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 .
Response to Amendment
The office action is responsive to the amendment filed on 02/10/2026. As directed by
the amendments claims 1-5, and 13-17 have been amended. Claims 1-20 are pending for examination.
Response to Arguments
Regarding Rejections Under 35 US.C. § 103:
Applicant’s arguments with respect to claims 1-20 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
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 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Ahmed 2018/0046149 A1 US (hereinafter Ahmed (I)) in view of Greene et al. US 2018/0046910 A1 (hereinafter Greene).
Ahmed (I) teaches the following:
A method, comprising: (Ahmed (I) FIG. 15 shows one embodiment of a method for building automation prediction in a building management system).
receiving, by one or more processors, a plurality of measurements for one or more points that are associated with a piece of building equipment, the plurality of measurements measured during a first time period; (Ahmed (I) [0045] teaches how the building automation system can collect operational data for a given building and [0046] teaches the operational data includes input and output. The input data includes any data used to control the operation of the building automation system and the output data includes any measuring performance of the building automation system. Further, Ahmed (I) [0047] teaches how the operational data can be provided for different times).
executing, by the one or more processors, a machine learning model using the plurality of measurements as an input to generate fault data for a plurality of future time periods subsequent to the first time period and assign a [prediction] to each of the plurality of future time periods, the [prediction] indicating a likelihood that a fault will occur in the piece of building equipment during a corresponding time period of the plurality of future time periods; ( Ahmed (I) [0052] and FIG. 3 teaches the building analytics system can use either building level analytics (element 22) or enterprise level analytics (element 24) and teaches the building automation systems may be diagnosed by the building analytics system using building level performance analytics. In particular, Ahmed (I) [0064] teaches “ the processor 16 of the building analytics system 17 uses a recurrent neural network or other machine learning based on a time series to predict the future time series and/or an event”. To be specific, Ahmed (I) [0112] & [0115] teaches the building analytic system, building process or other device of the building managements system can access input data (i.e., data associated with a given building automation system) and based on the input data, a machine-learnt classifier employed by the processor outputs the forecast or prediction such that a failure of a part is predicted based on the input times-series data with or without other data, a time of degraded performance (e.g., starting or range of times) is predicted and operation over time is forecast. Therefore, the machine-learnt classifier is able to generate fault data for a plurality of future time periods (see [0005])).
selecting, by the one or more processors, a second future time period from the plurality of future time periods responsive to satisfying a criterion indicating a fault will likely occur in the piece of building equipment during the second future time period of the plurality of future time periods; and ( Ahmed (I) [0087] teaches selecting a second period time that indicate a fault will occur as a future prediction degradation. Specifically, Ahmed teaches that the future predicted degradation may be of a likely time (second time period) for the degradation to occur, may be of a failure in the building automation system and may additional be of a part that will have degraded performance or a source of degraded performance. Further, Ahmed (I) [0118] teaches that any results such as “as failure occurring, the building automation system expected to fail, the part expected to fail, the time of expected failure...forecast operation may be output... [and] information derived from the forecast operation is output, such as a time and/or device operating in an undesired way”).
performing, by the one or more processors, an automated action responsive to the selection of the second future time period to prevent the fault from occurring during the second future time period, the automated action comprising at least one of operating the piece of building equipment in a different mode or at a different setpoint, shutting-down the piece of building equipment, directing control activities around the piece of building equipment, attempting to repair or work around the fault, or adjusting a speed, flow rate, temperature, or amount of light provided by the piece of building equipment ( Examiner will like to emphasize, the claim as presented teaches performing at least of the automated action for which Ahmed (I) [0064] teaches “the building processor trains a machine-learnt classifier to predict whether degradation will occur based on the input time series” such degradation can be “a particular part of the building automation system or degradation in general” ([0082]). Therefore, the prediction made by the machine-learnt classifier can be seen as an automated action that the building analytics system will use in order to alter operation of the enterprise, such as relocating employees before an expected shutdown that is working around the fault (see [0119])).
Ahmed (I) does not teach or suggest the prediction is a confidence score being assigned to each of the plurality of time periods, the prediction being a confidence score indicating a likelihood that a fault will occur in the piece of building equipment during a corresponding time period of the plurality of future time periods and assigning a confidence score to the second future time period.
Nevertheless, Greene teaches the following:
executing, by the one or more processors, a machine learning model using the [output] data for a plurality of future time periods subsequent to the first time period and assign a confidence score to each of the plurality of future time periods, the confidence score indicating a likelihood that a [condition] will occur in the building during a corresponding time period of the plurality of future time periods; ( Greene [0106] teaches using one or more quantitative measures” as input to a “condition prediction model” (i.e., machine learning model ) to “receiving, as an output of the condition prediction model, at least one prediction regarding the occurrence of an adverse condition at the first structural asset during one or more future time periods (i.e., plurality of future time periods subsequent to the first time period ) and a “plurality of confidence scores that respectively describe a confidence of the adverse condition respectively occurring at the first structural asset during a plurality of future time periods” (see [0108]).
…the confidence score assigned to the second future time period… (Greene [0108] teaches the confidence score is assigned to the “plurality of future time periods”, which will include a second future time period).
Greene is also in the same field of endeavor as Ahmed (I) (machine learning). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality confidence score indicating a likelihood that a condition will occur in the building during a corresponding time period of the plurality of future time periods, as being disclosed and taught by Greene, in the system taught by Ahmed (I) to yield the predictable results of predict and prevent adverse conditions (see [0037]).
Regarding claim 13: it is rejected under the same rationale of claim 1. Claim 13, only recites the additional elements of A system comprising one or more memory devices configured to store instructions thereon that, when executed by one or more processors, cause the one or more processors to... for which Ahmed (I) [0038] teaches one or more memory and [0035] teaches a system with a building processor that is a is a computer, server, panel, workstation, general processor, digital signal processor, application specific integrated circuit, field programmable gate array, analog circuit, digital circuit, combinations thereof, or other now known or later developed device for processing ([0059]).
-
Claims 2-3, 9 and 14-15 are rejected under 35 U.S.C. 103 as being unpatentable over Ahmed (I), Greene in further view of Papadopoulos et al. US 2021/0240147 A1 (hereinafter Papadopoulos).
Regarding claim 2:
Ahmed (I) and Greene each The method of claim 1. While Greene paragraphs [0106] and [0108] teach wherein executing, by the one or more processors, the machine learning model using the plurality of measurements to obtain a plurality of confidence scores for the plurality of future time periods.
Neither Ahmed (I) or Greene teach wherein executing the machine learning model using the plurality of measurements further comprises: executing, by the one or more processors, the machine learning model using the plurality of measurements to obtain a plurality of confidence scores for the plurality of time periods; and wherein selecting the second time period from the plurality of time periods is performed responsive to determining that the second time period is associated with a confidence score that satisfies a predetermined criteria.
Nevertheless, Papadopoulos teaches the following:
wherein executing, by the one or more processors, the machine learning model using the plurality of measurements to obtain a plurality of confidence scores for the plurality of future time periods; and (Papadopoulos [0151] teaches the output of the machine learning models that predict values of the variable for various time-steps include confidence scores. Such confidence score “indicate a degree of confidence that the machine learning model has that the predicted value for the variable is correct for a given time-step”).
wherein selecting the second future time period from the plurality of future time periods is performed responsive to determining that the second future time period is associated with a confidence score that satisfies a predetermined criteria (Papadopoulos [0151] teaches a machine leaning model can make future prediction (second time period) regarding the zone air temperature for a room and output a confidence score associated with it. The confidence score can be compared to a threshold and if it exceeds the threshold, the predicted value may be selected).
Papadopoulos is also in the same field of endeavor as Ahmed (I) and Greene (building management system). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality confidence scores, as being disclosed and taught by Papadopoulos in the system taught by Ahmed (I) and Greene to yield the predictable results of “optimize resource usage (e.g., electricity use, natural gas use, water use, etc.) and/or the monetary cost of such resource usage in response to satisfy the demand of building” (Papadopoulos [0072]).
Regarding claim 3:
Ahmed (I), Greene and Papadopoulos teach The method of claim 2. Papadopoulos specifically teaches, wherein determining the second future time period is associated with a confidence score that satisfies a predetermined criteria comprises determining, by the one or more processors, that the confidence score exceeds a threshold ( Papadopoulos [0091] teaches the data handler is able discard any values that are associated with confidence scores that do not exceed the threshold as being received or generated in error and any values that are associated with a confidence score that exceed the threshold are user and/or stored).
Regarding claim 9:
Ahmed (I) and Greene teach The method of claim 1.
Neither Ahmed (I) or Greene teach storing, by the one or more processors, an association between the machine learning and the piece of building equipment; retrieving, by the one or more processors, measurement data based on the stored association; and training, by the one or more processors, the machine learning model based on the retrieved measurement data.
Nevertheless, Papadopoulos teaches the following:
further comprising: storing, by the one or more processors, an association between the machine learning and the piece of building equipment; (Papadopoulos [0084] teaches storing building data for each business management system (BMS), for example data of a temperature sensor or an outside air temperature sensor may be collected such that it can be used to train the machine learning models of intervention prediction system to predict the impact that the intervention would likely have on a point or variable of the BMS).
retrieving, by the one or more processors, measurement data based on the stored association; and (Papadopoulos [0086] teaches the historical database hold data including various measurements and states associated with building equipment. Further, [0102] teaches how the data handler can configure to use building data that is stored ion historical database, therefore retrieving such data in order to use it).
training, by the one or more processors, the machine learning model based on the retrieved measurement data (Papadopoulos [0113] teaches how the data handler may be configured to use building data that is stored in historical database to train interrupted control model (a machine learning model)).
Regarding claim 14: it is rejected under the same rationale as claim 2. Claim 14 only recites the additional elements of the system... for which Ahmed (I) [0035] teaches the system.
Regarding claim 15: it is rejected under the same rationale as claim 3. Claim 15 only recites the additional elements of the system... for which Ahmed (I) [0035] teaches the system.
Claims 4-5, and 16-17 are rejected under 35 U.S.C. 103 as being unpatentable over Ahmed (I), Greene in further view of Ahmed US 2018/0046151 A1 (hereinafter Ahmed (II)).
Regarding claim 4:
Ahmed (I) and Greene teach The method of claim 1.
Neither Ahmed (I) and Greene disclose wherein the machine learning model is a first machine learning model, and further comprising: responsive to the selection of the second time period, executing, by the one or more processors, a second machine learning model using the plurality of measurements to obtain an output indicating a predicted root cause of the predicted fault in the piece of building equipment; wherein performing the automated action comprises generating, by the one or more processors, a record comprising a recommendation for resolving the predicted fault based on the predicted root cause.
However, Ahmed (II) teaches the following:
wherein the machine learning model is a first machine learning model, and further comprising: responsive to the selection of the second time period, executing, by the one or more processors, a second machine learning model using the plurality of measurements to obtain an output indicating a predicted root cause of the fault in the piece of building equipment; ( Ahmed (II), [0007] teaches “The first machine-learnt classifier is configured to identify a fault in the building automation system, and the second machine-learnt classifier is configured to identify a source of the fault” this implies that the failure prediction or forecast of the first machine-learnt classifier is used by the second machine learning model to identify the root cause. Furthermore, [0008] teaches how the second machine learning model uses data “related to operation of the building automation system”. In addition, [0065] teaches how the machine learner classifier (second machine leaning model) is able to identify a building automation system and/or subsystem that is malfunctioning. Moreover, Ahmed (II) [0071] teaches system level data (e.g., hot water supply temperature) is applied to other machine learning (such as the second machine learning model) in order to identify a source of fault (root cause) within the building).
wherein performing the automated action comprises generating, by the one or more processors, a record comprising a recommendation for resolving the predicted fault based on the predicted root cause (Ahmed (II) [0106] teaches outputting to the display the source, fault and/or the cost predicted by the cascades of the building analytics system. Further, [0133 & 0134] teaches the fault source is presented by the processor on a display of the building analytic system or might be printed or displayed remotely. Such that the source of the problem can be used to indicate a solution (recommendation) to be taken, for instance fixing or replacing a damper or valve).
Ahmed (II) is also in the same field of endeavor as Ahmed (I) and Greene (building automation). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of identifying the source fault using a second machine learning model as being disclosed and taught by Ahmed (II) in the system taught by Ahmed (I) and Greene to yield the predictable results of provide improvements for building automation systems employing analytics (see Ahmed (II) [0029]).
Regarding claim 5:
Ahmed (I), Greene and Ahmed (II) teach The method of claim 4. Ahmed (II) specifically teaches wherein executing the second machine learning model using the plurality of measurements further comprises executing, by the one or more processors, the second machine learning model using an identification of the second time period (Ahmed (II) [0009] teaches applying second data to a second machine learning model. Further, [0007] teaches a second machine learning model is configure to identify source of the fault and [0008] teaches how the second machine learning model uses data “related to operation of the building automation system”. Moreover, Ahmed (II) [0032] teaches meta-data from the enterprise is used to create “a classifier (second machine learning model) of building automation system”. Therefore, this implies that the second machine learning model uses such meta-data stored in order to identify fault).
Regarding claim 16: it is rejected under the same rationale as claim 4. Claim 16, only recites the additional elements of the system... for which Ahmed (I) [0035] teaches the system.
Regarding claim 17: it is rejected under the same rationale as claim 5. Claim 17 only recites the additional elements of the system... for which Ahmed (I) [0035] teaches the system.
Claims 6 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Ahmed (I), Greene, Ahmed (II) in further view of Mezic et al. US 2016/0203036 Al (hereinafter Mezic).
Regarding claim 6:
Ahmed (I), Greene and Ahmed (II) teach The method of claim 4. Ahmed (II) specifically teaches further comprising, presenting, by the one or more processors, the recommendation on a user interface; (Ahmed (II) [0133 & 0134] teaches the fault source is presented by the processor on a display of the building analytic system or might be printed or displayed remotely. Such that the source of the problem can be used to indicate a solution (recommendation) to be taken, for instance fixing or replacing a damper or valve).
training, by the one or more processors, the second machine learning model based on the predicted root cause and the input level of accuracy (Ahmed (II) [0101] teaches that in a feedback or online learning approach the training employed may include comparison of estimated verses actual performance and that by using this process, the building analytics system continues the training until a desired accuracy is reached. Therefore, the second machine learning model can be trained based on the input level of accuracy (feedback) provided by the user via the user interface).
Neither Ahmed (I), Greene or Ahmed (II) disclose receiving, by the one or more processors via the user interface, an input indicating a level of accuracy of the recommendation; and.
However, Mezic teaches the following:
receiving, by the one or more processors via the user interface, an input indicating a level of accuracy of the recommendation; and (Mezic [0056] teaches the user interference generator provides an opportunity for a user to provide feedback (input indicating a level of accuracy) on whether a detected fault can be confirmed as an actual fault. Such that, if the detected fault was misdiagnosed this feedback can be provided to the fault detection server such that artificial intelligence can be used to modify the behavior of the fault detection server in order to not identify similar type in the future ([0103])).
Mezic is also in the same field of endeavor as Ahmed (I), Greene and Ahmed (II) (fault detection). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of a user interface user interference generator that provides an opportunity for a user to provide feedback, as being disclosed and taught by Mezic, in the system taught by Ahmed (I), Greene and Ahmed (II) to yield the predictable results of improve fault detection in building such that the operation described can “provide significant improvements to the functioning of the fault detection server, reducing memory utilization and increasing processor performance through the reduction in the amount of data that needs to be stored and processed” ( Mezic [0062]).
Regarding claim 18: it is rejected under the same rationale as claim 6. Claim 18, only recites the additional elements of the system... for which Ahmed (I) [0035] teaches the system.
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Ahmed (I), Greene, Ahmed (II), Papadopoulos and Mezic.
Regarding claim 7:
Ahmed (I), Greene and Ahmed (II) teach The method of claim 4. While Ahmed (II) teaches wherein executing the second machine learning model using the plurality of measurements to obtain the output indicating the root cause (see rejection of claim 4) and teaches and teaches receiving, by the one or more processors via the user interface, a plurality of inputs indicating levels of accuracy of the plurality of confidence scores; and ( Ahmed (II) [0101] teaches that in a feedback or online learning approach the training employed may include comparison of estimated (i.e., the plurality of confidence scores) verses actual performance and that by using this process, the building analytics system continues the training until a desired accuracy is reached. Therefore, the second machine learning model can be trained based on the input level of accuracy (feedback) provided by the user via the user interface).
Neither Ahmed (I), Greene nor Ahmed (II) specifically teach further comprises executing, by the one or more processors, the second machine learning model using the plurality of measurements to obtain a plurality of confidence scores for a plurality of root causes for the fault, the method further comprising: presenting, by the one or more processors on a user interface, the plurality of confidence scores for the plurality of root causes; receiving, by the one or more processors via the user interface, a plurality of inputs indicating levels of accuracy of the plurality of confidence scores; and training, by the one or more processors, the second machine learning model based on the plurality of root causes and the plurality of inputs.
Nevertheless, Papadopoulos specifically teaches wherein executing the second machine learning model using the plurality of measurements to obtain the output indicating the root cause further comprises executing, by the one or more processors, the second machine learning model using the plurality of measurements to obtain a plurality of confidence scores for a plurality of root causes for the fault, the method further comprising: ( Papadopoulos [0151] teaches the output of the machine learning models ( this includes the second machine learning model) that predict values of the variable for various time-steps include confidence scores. Such confidence score “indicate a degree of confidence that the machine learning model has that the predicted value for the variable is correct for a given time-step”).
presenting, by the one or more processors on a user interface, the plurality of confidence scores for the plurality of root causes; (Papadopoulos [0134] teaches “The outputs of the machine learning models may be displayed in one user interface”).
Papadopoulos is also in the same field of endeavor as Ahmed (I), Greene and Ahmed (II) (building management system). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality confidence scores, as being disclosed and taught by Papadopoulos in the system taught by Ahmed (I), Greene and Ahmed (II) to yield the predictable results of “optimize resource usage (e.g., electricity use, natural gas use, water use, etc.) and/or the monetary cost of such resource usage in response to satisfy the demand of building” (Papadopoulos [0072]).
Ahmed (I), Greene and Ahmed (II) and Papadopoulos do not disclose training, by the one or more processors, the second machine learning model based on the plurality of root causes and the plurality of inputs.
However, Mezic teaches the following:
training, by the one or more processors, the second machine learning model based on the plurality of root causes and the plurality of inputs (Mezic [0103] teaches a user can provide feedback on whether the fault was accurately detected).
Mezic is also in the same field of endeavor as Ahmed (I), Greene and Ahmed (II) and Papadopoulos (fault detection). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of a user interface user interference generator that provides an opportunity for a user to provide feedback, as being disclosed and taught by Mezic, in the system taught by Ahmed (I), Greene and Ahmed (II) and Papadopoulos to yield the predictable results of improve fault detection in building such that the operation described can “provide significant improvements to the functioning of the fault detection server, reducing memory utilization and increasing processor performance through the reduction in the amount of data that needs to be stored and processed” ( Mezic [0062]).
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Ahmed (I), Greene in further view of Papadopoulos and Brown et al. US 2021/0081811 A1.
Regarding claim 8:
Ahmed (I) and Greene teach The method of claim 1.
Neither Ahmed (I) and Greene disclose storing, by the one or more processors, an association between the machine learning model and the piece of building equipment, wherein performing the automated action comprises: identifying, by the one or more processors, an identification of the piece of building equipment based on the stored association between the machine learning model and the piece of building equipment; and generating, by the one or more processors, a record comprising an identification of the piece of building equipment.
Nevertheless, Papadopoulos teaches the following:
further comprising: storing, by the one or more processors, an association between the machine learning model and the piece of building equipment, wherein performing the automated action comprises: (Papadopoulos [0084] teaches storing building data for each business management system (BMS), for example data of a temperature sensor or an outside air temperature sensor may be collected such that it can be used to train the machine learning models of intervention prediction system to predict the impact that the intervention would likely have on a point or variable of the BMS. Further, [0039] teaches performing the automated action such as displaying to a user the predicted times “for which the intervention would likely have a positive and/or a negative impact on a point or variable of a BMS” such that “it allows operators to view the impact that implementing the intervention will have and determine whether to follow through with the intervention”).
identifying, by the one or more processors, an identification of the piece of building equipment based on the stored association between the machine learning model and the piece of building equipment; and (Papadopoulos [0089] “In some embodiments, data handler can also tag the data with a device identifier tag indicating from which building device the building data was collected”).
Papadopoulos is also in the same field of endeavor as Ahmed (I) and Greene (building management system). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of storing an association between the machine learning model and the piece of building equipment and identifying an identification of the piece of building equipment based on the stored association between the machine learning model and the piece of building equipment, as being disclosed and taught by Papadopoulos in the system taught by Ahmed (I) and Greene to yield the predictable results of “optimize resource usage (e.g., electricity use, natural gas use, water use, etc.) and/or the monetary cost of such resource usage in response to satisfy the demand of building” (Papadopoulos [0072]).
Ahmed (I), Greene Papadopoulos do not teach or suggest generating, by the one or more processors, a record comprising an identification of the piece of building equipment.
However, Brown teaches the following:
generating, by the one or more processors, a record comprising an identification of the piece of building equipment (According to paragraph [0117] of the instant application a record if for example a file, document, table, listing, message, notification, etc. For which, Brown [0151] teaches the automated response includes the generation of a work order (i.e., document) that indicates the repair to equipment (e.g., HVAC equipment), thus it will include an identification of the piece of building equipment).
Brown is also in the same field of endeavor as Ahmed (I), Greene and Papadopoulos (building management system). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of generating a work order (i.e., record), as being disclosed and taught by Brown in the system taught Ahmed (I), Greene and Papadopoulos to yield the predictable results of “optimize building performance (e.g., efficiency, energy use, comfort, or safety) ” (Brown [0071]).
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Ahmed (I), Greene and Mezic.
Regarding claim 12:
Ahmed (I) and Greene teach The method of claim 1. Ahmed (I) specifically teaches wherein executing the machine learning model using the plurality of measurements further comprises executing, by the one or more processors, the machine learning model using the one or more setpoints (Ahmed (I) [0092] teaches the machine learner classifier uses the set point as input).
Neither Ahmed (I) nor Greene disclose identifying, by the one or more processors, one or more setpoints for the one or more points, the one or more setpoints configured for times within the first time period.
However, Mezic teaches the following:
identifying, by the one or more processors, one or more setpoints for the one or more points, the one or more setpoints configured for times within the first time period; (Mezic [0027] teaches “a first type or class of indicator function can define a setpoint ( e.g., a measured value in the time-series data, such as 70 degrees) and determine whether the setpoint is exceeded ( e.g., a true condition) or not exceeded ( e.g., a false condition) over time”).
Mezic is also in the same field of endeavor as Ahmed (I) and Greene (machine learning). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of a user interface user interference generator that provides an opportunity for a user to provide feedback, as being disclosed and taught by Mezic, in the system taught by Ahmed (I) and Greene to yield the predictable results of improve fault detection in building such that the operation described can “provide significant improvements to the functioning of the fault detection server, reducing memory utilization and increasing processor performance through the reduction in the amount of data that needs to be stored and processed” ( Mezic [0062]).
Claims 10 and 11 are rejected under 35 U.S.C. 103 as being unpatentable over Ahmed (I), Greene, Papadopoulos and in further view of Noboa et al. US 2012/0259583 A1 (hereinafter Noboa).
Regarding claim 10:
Ahmed (I) and Greene teach The method of claim 1.
Neither Ahmed (I) or Greene further comprising: grouping, by the one or more processors, the plurality of measurements into a plurality of time bins based on timestamps associated with the plurality of measurements, each time bin of the plurality of time bins associated with a different time window; and generating, by the one or more processors, a feature vector using the grouped plurality of measurements by labeling the plurality of measurements with labels identifying the time bins into which each of the plurality of measurements has been grouped, wherein executing the machine learning model using the plurality of measurements further comprises applying, by the one or more processors, the feature vector as an input into the machine learning model.
Nevertheless, Papadopoulos specifically teaches further comprising: generating, by the one or more processors, a feature vector using the grouped plurality of measurements by labeling the plurality of measurements with labels (Papadopoulos [0096] teaches the data handler can generate a feature vector and the feature vector may include numerical identifiers for the data. Further, teaches “the values of the data are divided and grouped based on the device with which the values are associated).
wherein executing the machine learning model using the plurality of measurements further comprises applying, by the one or more processors, the feature vector as an input into the machine learning model (Papadopoulos [0113] teaches how the feature vector can be used to predict values for time-steps into the future for a specific point and [0119] teaches each model may learn to handle feature vector inputs including point inputs).
Papadopoulos is also in the same field of endeavor as Ahmed (I) and Greene (machine learning). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality creating feature vector using the grouped plurality of measurements, as being disclosed and taught by Papadopoulos in the system taught by Ahmed (I) and Greene to yield the predictable results of “optimize resource usage (e.g., electricity use, natural gas use, water use, etc.) and/or the monetary cost of such resource usage in response to satisfy the demand of building” (Papadopoulos [0072]).
Neither Ahmed (I), Greene nor Papadopoulos disclose grouping, by the one or more processors, the plurality of measurements into a plurality of time bins based on timestamps associated with the plurality of measurements, each time bin of the plurality of time bins associated with a different time window; and identifying the time bins into which each of the plurality of measurements has been grouped....
However, Noboa teaches the following:
grouping, by the one or more processors, the plurality of measurements into a plurality of time bins based on timestamps associated with the plurality of measurements, each time bin of the plurality of time bins associated with a different time window; and (Noboa [0142] teaches the data may be collected in bins such that if the bins are based on time, each of the data collected may include a time stamp and it is separated and places into bins based on the corresponding time stamp).
...identifying the time bins into which each of the plurality of measurements has been grouped (Noboa [0139] teaches the performance values may be any data that can be used to determine whether the BMS is operating normally and [0142] teaches the performance values may be separated and placed into bins based on the corresponding time stamps).
Noboa is also in the same field of endeavor as Ahmed (I) and Papadopoulos (building management system). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of time-bins, as being disclosed and taught by Noboa, in the system taught by Ahmed (I) and Papadopoulos to yield the predictable results of “improve building efficiency, to enable greater or improved use of renewable energy sources, and to provide more comfortable and productive building” (see Noboa [0029]).
Noboa is also in the same field of endeavor as Ahmed (I), Greene and Papadopoulos (machine learning). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of time bins based on timestamps, as being disclosed and taught by Noboa, in the system taught by Ahmed (I), Greene and Papadopoulos to yield the predictable results of “improve building efficiency, to enable greater or improved use of renewable energy sources, and to provide more comfortable and productive buildings” ( see Noboa [0029]).
Regarding claim 11:
Ahmed (I), Greene and Papadopoulos and Noboa teach The method of claim 10. Papadopoulos specifically teaches wherein generating the feature vector using the received measurements further comprises generating, by the one or more processors, the feature vector using the determined averages and labeling, by the one or more processors, the determined averages with labels identifying the individual sub-time bins of the determined averages ( Papadopoulos [0096] teaches the data handler can generate a feature vector and teaches how the feature vector can include numerical identifier (labels) for the data such as values of the data (this can include values such as averages)).
While Papadopoulos [0095] teaches generating timeseries and apply a key identifier to it such as [<key, timestamp1, value1 >, <key, timestamp2, value2>, <key, timestamp3 , value3>]. Neither Ahmed (I), Greene or Papadopoulos disclose measurements for individual time bins or sub-time bins and how the feature vectors are being used to identify the individual sub-time bins.
Nonetheless, Noboa teaches the following:
wherein grouping the plurality of measurements into the plurality of time bins further comprises: (Noabo [0142] teaches collecting and separating data into time bins).
grouping, by the one or more processors, measurements of individual time bins of the plurality of time bins into a plurality of sub-time bins; and (Noboa [0141] teaches if the bins are based on time the performance values may be collected at equally space time intervals (sub-time bins). Further, [0142] teaches dividing each hours into 15 minutes sub-time bins such that that the performance values can be organized in a 6:00 am bins and other performance values can be organized in 6:15 am. Moreover, [0172] and FIG. 18 teaches sub-time bins such that “if the monitor period last 24 hours, the chart will include columns of bins for 15 minutes intervals, for the entire 24 hours”).
determining, by the one or more processors, averages of measurements of individual sub- time bins of the plurality of sub-time bins, ( Noboa [0098] teaches exponential weighted averages (EWMAs) to diagnose faults in building management controllers and teaches how it can be calculated using the following formula:
PNG
media_image1.png
43
127
media_image1.png
Greyscale
where x, is the EWMA at time t;
λ
is an exponential smoothing constant or filter value; and
x
t
-
j
is the value of the signal at time t-j. Further, [0107] teaches EWMAs may be calculated directly on building equipment controllers).
the determined averages with labels identifying the individual sub-time bins of the determined averages (Noboa FIG. 18 teaches sub-time bins such that “if the monitor period last 24 hours, the chart will include columns of bins for 15 minutes intervals, for the entire 24 hours”. Further, [0107] teaches EWMAs may be calculated directly on building equipment controllers).
Claim 19 is rejected under 35 U.S.C. 103 as being unpatentable over Ahmed (I), Ahmed (II), Greene in further view of Brown.
Regarding claim 19:
A method, comprising: (Ahmed (I) FIG. 15 shows one embodiment of a method for building automation prediction in a building management system).
receiving, by one or more processors, a plurality of measurements for one or more points that are associated with a piece of building equipment, the plurality of measurements measured during a first time period; (Ahmed (I) [0045] teaches how the building automation system can collect operational data for a given building and [0046] teaches the operational data includes input and output. The input data includes any data used to control the operation of the building automation system and the output data includes any measuring performance of the building automation system. Further, Ahmed [0047] teaches how the operational data can be provided for different times).
executing, by the one or more processors, a first machine learning model using the plurality of measurements to obtain an output predicting a fault will occur in the piece of building equipment within a second future time period subsequent to the first time period and assign a [prediction] to the second future time period, the [prediction] indicating a likelihood that a fault will occur in the piece of building equipment during the second future time period,the input times-series data with or without other data, a time of degraded performance (e.g., starting or range of times) is predicted and operation over time is forecast. Therefore, the machine-learnt classifier is able to generate fault data for a plurality of future time periods (see [0005])).
Ahmed (I) does not teach ...and assign a confidence score to the second future time period, the confidence score indicating a likelihood that a fault will occur in the piece of building equipment during the second future time period, the confidence score satisfying a criterion; responsive to the prediction that the fault will occur in the piece of building equipment within the second future time period, executing, by the one or more processors, a second machine learning model using the plurality of measurements and an identification of the second future time period to obtain an output indicating a predicted root cause of the fault in the piece of building equipment; and performing, by the one or more processors, an automated action responsive to the predicted root cause of the fault in the piece of building equipment to prevent the fault from occurring during the second future time period, the automated action comprising at least one of operating the piece of building equipment in a different mode or at a different setpoint, shutting- down the piece of building equipment, directing control activities around the piece of building equipment, attempting to repair or work around the fault, or adjusting a speed, flow rate, temperature, or amount of light provided by the piece of building equipment.
However, Ahmed (II) teaches the following:
responsive to the prediction that the fault will occur in the piece of building equipment within the second future time period, executing, by the one or more processors, a second machine learning model using the plurality of measurements and an identification of the second future time period to obtain an output indicating a predicted root cause of the fault in the piece of building equipment; and ( Ahmed (II) [0071] teaches system level data (e.g., hot water supply temperature) is applied to other machine learning (such as the second machine learning model) in order to identifying a source of fault (root cause) within the building).
performing, by the one or more processors, an automated action responsive to the predicted root cause of the predicted fault in the piece of building equipment to prevent the fault from occurring during the second future time period, the automated action comprising at least one of operating the piece of building equipment in a different mode or at a different setpoint, shutting- down the piece of building equipment, directing control activities around the piece of building equipment, attempting to repair or work around the fault, or adjusting a speed, flow rate, temperature, or amount of light provided by the piece of building equipment ( Ahmed (II) [0106] teaches outputting to the display the source, fault and/or the cost predicted by the cascades of the building analytics system. Further, [0056] teaches “The building analytics system 17 may use any of various sources of input data and building data with any of various data analytics to provide various information, such as control functions or a diagnosis of a potential fault for a building automation system 12 or other device within the enterprise 10” thus, the building analytics system can automatically perform an automated action to prevent a fault from occurring).
Ahmed (II) is also in the same field of endeavor as Ahmed (I) (building automation). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of identifying the source fault using a second machine learning model as being disclosed and taught by Ahmed (I) in the system taught by Ahmed (I) to yield the predictable results of provide improvements for building automation systems employing analytics (see Ahmed (II) [0029]).
Neither Ahmed (I) or Ahmed (II) teaches or suggests ...and assign a confidence score to the second time period, the confidence score indicating a likelihood that a fault will occur in the piece of building equipment during the second time period, the confidence score satisfying a criterion;
Nevertheless, Greene teaches the following:
executing, by the one or more processors, a first machine learning model using the plurality of measurements to obtain an output predicting a [condition] will occur in the building within a second future time period subsequent to the first time period and assign a confidence score to the second future time period, the confidence scoreindicating a likelihood that a [condition] will occur in the piece of building equipment during the second future time period, and a “plurality of confidence scores that respectively describe a confidence of the adverse condition respectively occurring at the first structural asset during a plurality of future time periods” (see [0108]).
Greene is also in the same field of endeavor as Ahmed (I) and Ahmed (II) (machine learning). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality confidence score indicating a likelihood that a condition will occur in the building during a corresponding time period of the plurality of future time periods, as being disclosed and taught by Greene, in the system taught by Ahmed (I) and Ahmed (II) to yield the predictable results of predict and prevent adverse conditions (see [0037]).
Neither Ahmed (I), Ahmed (II), Greene teach the confidence score satisfying a criterion.
Nonetheless, Brown teaches the following:
…the confidence score satisfying a criterion (Brown [0149] teaches a “confidence level may be represented as a percentage (e.g., where 100% indicates a completely accurate prediction), a number value, etc.” and teaches assigning a confidence level for a prediction. In addition, Brown [0149] teaches the confidence level can be compared to a criterion (i.e., threshold) in order to check if the confidence level satisfies the criterion).
Brown is also in the same field of endeavor as Ahmed (I), Ahmed (II), Greene (machine learning). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of confidence score satisfying a criterion, as being disclosed and taught by Brown in the system taught Ahmed (I), Ahmed (II), Greene to yield the predictable results of “optimize building performance (e.g., efficiency, energy use, comfort, or safety) ” (Brown [0071]).
Claim 20 is rejected under 35 U.S.C. 103 as being unpatentable over Ahmed (I), Ahmed (II), Greene, Brown in further view of Mezic.
Regarding claim 20:
Ahmed (I), Ahmed (II), Greene, and Brown teach The method of claim 19. Ahmed (II) specifically teaches wherein performing the automated action comprises generating, by the one or more processors, a record comprising a recommendation for resolving the predicted fault based on the predicted root cause, further comprising: (Ahmed (II) [0106] teaches outputting to the display the source, fault and/or the cost predicted by the cascades of the building analytics system. Further, [0133 & 0134] teaches the fault source is presented by the processor on a display of the building analytic system or might be printed or displayed remotely. Such that the source of the problem can be used to indicate a solution (recommendation) to be taken, for instance fixing or replacing a damper or valve).
presenting, by the one or more processors, the recommendation on a user interface; (Ahmed (II) [0133 & 0134] teaches the fault source is presented by the processor on a display of the building analytic system or might be printed or displayed remotely. Such that the source of the problem can be used to indicate a solution (recommendation) to be taken, for instance fixing or replacing a damper or valve)
training, by the one or more processors, the second machine learning model based on the predicted root cause and the input level of accuracy (Ahmed (II) [0101] teaches that in a feedback or online learning approach the training employed may include comparison of estimated verses actual performance and that by using this process, the building analytics system continues the training until a desired accuracy is reached. Therefore, the second machine learning model can be trained based on the input level of accuracy (feedback) provided by the user via the user interface).
Neither Ahmed (I), Ahmed (II), Greene, nor Brown disclose receiving, by the one or more processors via the user interface, an input indicating a level of accuracy of the recommendation.
However, Mezic teaches the following:
receiving, by the one or more processors via the user interface, an input indicating a level of accuracy of the recommendation; and (Mezic [0056] teaches the user interference generator provides an opportunity for a user to provide feedback (input indicating a level of accuracy ) on whether a detected fault can be confirmed as an actual fault. Such that, if the detected fault was misdiagnosed this feedback can be provided to the fault detection server such that artificial intelligence can be used to modify the behavior of the fault detection server in order to not identify similar type in the future ([0103])).
Mezic is also in the same field of endeavor as Ahmed (I), Ahmed (II), Greene, and Brown (fault detection). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the functionality of a user interface user interference generator that provides an opportunity for a user to provide feedback, as being disclosed and taught by Mezic, in the system taught by Ahmed (I), Ahmed (II), Greene, and Brown to yield the predictable results of improve fault detection in building such that the operation described can “provide significant improvements to the functioning of the fault detection server, reducing memory utilization and increasing processor performance through the reduction in the amount of data that needs to be stored and processed” (Mezic [0062]).
Conclusion
THIS ACTION IS MADE FINAL. 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 GISEL G FACCENDA whose telephone number is (703)756-1919. The examiner can normally be reached Monday - Friday 8:00 am - 4:00 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, Abdullah Al Kawsar can be reached at (571) 270-3169. 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.
/G.G.F./Examiner, Art Unit 2127
/ABDULLAH AL KAWSAR/Supervisory Patent Examiner, Art Unit 2127
Prosecution Insights