Computer-implemented method for controlling at least one vehicle function and vehicle

Abstract

The invention relates to a computer-implemented method (100) for controlling at least one vehicle function of a vehicle, comprising at least the following steps: extracting (102) a driver intention and general context data regarding at least one comfort situation in the vehicle from provided user data using a large-language model-based semantic module; determining (104) vehicle context data and weather context data using a context data orchestrator; predicting (106) at least one comfort need of the driver based on stored past driver behavior data, the extracted driver intention, the extracted general context data, and the determined vehicle context data and weather context data using a prediction logic; and deriving (108) and providing (110) at least one control command for the at least one vehicle function based on the prediction using a comfort action engine.Method (100) can simplify the operation of at least one vehicle function.

Description

[0001] The invention relates to a computer-implemented method for controlling at least one vehicle function and a vehicle.

[0002] Modern vehicles increasingly offer digital comfort features such as automatic pre-conditioning, seat and mirror adjustments, and personalized navigation suggestions. However, the current state of the art is largely based on reactive systems: the vehicle only performs comfort actions when the user actively intervenes or predefined triggers (e.g., fixed departure times) have been configured.

[0003] Current comfort systems have the following limitations: Lack of semantics: Calendar or appointment content is not understood contextually (e.g., "Meeting in 30 minutes" → time pressure). Lack of proactivity: Comfort functions mostly only react upon user interaction. Insufficient data fusion: Reactive, data-silo-specific logic cannot holistically process weather, appointment structure, location, and personal preferences. Low personalization granularity: Individual comfort patterns (e.g., preferred seat heating temperature depending on the season) are not automatically derived.

[0004] The object of the invention is to provide a computer-implemented method for controlling at least one vehicle function and a vehicle that simplifies the operation of the at least one vehicle function.

[0005] The problem is solved by the features of the independent claims. Advantageous further developments are the subject of the dependent claims and the following description.

[0006] According to one aspect, a computer-implemented method for controlling at least one vehicle function is described, comprising at least the following steps: extracting a driver intention and general context data regarding at least one comfort situation in the vehicle from provided user data using a Large Language Model (LLM)-based semantic module; determining vehicle context data and weather context data using a context data orchestrator; predicting at least one comfort need of the driver based on stored past driver behavior data, the extracted driver intention, the extracted general context data, and the determined vehicle context data and weather context data using a prediction logic; and deriving and providing at least one control command for the at least one vehicle function based on the prediction using a comfort action engine.

[0007] The invention thus represents a novel, intelligent, and highly personalized approach that elevates comfort functions to a proactive, intuitive level. The invention proposes an LLM-based comfort prediction agent. The LLM-based semantics module, the context data orchestrator, the prediction logic, and / or the comfort action engine can be part of the LLM-based comfort prediction agent. The proposed LLM-based comfort prediction agent can be an AI-based agent that automatically derives from various data sources which comfort actions the vehicle is likely to perform. Through natural language semantic processing and historical pattern recognition, the system can proactively prepare the vehicle for its next use.

[0008] An agent is understood to be a program that can act independently and autonomously to a certain extent.

[0009] The agent can aggregate and interpret the following data: Calendar information: appointment title, location details, duration, trip type (e.g., business / private), time pressure; Weather data: temperature, precipitation, humidity; Historical behavior: individual comfort preferences, driving profiles, seat and climate settings, preferred routes; and / or contextual data: parking situation, real-time traffic, battery charge level, vehicle surroundings.

[0010] Based on this analysis, the agent can make autonomous comfort decisions such as: pre-conditioning according to weather and personal history; seat, mirror and steering wheel settings; selection of the likely preferred route; recommendation or reservation of a suitable parking space; and / or preparation of specific application modes (e.g., business mode for business appointments).

[0011] Furthermore, a semantic interpretation of appointment content can take place: The LLM can recognize the meaning, urgency, and / or context. For example, context data might contain the information: "Meeting in 30 minutes in the city center." The process can then be used to select the fastest route, reserve a parking space, and / or initiate moderate pre-conditioning.

[0012] Furthermore, proactive vehicle preparation can take place before the journey begins. The vehicle then does not only act upon activation, but prepares itself autonomously for anticipated requirements.

[0013] Furthermore, multi-data fusion can take place at a semantic level. This allows data from the calendar, behavior, weather, location, and / or traffic to be integrated into a holistic comfort model.

[0014] Furthermore, adaptive personalization can be provided, whereby the system can learn and thus continuously optimize individual comfort preferences.

[0015] Furthermore, a modular agent architecture can be provided. The comfort prediction agent can integrate additional modules (e.g., stress level analysis, workload detection) and be extensible.

[0016] The LLM-based semantics module can, for example, extract intention and context from calendar and text data using prompting, embeddings, and ontologies for convenience situations.

[0017] The context data orchestrator can aggregate weather, traffic, vehicle and location data in real time.

[0018] The prediction logic can combine LLM reasoning and probabilistic models to predict comfort needs.

[0019] The Comfort Action Engine can derive specific vehicle actions, e.g., pre-conditioning, seat settings, routing, parking suggestions.

[0020] Furthermore, in some implementation examples, a privacy layer may be provided, whereby on-device processing of sensitive user data with cloud optionality for non-sensitive models can be provided.

[0021] The invention thus provides an intelligent system that proactively recognizes a driver's comfort needs, translates them into actions, and prepares the vehicle, e.g., before the journey begins or adjusts it during the journey.

[0022] According to some embodiments, it is conceivable that the at least one vehicle function may include a vehicle air conditioning system, a seat adjustment, a mirror adjustment, a steering wheel adjustment, a navigation device, a communication device, in particular for reserving a parking space at a destination, and / or an application mode control for the vehicle.

[0023] According to some examples, it is conceivable that the user data may contain calendar information, in particular appointment titles, location details, durations, professional and private appointment types and / or date and time information.

[0024] According to some embodiments, it is conceivable that the stored past driver behavior data may contain data on individual comfort preferences, driving profiles, seat and climate settings and / or preferred routes.

[0025] According to some examples, it is conceivable that the weather context data could include temperature data, precipitation data and / or humidity data.

[0026] According to some embodiments, it is conceivable that the general context data could include data on a parking situation at a destination, real-time traffic volume, the vehicle's state of charge and / or the vehicle's surroundings.

[0027] According to some embodiments, it is conceivable that the step could include: predicting at least one comfort need, predicting a climate setting, a seat setting, a mirror setting, a steering wheel setting, selecting a preferred route, recommending and / or reserving a parking space at a destination, and / or preparing an application mode of the vehicle.

[0028] According to a second aspect, a computer program product is described, comprising instructions that, when the program is executed by a computer, cause it to perform the steps of the procedure according to the preceding description.

[0029] The advantages, effects, and further developments of the computer program product result from the advantages, effects, and further developments of the method described above. Therefore, reference is made to the preceding description in this regard. A computer program product can be understood, for example, as a data carrier on which a computer program element is stored, containing instructions executable by a computer. Alternatively or additionally, a computer program product can also be understood, for example, as a permanent or volatile data storage medium, such as flash memory or main memory, that contains the computer program element. However, this does not exclude other types of data storage media that contain the computer program element.

[0030] According to a third aspect, a vehicle is described, comprising at least one vehicle function and at least one computer that is trained to perform the steps of the procedure according to the preceding description.

[0031] The advantages, effects, and further developments of the vehicle result from the advantages, effects, and further developments of the procedure described above. To avoid repetition, reference is therefore made to the preceding description in this regard.

[0032] The invention is described below with reference to an exemplary embodiment and the accompanying drawing. It shows: Fig. 1. A flowchart of the computer-implemented procedure.

[0033] The computer-implemented method for controlling at least one vehicle function of a vehicle is defined according to Fig. 1 hereinafter referred to in its entirety by reference numeral 100.

[0034] The computer-implemented method 100 for controlling at least one vehicle function of a vehicle can be carried out in conjunction with a vehicle. The vehicle can have at least one vehicle function and at least one computer, which can be configured to perform the steps of method 100.

[0035] The at least one vehicle function may include vehicle air conditioning, seat adjustment, mirror adjustment, steering wheel adjustment, navigation device, communication device, in particular for reserving a parking space at a destination, and / or application mode control for the vehicle.

[0036] According to step 102, a driver intention and general contextual data regarding at least one comfort situation in the vehicle can be extracted from the provided user data. This extraction can be performed using a large-language model-based semantic module.

[0037] User data may contain calendar information, in particular appointment titles, locations, durations, types of professional and private appointments and / or date and time information.

[0038] The general contextual data may include data on a parking situation at a destination, real-time traffic volume, vehicle charge level and / or vehicle environment.

[0039] In a further step, vehicle context data and weather context data can be determined. This data can be determined using a context data orchestrator.

[0040] The weather context data can include temperature data, precipitation data and / or humidity data.

[0041] Furthermore, in step 106, at least one comfort requirement of the driver will be predicted based on stored past driver behavior data, the extracted driver intention, the extracted general context data, and the determined vehicle and weather context data. The prediction can be executed using a prediction logic.

[0042] The stored past driver behavior data may contain data on individual comfort preferences, driving profiles, seat and climate settings and / or preferred routes.

[0043] A prediction can include information on a climate setting, a seat setting, a mirror setting, a steering wheel setting, a selection of a preferred route, a recommendation and / or reservation of a parking space at a destination and / or a preparation of an application mode of the vehicle.

[0044] Furthermore, in step 108, at least one control command for at least one vehicle function is derived based on the prediction. This control command can be provided in step 110, for example, to a controller for the at least one vehicle function. Steps 108 and 110 can be executed using a comfort action engine.

[0045] The example described above does not in any way limit the invention. Rather, the invention can be modified in numerous ways. All features of the invention described above can be essential to the invention, either alone or in combination.

Claims

[1] Computer-implemented method (100) for controlling at least one vehicle function of a vehicle, comprising at least the following steps: a. Extracting (102) a driver intention and general context data regarding at least one comfort situation in the vehicle from provided user data using a large language model-based semantic module; b. Determining (104) vehicle context data and weather context data using a context data orchestrator; c. Predictions (106) of at least one driver comfort need based on stored past driver behavior data, the extracted driver intention, the extracted general context data, and the determined vehicle context data and weather context data using a prediction logic; and d. Deriving (108) and providing (110) at least one control command for at least one vehicle function based on prediction by means of a comfort action engine. [2] Computer-implemented method (100) according to claim 1, characterized by that at least one vehicle function includes a vehicle air conditioning system, a seat adjustment, a mirror adjustment, a steering wheel adjustment, a navigation device, a communication device, in particular for reserving a parking space at a destination, and / or an application mode control for the vehicle. [3] Computer-implemented method (100) according to claim 1 or 2, characterized by that the user data contains calendar information, in particular appointment titles, location details, durations, professional and private appointment types and / or date and time information. [4] Computer-implemented method (100) according to any one of the preceding claims, characterized bythat the stored past driver behavior data contains data on individual comfort preferences, driving profiles, seat and climate settings and / or preferred routes. [5] Computer-implemented method (100) according to any one of the preceding claims, characterized by that the weather context data includes temperature data, precipitation data and / or humidity data. [6] Computer-implemented method (100) according to any one of the preceding claims, characterized by that the general context data includes data on a parking situation at a destination, real-time traffic volume, vehicle charge level and / or vehicle environment. [7] Computer-implemented method (100) according to any one of the preceding claims, characterized by, that the step: Prediction (106) includes at least one comfort need, a prediction of a climate setting, a seat setting, a mirror setting, a steering wheel setting, a selection of a preferred route, a recommendation and / or reservation of a parking space at a destination and / or a preparation of an application mode of the vehicle. [8] Computer program product comprising instructions which, when the program is executed by a computer, cause it to perform the steps of the method (100) according to any one of claims 1 to 7. [9] Vehicle comprising at least one vehicle function and at least one computer configured to perform the steps of the method (100) according to any one of claims 1 to 7.

Images