections may need to be made in the specification as well.
c) In claim 11 lines 6-7, please change
“obtaining a neural network, the neural network being trained to determine a state change of a physical system over time, training data of the neural network indicating states” to:
--obtaining a neural network, the neural network being trained to determine a state change of a physical system over time, and training data of the neural network indicating states--.
d) In claim 16 lines 7-10, please change
“determining a negative value of the first message feature representation as a second message feature representation from the second physical unit to the first physical unit, the first message feature representation characterizing an effect of the second physical unit on the first physical unit” to:
--determining a negative value of the first message feature representation as a second message feature representation from the second physical unit to the first physical unit, the second message feature representation characterizing an effect of the second physical unit on the first physical unit--.
*If this is indeed an inadvertent typo, appropriate corrections may need to be made in the specification as well.
e) In claim 22 lines 4-5, please change “obtaining a neural network, the neural network being trained to determine a state change of a physical system over time, training data of the neural network indicating states” to:
--obtaining a neural network, the neural network being trained to determine a state change of a physical system over time, and training data of the neural network indicating states--.
Appropriate correction is required.
Claim Rejections - 35 USC § 101
3. 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 1-19 and 22 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.
In view of the new 2019 Revised Patent Subject Matter Eligibility Guidance (Federal Register Vol. 84, No. 4, January 7, 2019), the Examiner has considered the claims and has determined that under step 1, claims 1-10 are to a process, claims 11-19 are to a machine, and claim 22 is to an article of manufacture. Next under the new step 2A prong 1 analysis, the claims are considered to determine if they recite an abstract idea (judicial exception) under the following groupings: (a) mathematical concepts, (b) certain methods of organizing human activity, or (c) mental processes. The independent claims contain at least the following bolded limitations (see representative independent claims) that fall into the grouping of mathematical concepts:
1. A method for state prediction, comprising:
obtaining a neural network, the neural network being trained to determine a state change of a physical system over time, training data of the neural network indicating states of a plurality of physical systems at a plurality of times;
obtaining state data corresponding to a state of a target physical system at a first time, the state data indicating a plurality of physical units comprised in the target physical system, material properties of the plurality of physical units, and interaction relationships between the plurality of physical units;
determining respective unit feature representations of the plurality of physical units in the target physical system based at least on target values of respective material properties of the plurality of physical units; and determining a state of the target physical system at a second time based on the state data by inputting at least the unit feature representations to the neural network.
11. An electronic device, comprising:
at least one processing unit; and
at least one memory coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit, the instructions, when executed by the at least one processing unit, causing the device to perform the following actions:
obtaining a neural network, the neural network being trained to determine a state change of a physical system over time, training data of the neural network indicating states of a plurality of physical systems at a plurality of times;
obtaining state data corresponding to a state of a target physical system at a first time, the state data indicating a plurality of physical units comprised in the target physical system, material properties of the plurality of physical units, and interaction relationships between the plurality of physical units;
determining respective unit feature representations of the plurality of physical units in the target physical system based at least on target values of respective material properties of the plurality of physical units; and determining a state of the target physical system at a second time based on the state data by inputting at least the unit feature representations to the neural network.
22. A non-transitory computer-readable storage medium having a computer program stored thereon, the computer program, when executed by a processing unit, causing the processing unit to perform the following actions: obtaining a neural network, the neural network being trained to determine a state change of a physical system over time, training data of the neural network indicating states of a plurality of physical systems at a plurality of times;
obtaining state data corresponding to a state of a target physical system at a first time, the state data indicating a plurality of physical units comprised in the target physical system, material properties of the plurality of physical units, and interaction relationships between the plurality of physical units;
determining respective unit feature representations of the plurality of physical units in the target physical system based at least on target values of respective material properties of the plurality of physical units; and determining a state of the target physical system at a second time based on the state data by inputting at least the unit feature representations to the neural network.
It is important to note that a mathematical concept need not be expressed in mathematical symbols, because "[w]ords used in a claim operating on data to solve a problem can serve the same purpose as a formula."(see MPEP 2106.04(a)(2) I.). The definition given in the specification is that a neural network is a “physics” engine (see Specification filed 05/28/2024, paragraph [0027]) based on a numerical parameter-based “model” (see Specification filed 05/28/2024, paragraph [0025]), and thus the limitations of “obtaining a neural network, the neural network being trained to determine a state change of a physical system over time, training data of the neural network indicating states of a plurality of physical systems at a plurality of times” amount to a mathematical description in words of building a physics-based mathematical model that is trained using data of a plurality of physical systems. The limitations of “determining respective unit feature representations of the plurality of physical units in the target physical system based at least on target values of respective material properties of the plurality of physical units” amounts to a mathematical calculation to represent values of respective material properties as unit feature representations, which is calculation/conversion of abstract data from one form to another. The limitations of “determining a state of the target physical system at a second time based on the state data by inputting at least the unit feature representations to the neural network” amount to a mathematical concept of using a function which takes input parameters of the unit feature representations and returns an output parameter of a state of the target physical system at a second time. Patents that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible under § 101 (see Recentive Analytics, Inc. v. Fox Corp., 134 F.4th 1205 (Fed. Cir. 2025)). If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it is still in the mental processes grouping unless the claim limitation cannot practically be performed in the mind. Given broadest reasonable interpretation of a very basic model trained from historical data in relation to known physical states, a person could equivalently calculate target values of respective material properties (represented as a unit feature representation) relating to a output state of a physical system. The neural network itself is not improved, but is merely used as a mathematical tool to perform what can equivalently be done by a person.
Next in step 2A prong 2, the independent claims are analyzed to determine whether there are additional elements or combination of elements that apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception such that it is more than a drafting effort designed to monopolize the exception, in order to integrate the judicial exception into a practical application. These limitations have been identified and underlined above, and are not indicative of integration into a practical application because: (1) the limitations of “obtaining state data corresponding to a state of a target physical system at a first time, the state data indicating a plurality of physical units comprised in the target physical system, material properties of the plurality of physical units, and interaction relationships between the plurality of physical units” amount to adding insignificant extra-solution data gathering activity to the judicial exception (see MPEP 2106.05(g)); and (2) the limitations of “an electronic device, comprising: at least one processing unit; and at least one memory coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit, the instructions, when executed by the at least one processing unit, causing the device to perform the following actions:”, and “a non-transitory computer-readable storage medium having a computer program stored thereon, the computer program, when executed by a processing unit, causing the processing unit to perform the following actions:” amount to mere instructions to implement an abstract idea on a computer or merely using a computer as a tool to perform an abstract idea (see MPEP 2106.05(f)).
Next in step 2B, the independent claims are considered to determine if they recite additional elements that amount to an inventive concept (“significantly more”) than the recited judicial exception.
The limitations of “obtaining state data corresponding to a state of a target physical system at a first time, the state data indicating a plurality of physical units comprised in the target physical system, material properties of the plurality of physical units, and interaction relationships between the plurality of physical units” do not add significantly more because such limitations amount to adding insignificant extra-solution data gathering activity to the judicial exception (see MPEP 2106.05(g)), and do not describe any gathering of data in an unconventional way. The limitations of “an electronic device, comprising: at least one processing unit; and at least one memory coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit, the instructions, when executed by the at least one processing unit, causing the device to perform the following actions:”, and “a non-transitory computer-readable storage medium having a computer program stored thereon, the computer program, when executed by a processing unit, causing the processing unit to perform the following actions:” do not add significantly more because such limitations amount to mere instructions to implement an abstract idea on a computer or merely using a computer as a tool to perform an abstract idea (see MPEP 2106.05(f)). The use of generic computer equipment is considered insignificant additional elements. As recited in the MPEP, 2106.07(b), merely adding a generic computer, generic computer components, or a programmed computer to perform generic computer functions does not automatically overcome an eligibility rejection (see Alice Corp. Pty. Ltd. v. CLS Bank Int'l, 134 S. Ct. 2347, 2359-60, 110 USPQ2d 1976, 1984 (2014). See also OIP Techs. v. Amazon.com, 788 F.3d 1359, 1364, 115 USPQ2d 1090, 1093-94).
Dependent claims 2-7, 9-10, 12-17, and 19 contain additional limitations that fall under the abstract idea grouping of mathematical concepts to describe additional details for calculating variables including unit feature representations, interpolation weights, a state of the physical target system, message feature representations, locations, graph data, etc. Dependent claims 8 and 18 describe the physical units comprising at least one boundary physical unit at a boundary of the physical system wherein a location of the at least one boundary physical unit remains unchanged when determining the locations of the plurality of physical units at the second time, but this amounts to generally linking the use of the judicial exception to a particular technological environment or field of use (without applying any improvement or change in operation to any physical technology), and is not indicative of integration into a practical application or something significantly more than the recited judicial exception (see MPEP 2016.05(h)).
4. An invention is not rendered ineligible for patent simply because it involves an abstract concept. Applications of such concepts "to a new and useful end" remain eligible for patent protection (see Alice Corp., 134 S. Ct. at 2354 (quoting Benson, 409 U.S. at 67)). However, "a claim for a new abstract idea is still an abstract idea" (see Synopsys v. Mentor Graphics Corp. _F.3d_, 120 U.S.P.Q. 2d1473 (Fed. Cir. 2016)). There needs to be additional elements or combination of additional elements in the claim to apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception or render the claim as a whole to be significantly more than the exception itself in order to demonstrate “integration into a practical application” or an “inventive concept.” For instance, particular physical arrangements for actively obtaining the input state data, or further physical applications using the calculated state of the target physical system at a second time to drive a transformation, change in physical operation, or repair/maintenance of a technology or technical process could provide integration into a practical application to demonstrate an improvement to the technology or technical field.
Claim Rejections - 35 USC § 102
5. In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
6. Claims 1, 4-5, 7-9, 11, 14-15, 17-19, and 22 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Riedmiller et al. (US Pat. Pub. 2021/0049467, hereinafter “Riedmiller”).
In regards to claim 1, Riedmiller teaches a method for state prediction (Riedmiller abstract and paragraph [0007] teach a method using a graph neural network for understanding a physical system, including inferring a static property or dynamic state of the physical system), comprising:
obtaining a neural network (Riedmiller abstract and paragraphs [0005] and [0007] teach obtaining a graph processing neural network), the neural network being trained to determine a state change of a physical system over time (Riedmiller paragraph [0007] teaches training the graph processing neural network to process an input graph to provide an output graph inferring a dynamic state of the physical system, and Riedmiller paragraph [0017] teaches where the output graph node features may define a change from the input, and a forward prediction made by the system may comprise a prediction for a single time step or a rollout prediction over multiple time steps), training data of the neural network indicating states of a plurality of physical systems at a plurality of times (Riedmiller paragraph [0029] teaches where the training data include randomly selected sequence (plurality of times) of observations of the physical system in combination with other systems at a different time step and subsequent time step; Riedmiller paragraph [0087] teaches where the training data may be captured from multiple different physical systems);
obtaining state data corresponding to a state of a target physical system at a first time (Riedmiller paragraphs [0007]-[0008] teach accepting a first input graph corresponding to a state of a target system at a first time), the state data indicating a plurality of physical units comprised in the target physical system, material properties of the plurality of physical units, and interaction relationships between the plurality of physical units (Riedmiller paragraph [0011] teaches where the input graphs (as the state data) are defined by nodes and edges of the physical system to represent the bodies (physical units) and joints in the target system; paragraph [0013] teaches where the input graph edges have directions indicating an expected physical influence (interaction relationship) of one body on another, and paragraph [0016] teaches where the input graph state data also comprises static graph data representing the static properties (material properties- see also paragraph [0020]) of the bodies and joints of the physical system);
determining respective unit feature representations of the plurality of physical units in the target physical system based at least on target values of respective material properties of the plurality of physical units (Riedmiller paragraph [0014] teaches determining respective unit feature representations of each node of the plurality of physical units based on aggregating the values of the edge features with the node features for each input graph node of the physical system, where paragraph [0008] teaches that the node and edge features are defined by feature vectors (and thus the combination of nodes and edges amount to a feature representation)); and
determining a state of the target physical system at a second time based on the state data by inputting at least the unit feature representations to the neural network (Riedmiller paragraph [0014] teaches providing the aggregated feature representation as input to the neural network to determine the output graph node features for the node; Riedmiller paragraph [0017] teaches where the output graph node features may define a predicted future (at a second time) dynamic state of the system given a current dynamic state of the system).
In regards to claim 4, Riedmiller teaches the method wherein determining the respective unit feature representations of the plurality of physical units (Riedmiller paragraph [0014] teaches determining respective unit feature representations of each node of the plurality of physical units based on aggregating the values of the edge features with the node features for each input graph node of the physical system, where paragraph [0008] teaches that the node and edge features are defined by feature vectors (and thus the combination of nodes and edges amount to a feature representation)) further comprises:
determining the respective unit feature representations of the plurality of physical units based at least on one of the following: respective velocities of the plurality of physical units at the first time (Riedmiller paragraph [0007] teaches where the input graph has input graph nodes comprising input graph node features representing dynamic data, and paragraph [0009] teaches where the dynamic data includes velocity data for each body (physical unit); paragraph [0027] teaches receiving dynamic data for a sequence of time steps including a first time), and an external force applied to the plurality of physical units respectively at the first time (Riedmiller paragraph [0007] teaches where the input graph has input graph edges comprising input graph edge features representing action data, and paragraph [0009] teaches where the action data comprises linear or angular force data associated with action at a joint (applied to the plurality of physical units); paragraph [0027] teaches receiving action data for a sequence of time steps including a first time).
In regards to claim 5, Riedmiller teaches the method wherein the determining a state of the target physical system at a second time based on the state data by inputting at least the unit feature representations to the neural network (Riedmiller paragraph [0014] teaches providing the aggregated feature representation as input to the neural network to determine the output graph node features for the node; Riedmiller paragraph [0017] teaches where the output graph node features may define a predicted future (at a second time) dynamic state of the system given a current dynamic state of the system) comprise:
determining respective relationship feature representations of the interaction relationships between the plurality of physical units, each relationship feature representation being determined based at least on relative locations of a pair of physical units having an interaction relationship at the first time (Riedmiller paragraph [0056] teaches where the edges of the input graph amount to relationship feature representations to represent joints or other interactions between the bodies (physical units); Riedmiller paragraph [0103] teaches where an edge feature may indicate the relative distance and/or direction of a node to a target node (i.e., based on relative locations of a pair of physical units)); and
determining a state of the target physical system at a second time based on the state data by inputting the unit feature representations and the respective relationship feature representations to the neural network (Riedmiller paragraphs [0014] and [0017] teaches determining an output graph node predicted future state (at a second time) of the target system based on inputting the unit feature representations and relationship edge feature representations to the graph neural network at a current dynamic state of the system).
In regards to claim 7, Riedmiller teaches the method wherein determining the state of the target physical system at the second time (Riedmiller paragraph [0017] teaches determining a future dynamic state of the physical system at a second time) comprises:
determining respective locations of the plurality of physical units in the target physical system at the second time (Riedmiller paragraph [0017] teaches determining output graph node features that define dynamic data as a change from the input to the second predicted future dynamic state time; Riedmiller paragraph [0062] defines dynamic data as including Cartesian position (location) for each of the bodies (physical units) in the target system and thus any changes in dynamic data would include changes in location).
In regards to claim 8, Riedmiller teaches the method, wherein the plurality of physical units comprise at least one boundary physical unit at a boundary of the physical system (Riedmiller Fig. 1a teaches the physical units comprising at least one boundary physical unit (see e.g., nodes at distant ends) of the physical system 102), wherein a location of the at least one boundary physical unit remains unchanged when determining the locations of the plurality of physical units at the second time (Riedmiller paragraph [0009] teaches where some bodies (including boundary units) may not provide dynamic data because they are stationary (i.e., location remains unchanged)).
In regards to claim 9, Riedmiller teaches the method wherein the state data comprises graph data with a plurality of nodes and a plurality of directed edges between the plurality of nodes, the plurality of nodes characterizing the plurality of physical units in the target physical system respectively, and the plurality of edges characterizing the interaction relationships between the plurality of physical units respectively (Riedmiller paragraphs [0007] and [0056] teach wherein the state data comprises graph data comprising nodes and edges, the nodes corresponding to the bodies (physical units) of the physical system and the edges corresponding to the joints or other interactions between the bodies (physical units)).
In regards to claim 11, Riedmiller teaches an electronic device (Riedmiller paragraph [0005] teaches a computer as an electronic device), comprising:
at least one processing unit (Riedmiller paragraph [0110] teaches at least one processor);
and at least one memory coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit, the instructions, when executed by the at least one processing unit, causing the device to perform the following actions (Riedmiller paragraphs [0114]-[0115] teach at least one memory for storing program instructions for execution by the at least one processing unit):
obtaining a neural network (Riedmiller abstract and paragraphs [0005] and [0007] teach obtaining a graph processing neural network), the neural network being trained to determine a state change of a physical system over time (Riedmiller paragraph [0007] teaches training the graph processing neural network to process an input graph to provide an output graph inferring a dynamic state of the physical system, and Riedmiller paragraph [0017] teaches where the output graph node features may define a change from the input, and a forward prediction made by the system may comprise a prediction for a single time step or a rollout prediction over multiple time steps), training data of the neural network indicating states of a plurality of physical systems at a plurality of times (Riedmiller paragraph [0029] teaches where the training data include randomly selected sequence (plurality of times) of observations of the physical system in combination with other systems at a different time step and subsequent time step; Riedmiller paragraph [0087] teaches where the training data may be captured from multiple different physical systems);
obtaining state data corresponding to a state of a target physical system at a first time (Riedmiller paragraphs [0007]-[0008] teach accepting a first input graph corresponding to a state of a target system at a first time), the state data indicating a plurality of physical units comprised in the target physical system, material properties of the plurality of physical units, and interaction relationships between the plurality of physical units (Riedmiller paragraph [0011] teaches where the input graphs (as the state data) are defined by nodes and edges of the physical system to represent the bodies (physical units) and joints in the target system; paragraph [0013] teaches where the input graph edges have directions indicating an expected physical influence (interaction relationship) of one body on another, and paragraph [0016] teaches where the input graph state data also comprises static graph data representing the static properties (material properties- see also paragraph [0020]) of the bodies and joints of the physical system);
determining respective unit feature representations of the plurality of physical units in the target physical system based at least on target values of respective material properties of the plurality of physical units (Riedmiller paragraph [0014] teaches determining respective unit feature representations of each node of the plurality of physical units based on aggregating the values of the edge features with the node features for each input graph node of the physical system, where paragraph [0008] teaches that the node and edge features are defined by feature vectors (and thus the combination of nodes and edges amount to a feature representation)); and
determining a state of the target physical system at a second time based on the state data by inputting at least the unit feature representations to the neural network (Riedmiller paragraph [0014] teaches providing the aggregated feature representation as input to the neural network to determine the output graph node features for the node; Riedmiller paragraph [0017] teaches where the output graph node features may define a predicted future (at a second time) dynamic state of the system given a current dynamic state of the system).
In regards to claim 14, Riedmiller teaches the device wherein determining the respective unit feature representations of the plurality of physical units (Riedmiller paragraph [0014] teaches determining respective unit feature representations of each node of the plurality of physical units based on aggregating the values of the edge features with the node features for each input graph node of the physical system, where paragraph [0008] teaches that the node and edge features are defined by feature vectors (and thus the combination of nodes and edges amount to a feature representation)) further comprises:
determining the respective unit feature representations of the plurality of physical units based at least on one of: respective velocities the plurality of physical units at the first time (Riedmiller paragraph [0007] teaches where the input graph has input graph nodes comprising input graph node features representing dynamic data, and paragraph [0009] teaches where the dynamic data includes velocity data for each body (physical unit); paragraph [0027] teaches receiving dynamic data for a sequence of time steps including a first time), and an external force applied to the plurality of physical units respectively at the first time (Riedmiller paragraph [0007] teaches where the input graph has input graph edges comprising input graph edge features representing action data, and paragraph [0009] teaches where the action data comprises linear or angular force data associated with action at a joint (applied to the plurality of physical units); paragraph [0027] teaches receiving action data for a sequence of time steps including a first time).
In regards to claim 15, Riedmiller teaches the device wherein the determining a state of the target physical system at a second time based on the state data by inputting at least the unit feature representations to the neural network (Riedmiller paragraph [0014] teaches providing the aggregated feature representation as input to the neural network to determine the output graph node features for the node; Riedmiller paragraph [0017] teaches where the output graph node features may define a predicted future (at a second time) dynamic state of the system given a current dynamic state of the system) comprises:
determining respective relationship feature representations of the interaction relationships between the plurality of physical units, each relationship feature representation being determined based at least on relative locations of a pair of physical units having an interaction relationship at the first time (Riedmiller paragraph [0056] teaches where the edges of the input graph amount to relationship feature representations to represent joints or other interactions between the bodies (physical units); Riedmiller paragraph [0103] teaches where an edge feature may indicate the relative distance and/or direction of a node to a target node (i.e., based on relative locations of a pair of physical units)); and
determining a state of the target physical system at a second time based on the state data by inputting the unit feature representations and the respective relationship feature representations to the neural network (Riedmiller paragraphs [0014] and [0017] teach determining an output graph node predicted future state (at a second time) of the target system based on inputting the unit feature representations and relationship edge feature representations to the graph neural network at a current dynamic state of the system).
In regards to claim 17, Riedmiller teaches the device wherein determining the state of the target physical system at the second time (Riedmiller paragraph [0017] teaches determining a future dynamic state of the physical system at a second time) comprises: determining respective locations of the plurality of physical units in the target physical system at the second time (Riedmiller paragraph [0017] teaches determining output graph node features that define dynamic data as a change from the input to the second predicted future dynamic state time; Riedmiller paragraph [0062] defines dynamic data as including Cartesian position (location) for each of the bodies (physical units) in the target system, and thus any changes in dynamic data would include changes in location).
In regards to claim 18, Riedmiller teaches the device wherein the plurality of physical units comprise at least one boundary physical unit at a boundary of the physical system (Riedmiller Fig. 1a teaches the physical units comprising at least one boundary physical unit (see e.g., nodes at distant ends) of the physical system 102), wherein a location of the at least one boundary physical unit remains unchanged when determining the locations of the plurality of physical units at the second time (Riedmiller paragraph [0009] teaches where some bodies (including boundary units) may not provide dynamic data because they are stationary (i.e., location remains unchanged)).
In regards to claim 19, Riedmiller teaches the device wherein the state data comprises graph data with a plurality of nodes and a plurality of directed edges between the plurality of nodes, the plurality of nodes characterizing the plurality of physical units in the target physical system respectively, and the plurality of edges characterizing the interaction relationships between the plurality of physical units respectively (Riedmiller paragraphs [0007] and [0056] teach wherein the state data comprises graph data comprising nodes and edges, the nodes corresponding to the bodies (physical units) of the physical system and the edges corresponding to the joints or other interactions between the bodies (physical units)).
In regards to claim 22, Riedmiller teaches a non-transitory computer-readable storage medium having a computer program stored thereon, the computer program, when executed by a processing unit, causing the processing unit to perform the following actions (Riedmiller paragraphs [0114]-[0115] teach at least one non-transitory computer readable medium storing a computer program for execution by the at least one processing unit):
obtaining a neural network (Riedmiller abstract and paragraphs [0005] and [0007] teach obtaining a graph processing neural network), the neural network being trained to determine a state change of a physical system over time (Riedmiller paragraph [0007] teaches training the graph processing neural network to process an input graph to provide an output graph inferring a dynamic state of the physical system, and Riedmiller paragraph [0017] teaches where the output graph node features may define a change from the input, and a forward prediction made by the system may comprise a prediction for a single time step or a rollout prediction over multiple time steps), training data of the neural network indicating states of a plurality of physical systems at a plurality of times (Riedmiller paragraph [0029] teaches where the training data include randomly selected sequence (plurality of times) of observations of the physical system in combination with other systems at a different time step and subsequent time step; Riedmiller paragraph [0087] teaches where the training data may be captured from multiple different physical systems);
obtaining state data corresponding to a state of a target physical system at a first time (Riedmiller paragraphs [0007]-[0008] teach accepting a first input graph corresponding to a state of a target system at a first time), the state data indicating a plurality of physical units comprised in the target physical system, material properties of the plurality of physical units, and interaction relationships between the plurality of physical units (Riedmiller paragraph [0011] teaches where the input graphs (as the state data) are defined by nodes and edges of the physical system to represent the bodies (physical units) and joints in the target system; paragraph [0013] teaches where the input graph edges have directions indicating an expected physical influence (interaction relationship) of one body on another, and paragraph [0016] teaches where the input graph state data also comprises static graph data representing the static properties (material properties- see also paragraph [0020]) of the bodies and joints of the physical system);
determining respective unit feature representations of the plurality of physical units in the target physical system based at least on target values of respective material properties of the plurality of physical units (Riedmiller paragraph [0014] teaches determining respective unit feature representations of each node of the plurality of physical units based on aggregating the values of the edge features with the node features for each input graph node of the physical system, where paragraph [0008] teaches that the node and edge features are defined by feature vectors (and thus the combination of nodes and edges amount to a feature representation)); and
determining a state of the target physical system at a second time based on the state data by inputting at least the unit feature representations to the neural network (Riedmiller paragraph [0014] teaches providing the aggregated feature representation as input to the neural network to determine the output graph node features for the node; Riedmiller paragraph [0017] teaches where the output graph node features may define a predicted future (at a second time) dynamic state of the system given a current dynamic state of the system).
Allowable Subject Matter
7. Claims 2-3, 6, 10, and 12-13 would be allowable if rewritten to overcome the rejection(s) under 35 U.S.C. 101, set forth in this Office action and to include all of the limitations of the base claim and any intervening claims.
The following is a statement of reasons for the indication of allowable subject matter: 8. Claim 2 contains allowable subject matter because the closest art, Riedmiller et al. (US Pat. Pub. 2021/0049467) fails to anticipate or render obvious the method wherein determining the respective unit feature representations of the plurality of physical units comprises: for a given physical unit among the plurality of physical units, determining a plurality of values of a physical unit having a same material property as the physical unit in the plurality of physical systems; and in response to the target value of the material property of the given physical unit falling between a first value and a second value of the plurality of values, determining a unit feature representation of the given physical unit based at least on a first feature representation corresponding to the first value of the material property and a second feature representation corresponding to the second value of the material property, in combination with the rest of the claim limitations as claimed and defined by the Applicant.
Claim 6 contains allowable subject matter because the closest prior art, Riedmiller et al. (US Pat. Pub. 2021/0049467) fails to anticipate or render obvious the method wherein determining the state of the target physical system at the second time comprises: determining, by the neural network, a first message feature representation from a first physical unit to a second physical unit among the plurality of physical units, the first message feature representation characterizing an effect of the first physical unit on the second physical unit; determining a negative value of the first message feature representation as a second message feature representation from the second physical unit to the first physical unit, the first message feature representation characterizing an effect of the second physical unit on the first physical unit; in combination with the rest of the claim limitations as claimed and defined by the Applicant.
Claim 10 contains allowable subject matter because the closest prior art, Riedmiller et al. (US Pat. Pub. 2021/0049467) fails to anticipate or render obvious the method wherein the neural network is trained for a particle-based discretized physical system, and the graph data comprises data based on a dynamics nearest neighbor graph, and wherein the neural network is trained for a grid-based discretized physical system, and the graph data comprises data based on a static multi-scale grid graph, in combination with the rest of the claim limitations as claimed and defined by the Applicant.
Claim 12 contains allowable subject matter because the closest art, Riedmiller et al. (US Pat. Pub. 2021/0049467) fails to anticipate or render obvious the device wherein determining the respective unit feature representations of the plurality of physical units comprises: for a given physical unit among the plurality of physical units, determining a plurality of values of a physical unit having the same material property as the physical unit in the plurality of physical systems; and in response to the target value of the material property of the given physical unit falling between a first value and a second value of the plurality of values, determining a unit feature representation of the given physical unit based at least on a first feature representation corresponding to the first value of the material property and a second feature representation corresponding to the second value of the material property, in combination with the rest of the claim limitations as claimed and defined by the Applicant.
Claim 16 contains allowable subject matter because the closest prior art, Riedmiller et al. (US Pat. Pub. 2021/0049467) fails to anticipate or render obvious the device wherein determining the state of the target physical system at the second time comprises: determining, by the neural network, a first message feature representation from a first physical unit to a second physical unit among the plurality of physical units, the first message feature representation characterizing an effect of the first physical unit on the second physical unit; determining a negative value of the first message feature representation as a second message feature representation from the second physical unit to the first physical unit, the first message feature representation characterizing an effect of the second physical unit on the first physical unit; in combination with the rest of the claim limitations as claimed and defined by the Applicant.
9. Dependent claim 3 depends from claim 2 and contains allowable subject matter for at least the same reasons as given for claim 2. Dependent claim 13 depends from claim 12 and contains allowable subject matter for at least the same reasons as given for claim 12.
Pertinent Art
10. Applicants are directed to consider additional pertinent prior art included on the Notice of References Cited (PTOL 892) attached herewith. The Examiner has pointed out particular references contained in the prior art of record within the body of this action for the convenience of the Applicant. Although the specified citations are representative of the teachings in the art and are applied to the specific limitations within the individual claim, other passages and figures may apply. Applicant, in preparing the response, should consider fully the entire reference as potentially teaching all or part of the claimed invention, as well as the context of the of the passage as taught by the prior art or disclosed by the Examiner. The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
B. Natsumeda et al. (US Pat. Pub. 2022/0004182) discloses Approach to Determining a Remaining Useful Life of a System.
C. Mnih et al. (US Pat. Pub. 2015/0100530) discloses Methods and Apparatus for Reinforcement Learning.
D. Li et al. (US Pat. No. 11,093,819) discloses Classifying Objects Using Recurrent Neural Network and Classifier Neural Network Subsystems.
Conclusion
11. Any inquiry concerning this communication or earlier communications from the examiner should be directed to PAUL D LEE whose telephone number is (571)270-1598. The examiner can normally be reached on M to F, 9:30 am to 6 pm.
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, Arleen Vazquez can be reached at 571-272-2619. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see https://ppair-my.uspto.gov/pair/PrivatePair. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/PAUL D LEE/Primary Examiner, Art Unit 2857 10/9/2026
Prosecution Insights