DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 07/01/2026 has been entered.
Status of Claims
Applicant filed an RCE on 07/01/2026. Claim 1 is Currently Amended. Claims 1–14 are pending.
Response to Arguments
In regards to the previous rejection of claims 1–3 and 5–14 under 35 U.S.C. § 103: Applicant argues that the prior art does not explicitly disclose the limitations "wherein the setting data packet contains context data acquired with sensors in the vehicle, the context data comprising at least one context value selected from the group consisting of interior temperature, exterior temperature, humidity, position of the sun, light intensity, and air pressure, and user data comprising at least one user value selected from the group consisting of skin temperature, pulse rate, respiratory rate, sitting position, emotional state, and biometric inputs" and "wherein the modeling data packet is created in response to manual user input from the user to change the current individualized settings of the air conditioner, and the modeling data packet comprises the setting data packet combined with the user input, wherein the user input comprises optimization data containing at least one optimization value regarding user-specific preferences". Applicant further argues on page 8 of the Remarks, "Stevanovic in view of Woods fail to disclose a method for creating an exportable individualized data set for controlling an air conditioner in a vehicle comprising a setting data packet and a modeling data packet wherein the setting data packet contains context data acquired with sensors in the vehicle, the context data comprising at least one context value selected from the group consisting of interior temperature, exterior temperature, humidity, position of the sun, light intensity, and air pressure, and user data comprising at least one user value selected from the group consisting of skin temperature, pulse rate, respiratory rate, sitting position, emotional state, and biometric inputs and wherein the modeling data packet is created in response to manual user input from the user to change the current individualized settings of the air conditioner, and the modeling data packet comprises the setting data packet combined with the user input, wherein the user input comprises optimization data containing at least one optimization value regarding user-specific preferences." Applicant's arguments filed July 1, 2026 with respect to Stevanovic in view of Woods, insofar as they are directed to the newly added limitation "wherein the user input comprises optimization data containing at least one optimization value regarding user-specific preferences", have been fully considered but they 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 in view of Saeki (US20200331320A1). Saeki is newly incorporated to teach that limitation, including a thermal sensation that is "an indicator representing how a user senses an in-vehicle temperature environment" and a preference estimating section that estimates "temperature preference of the user based on a temporal change in the thermal sensation" (see at least Saeki, para. [0005]; see also para. [0056], [0073]–[0074]).
As to the remaining argued features, which the new ground of rejection continues to attribute to Stevanovic and Woods, Examiner respectfully disagrees. Applicant is reminded claims must be given their broadest reasonable interpretation. Each of the two argued groups requires only "at least one" value "selected from the group." As recited in the rejection below, Stevanovic is incorporated to disclose wherein the setting data packet contains context data acquired with sensors in the vehicle, the context data comprising at least one context value selected from the group consisting of interior temperature, exterior temperature, humidity, position of the sun, light intensity, and air pressure. Stevanovic's adaptation application "monitors these instances and collects relevant data from the sensors 74 through the vehicle bus 32" (see at least Stevanovic, para. [0035]), and the collected readings include "outside temperature readings, inside temperature readings (e.g., interior roof temperature, interior dash board temperature, etc.), humidity readings, windshield wiper indication or settings, sun intensity, elevation and/or azimuth readings, ambient light status" (see at least Stevanovic, para. [0031]; see also para. [0040]). Stevanovic further discloses that "Temperature and/or humidity sensors may be positioned inside and/or outside of the vehicle 12" and that "a sun sensor (e.g., from Solar MEMS Technology) may be used to detect the sun's position or elevation" (see at least Stevanovic, para. [0032]). Stevanovic therefore discloses at least interior temperature, exterior temperature, humidity, position of the sun, and light intensity acquired with sensors in the vehicle, where any one of these meets the claimed group.
Examiner is aware that Stevanovic does not explicitly disclose the recited user data; however, Woods is incorporated to teach and user data comprising at least one user value selected from the group consisting of skin temperature, pulse rate, respiratory rate, sitting position, emotional state, and biometric inputs. Woods teaches that "the sensors 124 may include a temperature sensor to detect a skin temperature of the user 110" and that the vehicle computing device may detect "a drowsiness level of the user 110 (e.g., a driver) and/or a stress level of the user 110" (see at least Woods, para. [0036]). Woods stores this user status data, which may include "skin temperatures, a drowsiness level, a stress level" (see at least Woods, para. [0038]). Skin temperature is expressly recited in the claimed group, and a stress level of the user, under broadest reasonable interpretation, is considered an emotional state. Woods also places this user status data in the same data set used to train its personal inference model: "The information may include the historical in-vehicle sensor data 152, HVAC state data 154, vehicle state data 156, user identity data 158, user status data 160, user demographic data 162, and/or external data 164. Based on the received information, the computing device 122 may generate and train a personal inference model" (see at least Woods, para. [0042]).
Further, Stevanovic is incorporated to disclose wherein the modeling data packet is created in response to manual user input from the user to change the current individualized settings of the air conditioner, and the modeling data packet comprises the setting data packet combined with the user input. In Stevanovic, manual climate control settings "may be any setting of the HVAC system 70 that the user or another in-vehicle occupant manually inputs while in the vehicle 12" (see at least Stevanovic, para. [0029]). When "a user manually adjusts or changes a climate control setting", the adaptation application "recognizes that the adjustment, change, or report is being made and collects data associated with the particular climate control setting instance" (see at least Stevanovic, para. [0035]). Then, "After the data for the climate control setting instance is collected, the adaptation application 46 generates a comfort setting record for the climate control setting instance. The comfort setting record includes the data points (i.e., location and time/date data points for the climate control setting instance) and the climate control related setting(s)." (see at least Stevanovic, para. [0042]). The climate control related settings include both sensed context values and the user's own setting, e.g., "an outside air temperature data point, an outside humidity data point", "an interior air temperature data point", "a sun intensity data point, a sun elevation data point, a sun azimuth data point; an ambient light status indication", and "a driver temperature setting" (see at least Stevanovic, para. [0040]). These records are the training data for the prediction model: "the adaptation application 46 utilizes the retrieved data as training data for a machine learning algorithm and for building a prediction model" (see at least Stevanovic, para. [0052]). Under broadest reasonable interpretation, a comfort setting record that is generated in response to a manual change and that carries the sensed context data together with the user's manual setting is considered a modeling data packet that is created in response to manual user input and that comprises the setting data packet combined with the user input.
Applicant further argues on page 8 of the Remarks, "Neither Stevanovic nor Woods teaches or suggests incorporating optimization data containing user-specific perceptions (such as 'skin temperature, pulse rate, respiratory rate, sitting position, emotional state, and biometric inputs') into the modeling data packet for training the comfort model. Stevanovic's user input is limited to changing numerical HVAC settings. Woods' data collection focuses on sensor measurements and HVAC states rather than qualitative perception inputs from the user. There is no motivation to combine these references to arrive at the claimed optimization data, as neither reference discloses or suggests the concept of using qualitative perception-based inputs as training data for an AI comfort model." To the extent this argument is directed to the optimization data, it is moot for the reasons given above, because Saeki, not Stevanovic or Woods, is relied upon for wherein the user input comprises optimization data containing at least one optimization value regarding user-specific preferences. As to Stevanovic and Woods, Examiner respectfully disagrees with Applicant's characterizations. First, Stevanovic's user input is not limited to changing numerical HVAC settings. The monitored settings include non-numerical states such as "an air condition indication (e.g., on or off), a defrost indication" (see at least Stevanovic, para. [0029]) and non-HVAC comfort settings such as "window settings, sunroof status, heated and/or cooled seat settings" (see at least Stevanovic, para. [0031]). In any event, the claim requires only "manual user input from the user to change the current individualized settings of the air conditioner", and Stevanovic's manual climate control settings meet that language. Second, Woods' data collection is not limited to sensor measurements and HVAC states. As shown above, Woods also collects user status data (skin temperature, drowsiness level, stress level) and uses it to train the personal inference model (see at least Woods, para. [0036], [0038], [0042]), so that the HVAC control "adapts to the preferences of the occupant or occupants of a vehicle" (see at least Woods, para. [0027]). Third, the items Applicant lists as "user-specific perceptions" ("skin temperature, pulse rate, respiratory rate, sitting position, emotional state, and biometric inputs") are the user values of the claimed setting data packet, and Woods teaches them as shown above. Finally, the claims do not recite "qualitative perception inputs" or "qualitative perception-based inputs." Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims.
In response to Applicant's argument that "There is no motivation to combine these references," the examiner recognizes that obviousness may be established by combining or modifying the teachings of the prior art to produce the claimed invention where there is some teaching, suggestion, or motivation to do so found either in the references themselves or in the knowledge generally available to one of ordinary skill in the art. The rejection below states an explicit reason for each combination. Stevanovic is modified by Woods because both references address personalized vehicle HVAC comfort based on learned user-specific models, and Woods closes the loop from personal inference prediction to HVAC control while adding occupant physiological/status sensing (skin temperature; stress/drowsiness) that improves personalization beyond Stevanovic's climate-setting records alone. That reason comes from Woods itself, which provides HVAC control that "adapts to the preferences of the occupant or occupants of a vehicle" and that can "match occupants' preferences based on the history of occupant-initiated environmental controls" (see at least Woods, para. [0027]). This is the same goal as Stevanovic's method "for personalizing vehicular comfort settings for a specific user" (see at least Stevanovic, para. [0003]). The combination is further modified by Saeki because Stevanovic already records manual climate inputs in its modeling records, Woods already personalizes HVAC targets from preference differences, and Saeki supplies a thermal-sensation-to-temperature-preference estimation that converts user-related inputs into stored preference values used to control the in-vehicle temperature conditioner (see at least Saeki, para. [0005], [0056], [0074]). Further, in response to Applicant's arguments against the references individually (e.g., "Stevanovic's user input is limited to changing numerical HVAC settings" and "Woods' data collection focuses on sensor measurements and HVAC states"), one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references.
In regards to the previous rejection of claim 4 under 35 U.S.C. § 103: Applicant argues on page 9 of the Remarks, "Ostrowski has been carefully reviewed and Applicant submits that Ostrowski does not disclose or suggest the limitations of amended claim 1 missing from Stevanovic and Woods, let alone the limitations of claim 4." Applicant's arguments filed July 1, 2026 with respect to Stevanovic in view of Woods and further in view of Ostrowski have been fully considered but they 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 in view of Kelly (US20030160104A1) and Fusco et al. (US 6,454,178 B1). Ostrowski is no longer relied upon. Claim 4 is now rejected over Stevanovic in view of Woods and further in view of Saeki, and further in view of Kelly and Fusco, as set forth below.
Claim Objections
Claim 4 is objected to because of the following informalities: the claim recites "the offset settings" and "the offset data" without a clear antecedent basis in claim 1 (or in claim 4 itself prior to those definite articles). See 37 CFR 1.75; MPEP § 608.01. Appropriate correction is required.
Claim Rejections - 35 USC § 103
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 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 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 1–3 and 5–14 are rejected under 35 U.S.C. 103 as being unpatentable over Stevanovic et al. (US20170158023A1) in view of Woods et al. (US20200031195A1) and further in view of Saeki (US20200331320A1), hereinafter referred to as Stevanovic, Woods, and Saeki.
Regarding claims 1 and 14
Stevanovic discloses
A method for creating an exportable individualized data set for controlling an air conditioner in a vehicle for a user using a control structure ("upon launching the adaptation application, automatically downloading a historical comfort setting profile for the specific user, the historical comfort setting profile including individual comfort setting records" [0003]; "the specific user is recognized within the vehicle and his/her profile is downloaded to the vehicle" [0007]; "the server 14 may automatically download the historical comfort setting profile in response to a request from the VCP 28" [0047]; "the user's historical comfort setting profile is regularly updated, regardless of which vehicle 12, 12′ he/she is using" [0045]),
wherein, in a repeated modeling loop:
at least a first modeling data packet for the user is acquired ("the adaptation application 46 collects data related to instances of climate control settings in the vehicle 12, 12′ while the user is in the vehicle 12 in order to build up a historical comfort setting profile for the specific user" [0027]; "After the data for the climate control setting instance is collected, the adaptation application 46 generates a comfort setting record for the climate control setting instance" [0042]),
a comfort model is trained in an AI unit with data from at least one modeling data packet for the user ("the adaptation application 46 utilizes the retrieved data as training data for a machine learning algorithm and for building a prediction model. the prediction model is a product of applying the machine learning algorithm on the given data" [0052]; "the machine learning algorithm is C4.5 or J48 … each of which builds decision trees from the training data" [0052]),
revising the comfort model in the AI unit based on at least a second modeling data packet for the user ("the historical comfort setting profile is autonomously updated with a comfort setting record corresponding to each instance of an automatic climate control setting, a manual climate control setting, or an alternate climate control setting while the specific user is in the vehicle" [0003]; "Upon receiving a comfort setting record, the server 14 matches the comfort setting record with a profile of the specific user, and updates the historical comfort setting profile of the specific user with the comfort setting record" [0063]),
repeatedly revising the comfort model in the AI unit based on a plurality of additional modeling data packets for the user ("the adaptation application 46 may collect data and generate a comfort setting record for each instance of a climate control setting that is recognized while the specific user is in the vehicle 12, 12′. In a single trip … any number of climate control setting instances and corresponding comfort setting records may be made and generated" [0043]; "the user's historical comfort setting profile is regularly updated" [0045]; "at the end of one predetermined time increment, the adaptation application 46 retrieves data from the historical comfort setting profile (which may have been recently updated if changes or adjustments have been made, or new data is reported), and runs the prediction model to update the predicted in-vehicle setting(s)" [0054]),
wherein, in a repeated setting loop:
a setting data packet for the vehicle is acquired and sent to the trained comfort model ("the adaptation application 46 may also be programmed to collect the current state of the HVAC system 70 and vehicle 12 at predetermined/regular intervals" [0035]; "the adaptation application 46 retrieves data from comfort setting records within a predetermined time frame … [and] utilizes the retrieved data as training data for a machine learning algorithm and for building a prediction model" [0050]–[0052]),
individualized settings for the air conditioner are predicted by the comfort model based on the data in the setting data packet ("An in-vehicle setting is dynamically predicted within respective predetermined time increments while the specific user is in the vehicle or in response to a user request. the predicted in-vehicle setting is dependent upon geographic location data points and a set of climate control related settings retrieved from the individual comfort setting records" [0003]; "the prediction classifier includes one or more climate control setting(s) that is/are likely to be implemented" [0052]; "the predicted in-vehicle air temperature is 65° C. and the predicted air conditioner status is on" [0056]),
wherein the setting data packet contains context data acquired with sensors in the vehicle, the context data comprising at least one context value selected from the group consisting of interior temperature, exterior temperature, humidity, position of the sun, light intensity, and air pressure ("Examples of other non-HVAC system settings or readings that may be indicative of the climate include outside temperature readings, inside temperature readings (e.g., interior roof temperature, interior dash board temperature, etc.), humidity readings, … sun intensity, elevation and/or azimuth readings, ambient light status" [0031]; "Examples of the climate control related setting that may be received by the adaptation application 46 include an outside air temperature data point, an outside humidity data point, … an interior air temperature data point, … a sun intensity data point, a sun elevation data point, a sun azimuth data point; an ambient light status indication" [0040]),
and wherein the modeling data packet is created in response to manual user input from the user to change the current individualized settings of the air conditioner, and the modeling data packet comprises the setting data packet combined with the user input ("Manual climate control settings … may be any setting of the HVAC system 70 that the user or another in-vehicle occupant manually inputs while in the vehicle 12" [0029]; "When … a user manually adjusts or changes a climate control setting, … the adaptation application 46 recognizes that the adjustment, change, or report is being made and collects data associated with the particular climate control setting instance" [0035]; "the comfort setting record includes the data points … and the climate control related setting(s)" [0042]).
Stevanovic does not explicitly disclose
the air conditioner output is controlled on the basis of the individualized settings,
and user data comprising at least one user value selected from the group consisting of skin temperature, pulse rate, respiratory rate, sitting position, emotional state, and biometric inputs,
wherein the user input comprises optimization data containing at least one optimization value regarding user-specific preferences.
However, Woods does teach the air conditioner output is controlled on the basis of the individualized settings ("outputting the one or more target values including the personal target value for the vehicle or the user of the vehicle to control the HVAC system of the vehicle" [0003]; "the computing device 122 may output the predicted target value to the HVAC controller 126 to control the HVAC system 128 so as to provide a desired in-cabin environment" [0042]; "the computing device 122 may combine the general target value and the personal target adjustment value to generate a personalized target value to control the HVAC system 128" [0047]);
and user data comprising at least one user value selected from the group consisting of skin temperature, pulse rate, respiratory rate, sitting position, emotional state, and biometric inputs ("the sensors 124 may include a temperature sensor to detect a skin temperature of the user 110" [0036]; "the computing device 122 of the vehicle system 104 may analyze videos or images captured by the camera 124 to detect a drowsiness level of the user 110 … and/or a stress level of the user 110" [0036]; "these data 158, 160 and 162 associated with a user 110 may include the user's 110 characteristic, bio-information, images, … skin temperatures, a drowsiness level, a stress level" [0038]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Stevanovic's personalized climate prediction system to control the air conditioner output on the basis of the individualized settings and to include user data comprising at least one user value selected from the group consisting of skin temperature, pulse rate, respiratory rate, sitting position, emotional state, and biometric inputs as taught by Woods with a reasonable expectation of success, because that improves personalization of vehicle climate-settings See at least [0036, 0038]) .
However, Saeki does teach wherein the user input comprises optimization data containing at least one optimization value regarding user-specific preferences ("the preference estimating section is configured to estimate temperature preference of the user based on a temporal change in the thermal sensation. the preference memory section is configured to memorize the temperature preference. the control section is configured to control the in-vehicle temperature conditioner based on the temperature preference memorized by the preference memory section" [0005]; "the preference estimating section 370 estimates the temperature preference based on the thermal sensation estimated by the thermal sensation estimating section 330" [0056]; "the five groups of temperature preference are a heat-sensitive and quick heating preferring group, a heat-sensitive group, a neutral group, a cold-sensitive group, and a cold-sensitive and quick heating preferring group" [0073]; "the server control section 320 … causes the preference memory section 350 to memorize the temperature preference estimated by the preference estimating section 370" [0074]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the Stevanovic method (as modified by Woods) so that the user input comprises optimization data containing at least one optimization value regarding user-specific preferences as taught by Saeki with a reasonable expectation of success. Saeki supplies an explicit thermal-sensation, temperature-preference estimation pipeline that converts user-related inputs into stored preference optimization values usable to control the in-vehicle temperature conditioner (See at least [0056, 0073-0074]).
Regarding claim 2
Stevanovic discloses wherein the creation of the modeling data packet is initiated when user input is entered manually by the user to change the current individualized settings of the air conditioner ("When … a user manually adjusts or changes a climate control setting, … the adaptation application 46 recognizes that the adjustment, change, or report is being made and collects data associated with the particular climate control setting instance" [0035]; "Manual climate control settings … may be any setting of the HVAC system 70 that the user … manually inputs" [0029]),
and/or the creation of the modeling data packet is initiated after a predefined time interval, if the user does not manually enter any user input to change the current individualized settings of the air conditioner within a predefined time interval ("the adaptation application 46 may also be programmed to collect the current state of the HVAC system 70 and vehicle 12 at predetermined/regular intervals (e.g., every 5 minutes …). As such, the adaptation application 46 may collect data even when a particular climate control setting instance is not recognized. This will ensure that individual comfort setting records are recorded … even in the absence of a particular climate control setting instance" [0035]).
Regarding claim 3
Stevanovic discloses wherein the setting loop is executed at a predefined frequency ("An in-vehicle setting is dynamically predicted within respective predetermined time increments while the specific user is in the vehicle" [0003]; "the adaptation application 46 monitors the climate control activity every half second while the specific user is in the vehicle 12 or 12′" [0035]; "In an example, the predetermined time increment is every minute … In another example, the predetermined time increment is every 7 seconds" [0054]).
Regarding claim 5
Stevanovic discloses wherein the comfort model is trained periodically at a predefined time, wherein all modeling data packets for the user are stored between successive times, and used to train the comfort model ("the adaptation application 46 retrieves data from comfort setting records within a predetermined time frame (e.g., 2 weeks, 3 months, etc. prior to the current date). the predetermined time frame is a window period within which the historical comfort setting data is considered suitable for modeling the behavior of the user" [0050]; "the adaptation application 46 utilizes the retrieved data as training data for a machine learning algorithm and for building a prediction model" [0052]; "at the end of one predetermined time increment, the adaptation application 46 retrieves data from the historical comfort setting profile … and runs the prediction model" [0054]),
and/or the data from the modeling data packets are weighted when training the comfort model ("the adaptation application 46 associates probabilities to candidate climate control settings … at a requested time, or according to a predetermined schedule" [0051]; "a probability is generated for each climate control related setting in the data that is analyzed from the historical comfort setting profile" [0053]).
Regarding claim 6
Stevanovic discloses wherein the modeling loop is executed in the vehicle, and the comfort model is trained in the AI unit in the vehicle ("the adaptation application 46 may autonomously launch in response to the specific user being identified in the vehicle 12 or 12′" [0027]; "the electronic memory 44 of the VCP 28 may be an encrypted memory that is configured to store i) an adaptation application 46 to be executed by the processor 42" [0026]; "the adaptation application 46 utilizes the retrieved data as training data for a machine learning algorithm and for building a prediction model" [0052]).
Regarding claim 7
Stevanovic discloses wherein, in the modeling loop: the comfort model is stored in the vehicle after it has been trained ("the historical comfort setting profile is automatically downloaded for the user" [0003]; "the electronic memory 44 … store i) an adaptation application 46" [0026]);
and in the setting loop: the comfort model is executed in the vehicle and the individualized settings are predicted locally in the vehicle ("the adaptation application 46 also uses the historical comfort setting profile to predict in-vehicle setting(s) for the specific user while the user is in the vehicle 12 or 12′" [0027]; "An in-vehicle setting is dynamically predicted within respective predetermined time increments while the specific user is in the vehicle" [0003]).
Regarding claim 8
Stevanovic does not explicitly disclose wherein the modeling loop is executed in the cloud, the modeling data packet acquired for the user is sent to the cloud, and the comfort model is trained in the AI unit in the cloud.
However, Woods does teach wherein the modeling loop is executed in the cloud, the modeling data packet acquired for the user is sent to the cloud, and the comfort model is trained in the AI unit in the cloud ("the server 102 may be configured to generate and train a personal inference model for a vehicle system 104 or a user 110 of the vehicle system 104 based on the received information from the vehicle system 104" [0043]; "in the work flow 320, another data storage 322 and a model storage 324 may be included in the server 102 in a cloud computing context. … the model training component 308 is included in the server 102. the one or more inference models including a personal inference model may be trained by the model training component 308 in the server 102, and transmitted to the vehicle system 104" [0068]; "the computing device 122 may periodically send updated information from the sensors 124, the HVAC controller 126 and/or the user device to the server 102" [0043]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Stevanovic so that the modeling loop is executed in the cloud, the modeling data packet acquired for the user is sent to the cloud, and the comfort model is trained in the AI unit in the cloud as taught by Woods with a reasonable expectation of success because Stevanovic already uploads comfort setting records to a server/cloud and Woods teaches performing personal-inference-model training on the server/cloud and returning the trained model (See at least [0043, 0068]).
Regarding claim 9
Stevanovic discloses transmitting comfort setting records to the server when communication is available, including batching at end of trip ("The comfort setting record may be transmitted immediately after it is generated, or all comfort setting record(s) during a single driving event may be transmitted together at the end of the event" [0044]).
Stevanovic does not explicitly disclose wherein if it is not possible to send the modeling data packet for the user to the cloud, the modeling data packet is stored in the vehicle, and as soon as it is possible to send the modeling data packet for the user to the cloud, the modeling data packet is sent to the cloud and deleted in the vehicle.
However, Woods does teach wherein if it is not possible to send the modeling data packet for the user to the cloud, the modeling data packet is stored in the vehicle, and as soon as it is possible to send the modeling data packet for the user to the cloud, the modeling data packet is sent to the cloud and deleted in the vehicle. ("the data storage 306 in the vehicle system 104 may store recent in-vehicle sensor data 152 and HVAC state data 154 … the model training component 308 is included in the server 102" [0068]; "the computing device 122 may periodically send updated information … to the server 102" [0043]; “the personal model storage 360 and general model storage 362 in the vehicle system 104 may store the most updated models.” [0069]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the Stevanovic cloud modeling method so that if it is not possible to send the modeling data packet for the user to the cloud, the modeling data packet is stored in the vehicle, and as soon as it is possible to send the modeling data packet for the user to the cloud, the modeling data packet is sent to the cloud and deleted in the vehicle as taught by Woods with a reasonable expectation of success because Woods already stores recent data onboard pending server training, and clearing the local copy after successful upload avoids duplicate/conflicting training inputs while conserving vehicle storage (See at least [0043, 0068-0069]).
Regarding claim 10
Stevanovic does not explicitly discloses
wherein all modeling data packets for the user are stored in the vehicle, and periodically sent at predefined times from the vehicle to the cloud.
the comfort model is stored in a model pool in the cloud
and/or the comfort model is stored in a model pool in the cloud and sent from the model pool to the vehicle, and stored in the vehicle
and/or the comfort model is stored in a model pool in the cloud and sent from the model pool in the cloud to the vehicle after it has been fully trained.
However, Woods does teach
wherein all modeling data packets for the user are stored in the vehicle, and periodically sent at predefined times from the vehicle to the cloud (“while the model storage 310 in the vehicle 104 may store more up-to-date inference models.” [0068]; "the computing device 122 may periodically send updated information … to the server 102" [0043]).
the comfort model is stored in a model pool in the cloud ("another data storage 322 and a model storage 324 may be included in the server 102 in a cloud computing context" [0068]; "the personal inference model(s) and general inference model(s) may be stored at a general model storage 356 and a personal model storage 358 in the cloud respectively" [0069] / FIG. 3C);
and/or the comfort model is stored in a model pool in the cloud and sent from the model pool to the vehicle, and stored in the vehicle ("the server 102 may send the trained and/or updated personal inference model back to the vehicle system 104 and the vehicle system 104 may use the personal inference model to predict a target value" [0043]; "One or more general inference models may be sent to a vehicle 104" [0045]);
and/or the comfort model is stored in a model pool in the cloud, and the current comfort model is sent periodically, at predefined times, from the model pool to the vehicle ("the server 102 may further update the personal inference model … based on the periodically received updated information … The server 102 may send the trained and/or updated personal inference model back to the vehicle system 104" [0043]);
and/or the comfort model is stored in a model pool in the cloud and sent from the model pool in the cloud to the vehicle after it has been fully trained ("the one or more inference models including a personal inference model may be trained by the model training component 308 in the server 102, and transmitted to the vehicle system 104" [0068]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to store and distribute the Stevanovic/Woods comfort / personal inference model via cloud model storage as taught by Woods, with a reasonable expectation of success including sending the trained/updated model to the vehicle periodically or after training completes, because Woods teaches cloud model storage and push-back to the vehicle for onboard inference (See at least [0045, 0043, 0068-0069]).
Regarding claim 11
Stevanovic does not explicitly teach
wherein, in the modeling loop: the comfort model is stored in a model pool in the cloud after it has been trained, the comfort model is sent from the model pool to the vehicle and then stored in the vehicle ,
in the setting loop: the setting data packet is sent to the comfort model stored in the vehicle, the comfort model is executed in the vehicle and the individualized settings are predicted locally in the vehicle.
Woods teaches wherein, in the modeling loop: the comfort model is stored in a model pool in the cloud after it has been trained, the comfort model is sent from the model pool to the vehicle and then stored in the vehicle ("the server 102 may send the trained and/or updated personal inference model back to the vehicle system 104" [0043]; cloud model storage 324 / 356 / 358 — [0068], [0069]);
in the setting loop: the setting data packet is sent to the comfort model stored in the vehicle, the comfort model is executed in the vehicle and the individualized settings are predicted locally in the vehicle ("the vehicle system 104 may use the personal inference model to predict a target value for the vehicle system 104 or for the user 110" [0043]; "predicting one or more target values by inputting the collected data into one or more inference models that include a personal inference model" [0003]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Stevanovic's onboard prediction with Woods' train-in-cloud / send-model-to-vehicle / execute-locally pattern, because that architecture lets the vehicle use a freshly trained personal model without requiring continuous cloud connectivity during every prediction cycle, with a reasonable expectation of success (See at least [0003, 0043, 0068-0069]).
Regarding claim 12
Stevanovic does not explicitly teach wherein, in the setting loop: the setting data packet is sent from the vehicle to the comfort model stored in the model pool, the comfort model is executed in the cloud, and the individualized settings are predicted in the cloud and subsequently sent to the vehicle.
However, Woods does teach wherein, in the setting loop: the setting data packet is sent from the vehicle to the comfort model stored in the model pool, the comfort model is executed in the cloud, and the individualized settings are predicted in the cloud and subsequently sent to the vehicle ("the computing device 122 may be configured to send the received information from the sensors 124, the HVAC controller 126 and/or a user device to the server 102 the server 102 may further update the personal inference model for the vehicle system 104 or the user 110 of the vehicle system 104 based on the periodically received updated information from the vehicle system 104 and/or the collected external data 164. The server 102 may send the trained and/or updated personal inference model back to the vehicle system 104 and the vehicle system 104 may use the personal inference model to predict a target value for the vehicle system 104 or for the user 110 of the vehicle system 104.” [0043] see also [0068]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Stevanovic's onboard prediction to include the setting data packet is sent from the vehicle to the comfort model stored in the model pool, the comfort model is executed in the cloud, and the individualized settings are predicted in the cloud and subsequently sent to the vehicle as taught by Woods with a reasonable expectation of success because that architecture lets the vehicle use a freshly trained personal model without requiring continuous cloud connectivity during every prediction cycle (See at least [0043, 0068-0069]).
Regarding claim 13
Stevanovic does not explicitly teach the comfort model is stored after it has been trained in a model pool in the cloud , the comfort model is sent from the model pool to the vehicle and stored in the vehicle , in the setting loop : if it is not possible to send the setting data packet for the vehicle to the cloud , the setting data packet is sent to the comfort model stored in the vehicle , the comfort model is executed in the vehicle ,and the individualized settings are predicted locally in the vehicle , and when it is possible to send the setting data packet for the vehicle to the cloud , the setting data packet is sent from the vehicle to the comfort model stored in the model pool , the comfort model is executed in the cloud , the individualized settings are predicted in the cloud , and subsequently sent to the vehicle.
However, Woods does teach the comfort model is stored after it has been trained in a model pool in the cloud , the comfort model is sent from the model pool to the vehicle and stored in the vehicle , in the setting loop : if it is not possible to send the setting data packet for the vehicle to the cloud , the setting data packet is sent to the comfort model stored in the vehicle , the comfort model is executed in the vehicle ,and the individualized settings are predicted locally in the vehicle , and when it is possible to send the setting data packet for the vehicle to the cloud , the setting data packet is sent from the vehicle to the comfort model stored in the model pool , the comfort model is executed in the cloud , the individualized settings are predicted in the cloud , and subsequently sent to the vehicle (“ In some embodiments, the computing device 122 may periodically send updated information from the sensors 124, the HVAC controller 126 and/or the user device to the server 102. Responsively, the server 102 may further update the personal inference model for the vehicle system 104 or the user 110 of the vehicle system 104 based on the periodically received updated information from the vehicle system 104 and/or the collected external data 164. The server 102 may send the trained and/or updated personal inference model back to the vehicle system 104 and the vehicle system 104 may use the personal inference model to predict a target value for the vehicle system 104 or for the user 110 of the vehicle system 104.” [0043] see also [0068 -0069]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the method of Stevanovic to include Woods' local-execution path (model sent to vehicle) with Woods' cloud-connected path (vehicle data / models on server) into a hybrid offline/online setting loop as claimed — i.e., use the locally stored comfort model when the setting data packet cannot be sent to the cloud, and use cloud execution when connectivity permits — because maintaining HVAC personalization during connectivity outages is a known design goal, both paths are already taught, and selecting between them based on sendability is a predictable failover, with a reasonable expectation of success (See at least [0043, 0068-0069]).
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Stevanovic in view of Woods and further in view of Saeki as applied to claim 1 above, and further in view of Kelly (US20030160104A1) and Fusco et al. (US 6,454,178 B1) hereinafter referred to as Kelly and Fusco.
Regarding claim 4
The combination of Stevanovic, Woods, and Saeki does not explicitly teach wherein the offset settings are created from data in the modeling data packet, stored in the vehicle, and used to set the air conditioner, and as soon as the offset data are incorporated in the comfort model, they are deleted.
However, Kelly does teach the offset settings are created from data in the modeling data packet, stored in the vehicle, and used to set the air conditioner ("adjusting the constant K by an offset value adaptively determined based on operator adjustment of Tset, and adjusting the scheduled blower speed by an offset value adaptively determined based on operator override of the blower speed" [0003]); "The adaptive adjustment of the climate control commands is carried out by CMAC Neural Networks that take into account the environmental conditions and the dynamic state of the system when the respective overrides occur" [0004]; "can record a new training example (desired output for a given set of inputs) at each driver override, and maintain such training example in a queue if there are no further driver overrides of the corresponding input (AF, FAN or MODE) for a predetermined period of time. Then, at an opportune time, the training examples stored in the queue are submitted to the respective CMAC Neural Network for network training" [0025]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the Stevanovic/Woods/Saeki comfort-model method to include offset settings created from data in the modeling data packet, stored in the vehicle, and used to set the air conditioner, as collectively suggested by Kelly's adaptive offsets / queued override training examples that are later submitted into the neural (comfort) model as taught by Kelly, with a reasonable expectation of success because integrating Stevanovic/Woods iterative model updates would apply temporary local offsets from recent modeling-packet / override data until the comfort model is updated to avoid double-counting the same correction in both the offset layer and the updated model (See at least [0003-0004, 0025]).
and as soon as the offset data are incorporated in the comfort model, they are deleted.
However, Fusco does teach clearing temporary offset-like modifications after they have served their purpose / are no longer to be retained ("If the override and modification resulted from temporary meteorological conditions, the modification would be classified as a temporary control point modification in step 208 and would be reset upon ignition off of the vehicle. Accordingly, such adaptations would not be retained in the memory of the HVAC controller when the vehicle was cycled to ignition off" [Col.5-6 and ln 66-67 and ln 1-5]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the Stevanovic/Woods/Saeki comfort-model method to include as soon as the offset data are incorporated in the comfort model, they are deleted, as collectively suggested by Fusco's reset/non-retention of temporary control-point modifications after their interim purpose ends, as taught by Fusco because integrating Stevanovic/Woods iterative model updates would apply temporary local offsets from recent modeling-packet / override data until the comfort model is updated, then clear those interim offsets to avoid double-counting the same correction in both the offset layer and the updated model, with a reasonable expectation of success ([Col.5-6 and ln 66-67 and ln 1-5]).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to AHMED ALKIRSH whose telephone number is (703) 756-4503. The examiner can normally be reached M-F 9:00 am-5:00 pm EST.
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, FADEY JABR can be reached on (571) 272-1516. 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.
/A.A./Examiner, Art Unit 3668
/Fadey S. Jabr/Supervisory Patent Examiner, Art Unit 3668
