DETAILED ACTION
Claims 1-18 received on 05/22/2025 are considered in this office action. Claims 1-18 are pending for examination.
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 .
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 05/22/2025 is being considered by the examiner.
Claim Objections
Claim 1 is objected to because of the following informalities: the second language large model should read the second language model. Appropriate correction is required.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation(s) are:
Claims 11-13 and 17-18: control device (generic placeholder) implements (function)
Claim 14: manual input device (generic placeholder) confirms or denies (function)
Claims 15-16: supervising device (generic placeholder) examines (function)
Because this/these claim limitations are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, they are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
Regarding control device and supervising device, they are interpreted to cover the corresponding structure of processor and equivalents thereof as supported by paragraph [0025] of the specification reproduced below:
[0025] The control device for the motor vehicle also belongs to the disclosure. The control device can comprise a data processing device or a processor device, which is configured to perform an embodiment of the method according to the disclosure. Hereto, the processor device can comprise at least one microprocessor and/or at least one microcontroller and/or at least one FPGA (Field Programmable Gate Array) and/or at least one DSP (Digital Signal Processor). As the microprocessor, a CPU (Central Processing Unit), a GPU (Graphical Processing Unit) or an NPU (Neural Processing Unit) can in particular be respectively used. Furthermore, the processor device can comprise program code, which is configured, upon execution by the processor device, to perform the embodiment of the method according to the disclosure. The program code can be stored in a data memory of the processor device. The processor device can, e.g., be based on at least one circuit board and/or on at least one SoC (System on Chip).
Regarding manual input device, it is interpreted to cover the corresponding structure of touchscreen or button and equivalents thereof as supported by paragraph [0025] of the specification reproduced below:
[0011] According to an advantageous embodiment, a, preferably manual, input is received (in the next, optionally concluding step), by which it is confirmed that driving the motor vehicle by the first large language model sufficiently complies with the previous inputs in the opinion of an inputting person. Thus, in other words, a type of confirmation knob (or "button" on user interface like a touchscreen) can be pressed here. In this manner, it is communicated - wherein the second large language model does not have to be used anymore - to the first large language model that driving the motor vehicle can be continued as started
If applicant does not intend to have these limitations interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitations to avoid them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitations recite sufficient structure to perform the claimed function so as to avoid them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 4, 9 and 14-16 rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claims 14 and 15 recites the limitation "The motor vehicle according to claim 10". There is insufficient antecedent basis for this limitation in the claim, as claim 10 is directed to a method. For examination purposes, the Examiner will interpret The motor vehicle according to claim 10 as The motor vehicle according to claim [[10]] 11.
Claim 16 is dependent on claim 15, and fail to cure the deficiencies thereof, thus are rejected on the same basis.
The terms “practical” in claim 9 is a relative term which renders the claim indefinite. The term “practical” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention. For examination purposes, the Examiner will interpret practical circumstances as constraints, as supported by para. [0014] of the specification.
[0014] Alternatively or additionally, it can be provided that a supervision is effected whether the implementation of the instruction output by the first large language model complies with further conditions, such as for instance provided regulations: For example, if the user wishes that the motor vehicle is to drive with a speed of 60 km/h, but the motor vehicle is in an area with speed limitation to 30 km/h, a correction is to be performed by the automatic supervision and a command (preferably in the form of a voice command) is to be given to the first large language model, which corrects it. Similarly, it can also be examined if planned trajectories are in conflict with further objects, which the user possibly has not considered with his command. In this case, it should brake and an alternative trajectory can be planned, which could be situated as close as possible to the trajectory desired by the user
The terms “sufficient” in claims 9 and 14 is a relative term which renders the claim indefinite. The term “sufficient” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention. For examination purposes, the Examiner will interpret sufficient compliance as
The terms “opinion” in claims 4 and 14 is a relative term which renders the claim indefinite. The term “opinion” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention. For examination purposes, the Examiner will interpret in an opinion of an inputting person as based on an input of an inputting person.
[0011] According to an advantageous embodiment, a, preferably manual, input is received (in the next, optionally concluding step), by which it is confirmed that driving the motor vehicle by the first large language model sufficiently complies with the previous inputs in the opinion of an inputting person. Thus, in other words, a type of confirmation knob (or "button" on user interface like a touchscreen) can be pressed here
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 text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
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 1-18 are rejected under 35 U.S.C. 103 as being unpatentable over Ng-Thow-Hing (US 20140365228 A1), in view of Cui (NPL- Receive, Reason, and React: Drive as You Say with Large Language Models in Autonomous Vehicles), and further in view of Luan (NPL- Enhancing Robot Task Planning and Execution through Multi-Layer Large Language Models).
Cui was cited in the IDS received on 05/22/2025.
Regarding claim 1, Ng-Thow-Hing teaches a method for at least partially autonomously driving a motor vehicle (FIG. 2; para. [0017]: “autonomous control system configured to control the motion of the vehicle”), comprising:
providing a first (FIG. 2; para. [0037]: “FIG. 2 illustrates an exemplary vehicle interaction system 200 […] three autonomous control subsystems 214, 216, 218, and a command interpreter 220 configured to pass instructions to these subsystems. […] the command interpreter 220 may be implemented as a standalone element or the autonomous control modules 214, 216, 218 may be omitted. Further, it will be appreciated from the following description that the command interpreter 220 may be used to translate input information from virtually any group of input subsystems into instructions for virtually any group of "output" subsystems and, as such, the various subsystems disclosed are examples illustrating the various utilities of the command interpreter 220”);
driving the motor vehicle by way of the first (para. [0038]: “Autonomous control of a vehicle may be separated into multiple levels of control that are associated with different goals and detail granularities. For example, according to some models, autonomous control may be split into three tiers: strategic control 214, tactical control 216, and operational control 218” […] strategic control is the highest level of control and relates to the user's end goals in operation of a vehicle […] Operational control is the lowest level of control and is directed to the operation of the vehicle to achieve the goals of the tactical control […] such as monolithic implementations of all control levels into a single module or implementations according to different control paradigms than that described here will be apparent”);
receiving a first voice input by the second (FIG. 2; FIG. 4; FIG. 6; para. [0064]: “to step 420 where the command interpreter receives voice input from the user indicating a command to the vehicle interaction system”; para. [0039]: “command interpreter which receives instruction and context input and subsequently translates this information into instructions for the other subsystems”; para. [0066]: “the command interpreter may utilize a more linear, rules-based machine learning approach as will be described below”); and
in at least one query iteration, outputting a query by the second (FIG. 6; para. [0075]: “If, in step 620, the command interpreter is unable to resolve the correlation to a single command, the method proceeds to step 625 where the command interpreter outputs a request for clarification to the user. For example, the command interpreter may generate an instruction to the driver HUD subsystem or a speaker subsystem to present the request to the user.”);
after a last query iteration, transferring an instruction output by the second (FIG. 6; para. [0078]: “step 665 where the command interpreter identifies the target subsystem for the instruction by, for example, referring to the retrieved instruction metadata. Then, in steps 670 and 675, the command interpreter generates and transmits the instruction to the target subsystem. For example, the command interpreter may perform a function call that passes a parameter values to the subsystem. As another example, the command interpreter may construct an XML object or other data object identifying the command and parameter values, and subsequently pass the object to the target subsystem via any known method such as, for example, inter-process communication via the operating system”); and
driving the motor vehicle based on an instruction output by the first (para. [0038]: “Operational control is the lowest level of control and is directed to the operation of the vehicle to achieve the goals of the tactical control. For example, operational control may focus on acceleration, braking, and steering while staying in the lane and avoiding collision”), but fails to specifically teach a large language model.
However, in the same field of endeavor, Cui teaches receiving a voice input and driving the motor vehicle by way of the large language model (FIG. 1; FIG. 2; pg. 1 right col: “in the context of fully autonomous vehicles, the LLMs’ capabilities could even extend to taking charge of the vehicle and executing the instructed commands.”; pg 5: “The vehicle is afforded five distinct operational actions. These are 0, enabling a leftward lane change (’LANE LEFT’); 1, maintaining the current state or position (’IDLE’); 2, facilitating a rightward lane change (’LANE RIGHT’); 3, accelerating the current speed by 2 meters per second; and 4, reducing the speed by the same magnitude. This structured environment and action set are designed to test the LLM’s proficiency in varied highway scenarios”; pg 6 right col: “For instance, by stating “drive more conservatively” to the LLMs, you can trigger a change in the autonomous driving behavior towards a more conservative mode. This change is aimed at aligning the driving style closer to your comfort zone, which enhances the user experience”, wherein Cui teaches a LLM model capable of receiving a voice input and driving the motor vehicle).
Ng-Thow-Hing and Cui are considered analogous art to the claimed invention because they are in the same field of endeavor of performing vehicle functions based on voice command of user. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have a simple substitution of the command interpreter and subsystems of Ng-Thow-Hing with LLM models of Cui because both models are known to interpret voice commands and drive vehicle according to the command and would have obtained the predictable result of vehicles can dynamically interact with passengers and adapt to their preferences (Cui, Abstract). Ng-Thow-Hing in view Cui fails to specifically teach a second large language model.
However, in the same field of endeavor, Luan teaches transferring an instruction output by the second language large model to the first large language model (FIG. 2; FIG. 7; pg 5 Section 3.1: “The first large language model is responsible for comprehending human instructions and subsequently generating an executable coarse-grained plan for the robot. […] the outcomes of the coarse-grained task decomposition were input into the subsequent functional module. This functional module comprises a large language model and a visual model […] This iterative decomposition yields more precise fine-grained tasks, aligning with the specific nuances of the environment [25]. The generation of a sequence of fine-grained tasks is achieved by organizing these tasks based on the general knowledge embedded in the large language model”; pg. 2 para 4: “Consequently, the large language model can output tasks at the semantic level that precisely control the robot to execute the corresponding actions” ).
Luan is considered analogous art to the claimed invention because it is in the same field of endeavor of using LLM to perform human commands. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the single LLM of of Ng-Thow-Hing in view of Cui with a multi-layer LLM of Luan. Doing so will optimize the effectiveness and accuracy of motion planning by reducing the complexity of tasks (Luan Abstract, Section 1), as a single LLM leads to considerable latency and errors.
Regarding claim 2, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the method according to claim 1. The combination of Ng-Thow-Hing in view of Cui further teaches wherein the second large language model is provided in the motor vehicle (Ng-Thow-Hing Fig. 1-3; Ng-Thow-Hing Abstract: “vehicle control system in a vehicle”; Ng-Thow-Hing para. [0037]: “FIG. 2 illustrates an exemplary vehicle interaction system 200. As shown, the system 200 is centered around a vehicle control system 210 including a manual control 212 subsystem, three autonomous control subsystems 214, 216, 218, and a command interpreter 220 configured to pass instructions to these subsystems”; Cui: “LLM”).
Regarding claim 3, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the method according to claim 1. Ng-Thow-Hing further teaches wherein the input received in response to the query is a further voice input (para. [0002]: “generally to interpretation of user input and, more particularly but not exclusively, to interpretation of voice commands and gestures by a vehicle driver or passenger”).
Regarding claim 4, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the method according to claim 1. Ng-Thow-Hing further teaches further comprising: receiving an input which confirms that the driving of the motor vehicle by the first large language model sufficiently complies with previous inputs in an opinion of an inputting person (para. [0082]: “command interpreter may infer from various events that a previous decision was incorrect and consequently use the decision and subsequent user actions to further refine the learned information base. For example, when the user indicates that an action taken by the command interpreter is incorrect (e.g., the user shakes his head or says “no, that isn't right”) or issues a contradictory command shortly after the command interpreter sends an instruction to another subsystem (e.g., the user says “turn left on Elm street” after the command interpreter previously interpreted user input into an instruction to turn right on Elm street), the command interpreter may determine that the previous decision was incorrect.”).
Regarding claim 5, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the method according to claim 4. Ng-Thow-Hing teaches wherein the input which confirms that the driving of the motor vehicle by the first large language model sufficiently complies with previous inputs is a manual input (para. [0034]: “includes a microphone 130 and a camera 140 for receiving input from the user 110 and a display 150 for displaying output to the user. The display 150 may also include a touchscreen for receiving additional input from the user”; para. [0055]: “touchscreen 355 for receiving input from the user”; para. [0099]: “the command interpreter may output a confirmation message to the user.”; para. [0082]: “For example, when the user indicates that an action taken by the command interpreter is incorrect (e.g., the user shakes his head or says “no, that isn't right”)”, wherein various input methods can be made, which comprises touch input or manual input).
Regarding claim 6, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the method according to claim 1. Ng-Thow-Hing and Cui further teaches further comprising: receiving an input that revokes previous inputs, wherein a return to a driving style effected by the first large language model is caused before receiving the first voice input (Cui pg 6 right col: “For instance, by stating “drive more conservatively” to the LLMs, you can trigger a change in the autonomous driving behavior towards a more conservative mode. This change is aimed at aligning the driving style closer to your comfort zone, which enhances the user experience. […] On a different day, if you are running late and desire a swifter transit, instructing the LLMs with a “drive more aggressively” command would prompt a shift to faster and more aggressive driving behavior. This scenario illuminates the potential for dynamic adjustments in driving styles based on real-time user feedback”; Ng-Thow-Hing para. [0087]: “Using the rules illustrated, the vehicle would likely begin to turn left and the user could then follow up with a correction such as speaking the phrase “No, I meant turn right.” The command interpreter, in addition to sending a new instruction to the autonomous control subsystem, may reevaluate the previously issued left turn instruction as incorrect”, wherein based on the feedback the driving style can be returned).
Regarding claim 7, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the method according to claim 6. Ng-Thow-Hing wherein the input that revokes previous inputs is a manual input (para. [0034]: “includes a microphone 130 and a camera 140 for receiving input from the user 110 and a display 150 for displaying output to the user. The display 150 may also include a touchscreen for receiving additional input from the user”; para. [0055]: “touchscreen 355 for receiving input from the user”; para. [0099]: “the command interpreter may output a confirmation message to the user.”; para. [0082]: “For example, when the user indicates that an action taken by the command interpreter is incorrect (e.g., the user shakes his head or says “no, that isn't right”)”, wherein various input methods can be made, which comprises manual input).
Regarding claim 8, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the method according to claim 1. Cui further teaches further comprising: effecting an automatic supervision, based on sensor data of sensors of the motor vehicle, that determines whether implementation of the instruction output by the first large language model sufficiently complies with certain conditions (FIG. 2 Unsafe Overtaking Scenario; pg 6 left col: “Our experiments have revealed significant distinctions between LLMs prompted using the chain-of-thought prompting method and those trained using standard prompting. In the case of chain-of-thought prompting, the vehicle can intelligently assess the safety of overtaking the front vehicle. If the conditions are regarded safe, LLMs trained through this method will execute the most reasonable actions: they observe neighboring lanes for clearance, change lanes, accelerate to the speed limit to overtake, and subsequently return to their original lane when the lanes are clear. Conversely, if the situation is assessed as unsafe, these LLMs instruct the vehicle to accelerate to the speed limit. As the vehicle approaches the front one, it continuously evaluates the situation for overtaking opportunities. Upon determining that overtaking is not safe, it promptly reduces speed and abandons the overtaking attempt”).
Regarding claim 9, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the method according to claim 8, wherein the automatic supervision determines whether implementation of the instruction output by the first large language model is in sufficient compliance with the inputs by an inputting person and/or is in compliance with practical circumstances (FIG. 2 Unsafe Overtaking Scenario; pg 6 left col: “Our experiments have revealed significant distinctions between LLMs prompted using the chain-of-thought prompting method and those trained using standard prompting. In the case of chain-of-thought prompting, the vehicle can intelligently assess the safety of overtaking the front vehicle. If the conditions are regarded safe, LLMs trained through this method will execute the most reasonable actions: they observe neighboring lanes for clearance, change lanes, accelerate to the speed limit to overtake, and subsequently return to their original lane when the lanes are clear. Conversely, if the situation is assessed as unsafe, these LLMs instruct the vehicle to accelerate to the speed limit. As the vehicle approaches the front one, it continuously evaluates the situation for overtaking opportunities. Upon determining that overtaking is not safe, it promptly reduces speed and abandons the overtaking attempt”, wherein safe and unsafe indicates sufficient compliance with the inputs by an inputting person and/or is in compliance with practical circumstances).
Regarding claim 10, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the method according to claim 1. The combination of Ng-Thow-Hing in view of Cui further teaches wherein the first large language model is trained with training data (Cui pg 3: “The study by [6] exemplifies that LLMs can be fine-tuned to exhibit enhanced performance, particularly in tasks with limited training data.”, wherein LLMs are trained with training data), and the method further comprises:
receiving an instruction that instructs a change of a previous driving style according to driving by the first large language model (Cui pg 6 right col: “For instance, by stating “drive more conservatively” to the LLMs, you can trigger a change in the autonomous driving behavior towards a more conservative mode. This change is aimed at aligning the driving style closer to your comfort zone, which enhances the user experience”, wherein “drive more conservatively” indicates change of a previous driving style); and
implementing the instruction that instructs the change of the previous driving style according to driving by the first large language model, wherein, after a lapse of time and/or after termination of a driving situation and/or due to a user input, the change instructed by the instruction is canceled (Cui pg 6 right col: “On a different day, if you are running late and desire a swifter transit, instructing the LLMs with a “drive more aggressively” command would prompt a shift to faster and more aggressive driving behavior. This scenario illuminates the potential for dynamic adjustments in driving styles based on real-time user feedback”, wherein “drive more aggressively” drive more aggressively” indicates change of the previous driving style) and an entirety of the training data remains unchanged (pg 7 left col: “One of the most evident benefits of in-context learning in LLMs is its inherent adaptability. Traditional decision-making models often need retraining from scratch or employing a pre trained model when faced with a new scenario. Our high way overpassing and merging driving experiments exhibited the strengths of in-context learning. […] In-context learning in LLMs presents a more cost-effective alternative. By simply providing contextual guidance, we can recalibrate the model’s behavior, reducing both computational and financial overheads”, wherein LLM models are given contextual guidance unlike traditional decision-making models often need retraining, thus indicating entirety of the training data remains unchanged).
Regarding claim 11, it recites a motor vehicle, comprising:
a control device (Ng-Thow-Hing FIG. 2; Ng-Thow-Hing para. [0017]: “autonomous control system configured to control the motion of the vehicle”; para. [0061]: “the term “subsystem” or “system” may refer to both co-resident system (system that utilize the same processor or other hardware as another system)”) with first interface (Ng-Thow-Hing para. [0034]: “includes a microphone 130 and a camera 140 for receiving input from the user 110 and a display 150 for displaying output to the user) and second interface (Luan FIG. 2; Luan FIG. 7) that performs claim limitations similar to those of the method according to claim 1, and thus is rejected on the same basis.
Regarding claim 12, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the motor vehicle according to claim 11. Ng-Thow-Hing further teaches wherein the second large language model is a part of the motor vehicle (Abstract: “a vehicle control system in a vehicle”), and wherein the second large language model, in operation, outputs a query via the first interface in at least one query iteration and receives an input in response to the query, via the first interface, before an instruction is transferred to the control device via the second interface (FIG. 6; para. [0075]: “If, in step 620, the command interpreter is unable to resolve the correlation to a single command, the method proceeds to step 625 where the command interpreter outputs a request for clarification to the user. For example, the command interpreter may generate an instruction to the driver HUD subsystem or a speaker subsystem to present the request to the user.”; para. [0074]: “step 610 where the command interpreter receives user input. For example, the command interpreter may receive input from the voice recognition subsystem or gesture recognition subsystem.”).
Regarding claim 13, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the motor vehicle according to claim 12. Ng-Thow-Hing further teaches wherein the input received in response to the query is one of the voice inputs (FIG. 6; para. [0075]: “If, in step 620, the command interpreter is unable to resolve the correlation to a single command, the method proceeds to step 625 where the command interpreter outputs a request for clarification to the user. For example, the command interpreter may generate an instruction to the driver HUD subsystem or a speaker subsystem to present the request to the user.”; para. [0074]: “step 610 where the command interpreter receives user input. For example, the command interpreter may receive input from the voice recognition subsystem or gesture recognition subsystem.”; para. [0095]: “the voice input additionally identifies any environmental variables”).
Regarding claim 14, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the motor vehicle according to claim 10. Ng-Thow-Hing and Cui further comprising: a manual input device that, in operation, confirms or denies input in a query iteration (Ng-Thow-Hing para. [0034]: “includes a microphone 130 and a camera 140 for receiving input from the user 110 and a display 150 for displaying output to the user. The display 150 may also include a touchscreen for receiving additional input from the user”; Ng-Thow-Hing para. [0055]: “touchscreen 355 for receiving input from the user”; para. [0099]: “the command interpreter may output a confirmation message to the user.”; Ng-Thow-Hing para. [0082]: “For example, when the user indicates that an action taken by the command interpreter is incorrect (e.g., the user shakes his head or says “no, that isn't right”)”, wherein various input methods can be made, which comprises touch input or manual input) and/or confirms that driving the motor vehicle by the first large language model sufficiently complies with previous inputs in an opinion of an inputting person and/or revokes the previous inputs and causes a return to a driving style effected by the first large language model before a first one of the voice inputs is received (Cui pg 6 right col: “For instance, by stating “drive more conservatively” to the LLMs, you can trigger a change in the autonomous driving behavior towards a more conservative mode. This change is aimed at aligning the driving style closer to your comfort zone, which enhances the user experience. […] On a different day, if you are running late and desire a swifter transit, instructing the LLMs with a “drive more aggressively” command would prompt a shift to faster and more aggressive driving behavior. This scenario illuminates the potential for dynamic adjustments in driving styles based on real-time user feedback”; Ng-Thow-Hing para. [0087]: “Using the rules illustrated, the vehicle would likely begin to turn left and the user could then follow up with a correction such as speaking the phrase “No, I meant turn right.” The command interpreter, in addition to sending a new instruction to the autonomous control subsystem, may reevaluate the previously issued left turn instruction as incorrect”, wherein based on the feedback the driving style can be returned; Ng-Thow-Hing para. [0082]: “command interpreter may infer from various events that a previous decision was incorrect and consequently use the decision and subsequent user actions to further refine the learned information base. For example, when the user indicates that an action taken by the command interpreter is incorrect (e.g., the user shakes his head or says “no, that isn't right”) or issues a contradictory command shortly after the command interpreter sends an instruction to another subsystem (e.g., the user says “turn left on Elm street” after the command interpreter previously interpreted user input into an instruction to turn right on Elm street), the command interpreter may determine that the previous decision was incorrect”).
Regarding claim 15, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the motor vehicle according to claim 10. Ng-Thow-Hing in view of Cui and further in view of Luan further teaches further comprising: a supervising device that, in operation, receives data from sensors of the motor vehicle, and examines if an instruction transferred to the control device can be currently implemented (Cui FIG. 2 Unsafe Overtaking Scenario; Cui pg 6 right col: “Conversely, if the situation is assessed as unsafe, these LLMs instruct the vehicle to accelerate to the speed limit. As the vehicle approaches the front one, it continuously evaluates the situation for overtaking opportunities. Upon determining that overtaking is not safe, it promptly reduces speed and abandons the overtaking attempt”; Ng-Thow-Hing para. [0038]: “implementation of autonomous control may also involve separating these tiers among different modules or processes that each perform the assigned functionality […] monolithic implementations of all control levels into a single module or implementations according to different control paradigms than that described here will be apparent”, wherein “evaluates the situation” indicates receives data from sensors of the motor vehicle, and examines if an instruction transferred to the control device can be currently implemented, and wherein Ng-Thow-Hing and Luan teaches different modules for different roles).
Regarding claim 16, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the motor vehicle according to claim 15. The combination of Ng-Thow-Hing in view of Cui and further in view of Luan further teaches wherein the supervising device, in operation, transfers correcting instructions to the first large language model (FIG. 2 Unsafe Overtaking Scenario; pg 6 right col: “Conversely, if the situation is assessed as unsafe, these LLMs instruct the vehicle to accelerate to the speed limit. As the vehicle approaches the front one, it continuously evaluates the situation for overtaking opportunities. Upon determining that overtaking is not safe, it promptly reduces speed and abandons the overtaking attempt”, wherein “abandons the overtaking attempt” indicates correcting instructions to the first large language model).
Regarding claim 17, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the motor vehicle according to claim 11. The combination of Ng-Thow-Hing in view of Cui further teaches wherein the first large language model is trained with training data (Cui pg 3: “The study by [6] exemplifies that LLMs can be fine-tuned to exhibit enhanced performance, particularly in tasks with limited training data.”, wherein LLMs are trained with training data),
wherein the first large language model, in operation, receives an instruction that causes a previous driving style according to driving by the first large language model to be changed and implements the instruction (Cui pg 6 right col: “For instance, by stating “drive more conservatively” to the LLMs, you can trigger a change in the autonomous driving behavior towards a more conservative mode. This change is aimed at aligning the driving style closer to your comfort zone, which enhances the user experience […] On a different day, if you are running late and desire a swifter transit, instructing the LLMs with a “drive more aggressively” command would prompt a shift to faster and more aggressive driving behavior. This scenario illuminates the potential for dynamic adjustments in driving styles based on real-time user feedback”, wherein “drive more conservatively or agressively” indicates change of a previous driving style), and
wherein the first large language model remains unchanged with respect to an entirety of the training data after implementation of the instruction (pg 7 left col: “One of the most evident benefits of in-context learning in LLMs is its inherent adaptability. Traditional decision-making models often need retraining from scratch or employing a pre trained model when faced with a new scenario. Our high way overpassing and merging driving experiments exhibited the strengths of in-context learning. […] In-context learning in LLMs presents a more cost-effective alternative. By simply providing contextual guidance, we can recalibrate the model’s behavior, reducing both computational and financial overheads”, wherein LLM models are given contextual guidance unlike traditional decision-making models often need retraining, thus indicating entirety of the training data remains unchanged).
Regarding claim 18, Ng-Thow-Hing in view of Cui and further in view of Luan teaches the motor vehicle according to claim 17. Cui further teaches wherein the instruction is in language form (pg 6 right col: “For instance, by stating “drive more conservatively” to the LLMs, you can trigger a change in the autonomous driving behavior towards a more conservative mode. This change is aimed at aligning the driving style closer to your comfort zone, which enhances the user experience. On a different day, if you are running late and desire a swifter transit, instructing the LLMs with a “drive more aggressively” command would prompt a shift to faster and more aggressive driving behavior.”).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
LEE (US 20250087208 A1) teaches classifying the intent of an utterance in consideration of context surrounding a vehicle and a driver.
CHEN (GB 2632656 A) teaches method of providing improved systems and methods for autonomous driving that provide for enhanced explainability.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ANDREW S KIM whose telephone number is (571)272-7356. The examiner can normally be reached Mon - Fri 8AM - 5PM.
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, James J Lee can be reached on (571) 270-5965. 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.
/ANDREW SANG KIM/Examiner, Art Unit 3668
Prosecution Insights