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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 02/19/2026 has been entered.
Response to Remarks
Claim Rejections – 35 U.S.C. 103
Applicant’s amendments have been fully considered and they are persuasive.
Applicant argues (pg. 9-12) that the cited references do not teach the amended limitations that further clarify that determining the combined objective-defining value comprises determining a linear combination of the first machine learning objective-defining value and the machine learning objective-defining values received from the other machine learning systems each weighted by a respective one of the mixing parameters and that the efficiency estimate is to adjust how the machine learning objective-defining values received from the other machine learning systems are weighted in the linear combination.
Examiner agrees. Accordingly, a new reference, Jaques et al. (“Social Influence as Intrinsic Motivation for Multi-Agent Deep Reinforcement Learning”) has been added to the rejection, as further detailed below.
The foregoing applies to all independent claims and their dependent claims.
Claim Rejections – 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-3, 10-11, 13-21, 23 are rejected under 35 U.S.C. 103 as being unpatentable over Wang et al. (“Integrating Reinforcement Learning with Multi-Agent Techniques for Adaptive Service Composition”) hereinafter known as Wang in view of Czarnecki et al. (US20190354867A1) hereinafter known as Czarnecki in view of Jaques et al. (“Social Influence as Intrinsic Motivation for Multi-Agent Deep Reinforcement Learning”) hereinafter known as Jaques.
Regarding independent claim 1, Wang teaches:
A computer-implemented method of training a first machine learning system to select actions to be performed by a first agent of a group of agents to control the first agent to perform a task in an environment, wherein whilst performing the task the first agent interacts with one or more other agents of a group of agents in the environment respectively controlled by one or more other machine learning systems to perform one or more other tasks, the method comprising: (Wang [Page 1, Paragraph 1]: “enable agents to work for a common task (i.e., composition)” Wang teaches that the agents are in the shared environment to perform a common task. Wang [Page 8:31, Paragraph 2]: “With the increase of agents, the complexity of multi-agent and the communication consumption between agents in the process of fictitious play increase, too. When the agents increase to a certain number, the acceleration of multi-agent will be balanced off by the complexity of multi-agent and the communication consumption.” Wang teaches that the agents interact with each other by communicating with one another. Such actions are selected to be performed by the agents.)
receiving, from each of the other machine learning systems, a respective machine learning objective-defining value used for training the other machine learning system, wherein the respective machine learning objective-defining value is a respective reward value received by the respective other agent from the environment or a respective value of a loss function that is used to define a training objective for training the respective machine learning system; (Wang [Page 8:20, Paragraph 1]: “Each agent is aware of the states of other agents through the reward and updates their own rewards after execution.” Wang teaches that an agent receives the state of other agents, which has the objective-defining value. Using this, the agent updates their own rewards, for subsequent training.)
determining a combined objective-defining value from a combination of a first machine learning objective-defining value for the first machine learning system and the machine learning objective-defining values received from the other machine learning systems, wherein the first machine learning objective-defining value is a reward value received by the first agent from the environment or a value of a loss function that is used to define a training objective for training the first machine learning system, wherein the combination is defined by a set of mixing parameters, … (Wang [Page 8:8, Paragraph 2]: “agents construct different composite service results on the basis of different collaboration schemes and various component services” Wang teaches that from the various different collaboration schemes and component services from the different agents, agents construct a composite objective-defining value. Based on the importance of the different schemes, the composite scheme shows that the agents will have a set of parameters that correspond to the importance.)
training the first machine learning system using the combined objective-defining value; (Wang [Page 8:25, Paragraph 3]: “Later, the WEQ function should be combined into the learning policy for the Team Markov Game to select appropriate behavior at every step of the learning process.” Wang teaches that the combined objective defining function, which is the learning policy for the Team Markov Game, is used as the learning policy during training.)
…
after the training the first machine learning system using the combined objective-defining value and the adjusting the set of mixing parameters, controlling the first machine learning system using the first agent of the first machine learning system to perform one or more actions; (Wang [Page 8:8, Paragraph 2]: “agents construct different composite service results on the basis of different collaboration schemes and various component services” Wang teaches that from the various different collaboration schemes and component services from the different agents, agents construct a composite objective-defining value. Based on the importance of the different schemes, the composite scheme shows that the agents will have a set of parameters that correspond to the importance.)
receiving a new reward received by the first agent from the environment or a new loss from the first agent performing the one or more actions; (Wang [Page 8:20, Paragraph 1]: “Each agent is aware of the states of other agents through the reward and updates their own rewards after execution.” Wang teaches that the states of the various agents comprise of the reward that corresponds to each of the agents.)
receiving, from the each of the other machine learning systems, a respective new reward received by the respective other agent from the environment or a new loss from the respective other agent performing one or more other actions; (Wang [Page 8:20, Paragraph 1]: “Each agent is aware of the states of other agents through the reward and updates their own rewards after execution.” Wang teaches that the states of the various agents comprise of the reward that corresponds to each of the agents.)
determining a new combined objective-defining value using the adjusted set of mixing parameters; (Wang [Page 8:8, Paragraph 2]: “agents construct different composite service results on the basis of different collaboration schemes and various component services” Wang teaches that from the various different collaboration schemes and component services from the different agents, agents construct a composite objective-defining value. Based on the importance of the different schemes, the composite scheme shows that the agents will have a set of parameters that correspond to the importance.)
and training the first machine learning system on the new combined objective-defining value. (Wang [Page 8:25, Paragraph 3]: “Later, the WEQ function should be combined into the learning policy for the Team Markov Game to select appropriate behavior at every step of the learning process.” Wang teaches that the combined objective defining function, which is the learning policy for the Team Markov Game, is used as the learning policy during training.)
Wang does not explicitly teach:
adjusting the set of mixing parameters using gradient descent to optimize an efficiency estimate to adjust how the machine learning objective-defining values received from the other machine learning systems are weighted in the linear combination, wherein the efficiency estimate is dependent upon a rate of change of the combined objective value with time.
However, Czarnecki teaches:
adjusting the set of mixing parameters using gradient descent to optimize an efficiency estimate to adjust how the machine learning objective-defining values received from the other machine learning systems are weighted in the linear combination, wherein the efficiency estimate is dependent upon a rate of change of the combined objective value with time. (Czarnecki [¶ 0090]: “To train the neural network, the system computes gradients of a reinforcement learning loss function … that is appropriate for the kinds of network outputs that the policy networks are configured to generate and that encourages the combined policies to show improved performance on the reinforcement learning task” Czarnecki teaches that the gradients are computed such that the training can converge toward a combined policy with improved performance on the reinforcement learning task. This shows that the weights to each of the agents’ policies are changed during the gradient descent.)
Wang and Czarnecki are in the same field of endeavor as the present invention, as the
references are directed to using reinforcement learning to combine collaboration schemes. It would have been obvious, before the effective filing date of the claimed invention, to a person of ordinary skill in the art, to combine the composition of multiple different objective functions from different agents as taught in Wang with using gradient descent during training to converge toward a combined policy with improved policy as taught in Czarnecki. Czarnecki provides this additional functionality. As such, it would have been obvious to one of ordinary skill in the art to modify the teachings of Wang to include teachings of Czarnecki because the combination would allow for the agents to collaborate on their objective function and policy while efficiently training by gradient descent using this combined policy. This has the potential benefit of increasing the efficiency of the reinforcement learning of the agents, as the path that is orthogonal to the path of least efficiency can be taken to maximize the efficiency.
Wang and Czarnecki do not explicitly teach:
… wherein determining the combined objective-defining value comprises determining a linear combination of the first machine learning objective-defining value and the machine learning objective-defining values received from the other machine learning systems each weighted by a respective one of the mixing parameters;
However, Jaques teaches:
… wherein determining the combined objective-defining value comprises determining a linear combination of the first machine learning objective-defining value and the machine learning objective-defining values received from the other machine learning systems each weighted by a respective one of the mixing parameters; (Jaques [Page 9, Col. 2, Paragraph 1]: “we implemented agents that optimize for a convex combination of their own individual reward e_{t}^{k} and the collective reward of all other agents, sum_{i}^{N} e_{t}^{i}, where i != k. Thus, the reward function for agent k is r_{t}^{k} = e_{t}^{k} + eta sum_{i}^{N} e_{t}^{i}, where i != k” Jaques teaches that for each of the agents, the objective depends on the objective defining value for the other agents, as can be seen in the summation of the influences from the other agents. The mixing parameters is baked into the objective value from the other agents but is also scaled using the eta parameter.)
Jaques is in the same field as the present invention, since it is directed to a multi-agent system in reinforcement learning. It would have been obvious, before the effective filing date of the claimed invention, to a person of ordinary skill in the art, to combine collaboration of agents on their objective function and policy while efficiently training by gradient descent using this combined policy as taught in Wang as modified by Czarnecki with using a linear combination of objective defining values from the other agents as taught in Jaques. Jaques provides this additional functionality. As such, it would have been obvious to one of ordinary skill in the art to modify the teachings of Wang as modified by Czarnecki to include teachings of Jaques because the combination would allow for the objectives of the other agents to be weighted in when defining the objective for a particular agent. This has the potential benefit of an agent behaving in the best interest of the overall problem at hand in conjunction with the other agents.
Regarding dependent claim 2, Wang and Czarnecki teach:
A method as claimed in claim 1,
Czarnecki teaches:
wherein the efficiency estimate comprises a cost which has a higher value when the combined objective-defining value is worsening with time than when the combined objective-defining value is improving with time. (Czarnecki [¶ 0046]: “For example, a return may refer to a long-term time-discounted reward received by the system. In some of these implementations, the system 100 can select the action that has the highest return as the action to be performed or can apply an epsilon-greedy action selection policy.” Czarnecki teaches that the system can select an action with the highest return. Since the return is the long-term time-discounted reward, the cost is higher when the objective-defining value is worsening with time. This is because the return is lower with the time discount and the cost is higher.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 3, Wang and Czarnecki teach:
A method as claimed in claim 1,
Czarnecki teaches:
wherein adjusting the set of mixing parameters using gradient descent comprises determining a set of gradients of the efficiency estimate with respect to the set of mixing parameters, and adjusting the set of mixing parameters using the set of gradients. (Czarnecki [¶ 0090]: “To train the neural network, the system computes gradients of a reinforcement learning loss function … that is appropriate for the kinds of network outputs that the policy networks are configured to generate and that encourages the combined policies to show improved performance on the reinforcement learning task” Czarnecki teaches that the gradients are computed such that the training can converge toward a combined policy with improved performance on the reinforcement learning task. This shows that the weights to each of the agents’ policies are changed during the gradient descent.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 10, Wang and Czarnecki teach:
A method as claimed in claim 1,
Wang teaches:
further comprising sending the first machine learning objective-defining value to each of the one or more other machine learning systems. (Wang [Page 8:20, Paragraph 1]: “Each agent is aware of the states of other agents through the reward and updates their own rewards after execution.” Wang teaches that an agent sends its state to other agents, which has the objective-defining value. Using this, the agents update their own rewards, for subsequent training.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 11, Wang and Czarnecki teach:
A method as claimed in claim 1,
Wang teaches:
wherein the first machine learning system includes a policy neural network to receive observations of the environment and to select the actions to be performed by the first agent in response to the observations, wherein the policy neural network has a plurality of policy neural network parameters, … (Wang [Page 8:20, Paragraph 1]: “Each agent is aware of the states of other agents through the reward and updates their own rewards after execution.” Wang teaches that an agent sends its state to other agents, which has the objective-defining value. Using this, the agents update their own rewards, for subsequent training.)
Czarnecki teaches:
… and wherein training the first machine learning system comprises adjusting the policy neural network parameters using gradient descent to optimize an objective defined by the combined objective-defining value. (Czarnecki [¶ 0090]: “To train the neural network, the system computes gradients of a reinforcement learning loss function … that is appropriate for the kinds of network outputs that the policy networks are configured to generate and that encourages the combined policies to show improved performance on the reinforcement learning task” Czarnecki teaches that the gradients are computed such that the training can converge toward a combined policy with improved performance on the reinforcement learning task. This shows that the weights to each of the agents’ policies are changed during the gradient descent.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 12, Wang and Czarnecki teach:
A method as claimed in claim 1,
Wang teaches:
wherein the first machine learning objective-defining value and the machine learning objective-defining values from the other machine learning systems each comprise the value of a loss function dependent on, or the value of a reward received in response to, respectively, an action of the first agent and an action each of the one or more other agents. (Wang [Page 8:20, Paragraph 1]: “Each agent is aware of the states of other agents through the reward and updates their own rewards after execution.” Wang teaches that the states of the various agents comprise of the reward that corresponds to each of the agents.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 13, Wang and Czarnecki teach:
A method as claimed in claim 1,
Czarnecki teaches:
wherein the efficiency estimate is component of a cost of inefficiency of the group of agents performing their respective tasks in the environment. (Czarnecki [¶ 0046]: “For example, a return may refer to a long-term time-discounted reward received by the system. In some of these implementations, the system 100 can select the action that has the highest return as the action to be performed or can apply an epsilon-greedy action selection policy.” Czarnecki teaches that the system can select an action with the highest return. Since the return is the long-term time-discounted reward, the cost is higher when the objective-defining value is worsening with time. This is an estimate of inefficiency as a function of the time.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 14, Wang and Czarnecki teach:
A method as claimed in claim 1,
Wang teaches:
wherein the method is also implemented by each of the one or more other machine learning systems. (Wang [Page 8:20, Paragraph 1]: “Each agent is aware of the states of other agents through the reward and updates their own rewards after execution.” Wang teaches that an agent sends its state to other agents, which has the objective-defining value. Using this, the agents update their own rewards, for subsequent training. Thus, this updating is implemented on each of the machine learning systems of the various agents.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 15, Wang and Czarnecki teach:
A method as claimed in claim 1,
Wang teaches:
wherein the first machine learning system is a first reinforcement learning system, and wherein each of the other machine learning systems are reinforcement learning systems. (Wang [Page 8:26, Paragraph 3]: “By reinforcement learning, any tiny changes of the QoS values will be accumulated in the Q-value and thus drive the system to react to it and have an effect on the future action selections for learning episodes” Wang teaches that the systems that the agents use are reinforcement learning systems.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 16, Wang and Czarnecki teach:
A method as claimed in claim 1,
Czarnecki teaches:
wherein each agent of the group of agents comprises a robot or autonomous vehicle, wherein each of the tasks comprises navigating a path through the environment from a start point to an end point, and wherein the first machine learning objective-defining value is dependent on an estimated time or distance for the agent physically to move from the start point to the end point. (Czarnecki [¶ 0030]: “In some implementations, the environment is a real-world environment and the agent is a mechanical agent interacting with the real-world environment, e.g., a robot or an autonomous or semi-autonomous land, air, or sea vehicle navigating through the environment.” Czarnecki teaches that the agents are a system of robots or autonomous vehicles navigating the environment.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 17, Wang and Czarnecki teach:
A method as claimed in claim 1,
Wang teaches:
wherein the environment is a data packet communications network environment, wherein each agent of the group of agents comprises a router to route packets of data over the communications network, and wherein the first machine learning objective-defining value is dependent on a routing metric for a path from the router to a next or further node in the data packet communications network. (Wang [Page 8:8, Paragraph 2]: “agents construct different composite service results on the basis of different collaboration schemes and various component services” Wang teaches that from the various different collaboration schemes and component services from the different agents, agents construct a composite objective-defining value. This necessarily shows that each agent acts like a router in an environment of packets communications.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 18, Wang and Czarnecki teach:
A method as claimed in claim 1,
Wang teaches:
wherein the environment is an electrical power distribution environment, wherein each agent of the group of agents is configured to control routing of electrical power from a node associated with the agent to one or more other nodes over one or more power distribution links, and wherein the first machine learning objective-defining value is dependent on one or both of a loss and a frequency or phase mismatch over the one or more power distribution links. (Wang [Page 8:8, Paragraph 2]: “agents construct different composite service results on the basis of different collaboration schemes and various component services” Wang teaches that from the various different collaboration schemes and component services from the different agents, agents construct a composite objective-defining value. This necessarily shows that each agent acts like a router in an environment of electrical power.)
The reasons to combine are substantially similar to those of claim 1.
Regarding dependent claim 19, Wang and Czarnecki teach:
A method as claimed in claim 1,
Wang teaches:
wherein the environment is a plant or service facility, wherein and each agent of the group of agents is configured to control an item of equipment in the plant or service facility, and wherein the first machine learning objective-defining value is dependent on resource usage by the plant or service facility. (Wang [Page 8:8, Paragraph 2]: “agents construct different composite service results on the basis of different collaboration schemes and various component services” Wang teaches that from the various different collaboration schemes and component services from the different agents, agents construct a composite objective-defining value. This necessarily shows that each agent acts like a controller in a service facility.)
The reasons to combine are substantially similar to those of claim 1.
Claim 20 is substantially similar to claim 1, but has the following additional elements:
Regarding independent claim 20, Czarnecki teaches:
One or more non-transitory computer-readable storage media storing instructions that when executed by one or more computers cause the one or more computers to perform the operations (Czarnecki [¶ 0102]: “one or more modules of computer program instructions encoded on a tangible non transitory storage medium for execution by, or to control the operation of, data processing apparatus” Czarnecki teaches non-transitory storage mediums that store instructions to control the operation of the data processing apparatus.)
The reasons to combine are substantially similar to those of claim 1.
Claim 21 is substantially similar to claim 1, but has the following additional elements:
Regarding independent claim 21, Czarnecki teaches:
A system comprising one or more computers and one or more storage devices storing instructions that when executed by one or more computers cause the one or more computers to perform operations (Czarnecki [¶ 0108]: “The essential elements of a computer are a central processing unit for performing or executing instructions and one or more memory devices for storing instructions and data.” Czarnecki teaches computers and memory devices for storing the computer instructions.)
The reasons to combine are substantially similar to those of claim 1.
Claim 23 is rejected on the same grounds under 35 U.S.C. 103 as claim 2 as they are substantially similar. Mutatis mutandis.
Claims 4-8 are rejected under 35 U.S.C. 103 as being unpatentable over Wang in view of Czarnecki in view of Jaques in view of Brownlee (“Use Weight Regularization to Reduce Overfitting of Deep Learning Models”) hereinafter known as Brownlee.
Regarding dependent claim 4, Wang and Czarnecki teach:
A method as claimed in claim 3,
Wang and Czarnecki do not explicitly teach:
wherein determining the set of gradients of the efficiency estimate includes adding a regularization term to the set of gradients to inhibit adjusting the set of mixing parameters away from the first machine learning objective-defining value.
However, Brownlee teaches:
wherein determining the set of gradients of the efficiency estimate includes adding a regularization term to the set of gradients to inhibit adjusting the set of mixing parameters away from the first machine learning objective-defining value. (Brownlee [Page 3, Paragraph 1]: “The addition of a weight size penalty or weight regularization to a neural network has the effect of reducing generalization error and of allowing the model to pay less attention to less relevant input variables.” Brownlee teaches using a regularization to adjust the weights away from less relevant input variables, which may be the first machine learning objective-defining value.)
Brownlee is in the same field as the present invention, since it is directed to using regularization to maintain the different terms in the objective function of a machine learning task. It would have been obvious, before the effective filing date of the claimed invention, to a person of ordinary skill in the art, to combine agents collaborating on their objective function and policy while efficiently training by gradient descent using this combined policy as taught in Wang as modified by Czarnecki with using regularization to make sure that a term is not overpowering as taught in Brownlee. Brownlee provides this additional functionality. As such, it would have been obvious to one of ordinary skill in the art to modify the teachings of Wang as modified by Czarnecki to include teachings of Brownlee because the combination would allow for the objective function to take in terms from different agents without one term being too influential. This has the potential benefit of balancing the objective function such that the goals of the various different agents can be fairly represented during the training.
Regarding dependent claim 5, Wang and Czarnecki teach:
A method as claimed in claim 3,
Brownlee teaches:
wherein determining the set of gradients of the efficiency estimate comprises applying a trial modification to the set of mixing parameters to determine a trial set of mixing parameters, and determining the efficiency estimate using a trial combined objective-defining value for the first machine learning system where the trial combined objective-defining value is defined by the trial set of mixing parameters. (Brownlee [Page 3, Paragraph 11]: “Rather than adding each weight to the penalty directly, they can be weighted using a new hyperparameter called alpha (a) or sometimes lambda. This controls the amount of attention that the learning process should pay to the penalty.” Brownlee teaches that the weights of the objective-defining value can be placed as an initial alpha or lambda value.)
The reasons to combine are substantially similar to those of claim 4.
Regarding dependent claim 6, Wang, Czarnecki, and Brownlee teach:
A method as claimed in claim 5,
Brownlee teaches:
wherein the trial modification to the set of mixing parameters defines a direction, and wherein adjusting the set of mixing parameters using the set of gradients comprises adjusting the set of mixing parameters in the opposite direction in response to the combined objective-defining value worsening while the trial modification to the set of mixing parameters is applied. (Brownlee [Page 4, Paragraph 3]: “Because it can be expensive to search for the correct value of multiple hyperparameters, it is still reasonable to use the same weight decay at all layers just to reduce the size of search space.” Brownlee teaches that after the initial set of mixing parameters, there is a search to find the correct values of the weights of the mixing parameters.)
The reasons to combine are substantially similar to those of claim 4.
Regarding dependent claim 7, Wang, Czarnecki, and Brownlee teach:
A method as claimed in claim 5,
Brownlee teaches:
wherein determining the efficiency estimate comprises estimating a rate of change of the trial combined objective-defining value with time by determining a change in the trial combined objective- defining value over multiple machine learning time steps. (Brownlee [Page 4, Paragraph 3]: “Because it can be expensive to search for the correct value of multiple hyperparameters, it is still reasonable to use the same weight decay at all layers just to reduce the size of search space.” Brownlee teaches that after the initial set of mixing parameters, there is a search to find the correct values of the weights of the mixing parameters. Since every search has time complexity, it follows that the assessment of whether the mixing parameters are correctly found (rate of change) is found throughout time steps in the search.)
The reasons to combine are substantially similar to those of claim 4.
Regarding dependent claim 8, Wang, Czarnecki, and Brownlee teach:
A method as claimed in claim 7,
Brownlee teaches:
wherein determining the efficiency estimate comprises determining a difference between first and second mean returns from the trial combined objective-defining value at the respective start and end of a trial period. (Brownlee [Page 4, Paragraph 3]: “Because it can be expensive to search for the correct value of multiple hyperparameters, it is still reasonable to use the same weight decay at all layers just to reduce the size of search space.” Brownlee teaches that after the initial set of mixing parameters, there is a search to find the correct values of the weights of the mixing parameters. The search involves a difference between the starting parameter values and the ending parameter values.)
The reasons to combine are substantially similar to those of claim 4.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KYU HYUNG HAN whose telephone number is (703) 756-5529. The examiner can normally be reached on MF 9-5.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Alexey Shmatov can be reached on (571) 270-3428. 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.
/Kyu Hyung Han/
Examiner
Art Unit 2123
/ALEXEY SHMATOV/Supervisory Patent Examiner, Art Unit 2123
Prosecution Insights