Vehicle assistance system for operating support systems and motor vehicles

Abstract

A vehicle assistance system for operating the support systems of a motor vehicle is provided, comprising a conflict detection module for recognizing conflicting settings of different vehicle occupants for the support system, a conflict resolution module for generating a compromise between the conflicting settings, wherein the conflict resolution module includes an AI module, in particular based on an LLM, for simulating the occupants' perspectives and / or communicating with the occupants to detect their perspectives, and a communication module for controlling the support system with the generated compromise. The AI-based resolution of conflicts in the settings of different occupants enables a good adaptation of the vehicle's support systems to the different preferences of multiple occupants.

Description

[0001] The invention relates to a vehicle assistance system by which several occupants of a motor vehicle can operate a support system, and to a motor vehicle with such a vehicle assistance system.

[0002] From DE 10 2023 133 909 A1, a driving assistant with an AI agent for operating support systems of a motor vehicle is known, in which the AI ​​agent receives a reaction from the driver via interior sensors to a control parameter displayed before the start of the journey, in order to individually adapt the respective control parameter to the wishes of the driver and to change it depending on the reaction of the driver.

[0003] There is a constant need to be able to adapt the assistance systems of a motor vehicle to the different preferences of multiple occupants.

[0004] The object of the invention is to demonstrate measures that enable a good adaptation of support systems of a motor vehicle to the different preferences of several occupants.

[0005] The problem is solved according to the invention by a vehicle assistance system with the features of claim 1 and a motor vehicle with the features of claim 10. Preferred embodiments of the invention are specified in the dependent claims and the following description, each of which can individually or in combination represent an aspect of the invention, wherein the scope of protection is determined by the claims.

[0006] One aspect of the invention relates to a vehicle assistance system for operating the support systems of a motor vehicle, comprising a conflict detection module for recognizing conflicting settings of different occupants of the motor vehicle for the support system, a conflict resolution module for generating a compromise between the conflicting settings, wherein the conflict resolution module includes an AI module, in particular based on an LLM, for simulating the occupants' perspectives and / or communicating with the occupants to detect their perspectives, and a communication module for controlling the support system with the generated compromise. The AI-based resolution of conflicts in the settings of different occupants enables a good adaptation of the support systems of a motor vehicle to the different preferences of multiple occupants.

[0007] Modern vehicles offer a wide range of individual personalization options, from climate settings and music preferences to navigation routes. However, it has been recognized that conflicts can arise in vehicles with multiple occupants due to differing comfort, entertainment, and route preferences. Existing systems typically treat comfort parameters in isolation, prioritizing only the driver or individual user profiles. Dynamic detection of conflicts between occupants and intelligent, socially responsible mediation are not currently provided. It has been recognized that the use of AI in vehicles opens up new possibilities for supporting social interactions between occupants through digital assistance systems. It can be taken into account that existing vehicle assistance systems do not automatically detect conflicts arising from competing user preferences, such as...Temperature, music volume, and navigation destinations are all possible settings. However, the system lacks social intelligence, meaning it doesn't offer conflict resolution mechanisms. Collective optimization approaches are absent, and settings are frequently made in favor of one individual. Decision-making is not transparent, leaving users unable to understand why the system makes certain adjustments. Furthermore, the user experience can be inadequate when traveling with multiple people, potentially leading to dissatisfaction with decision-making and / or suboptimal interaction between occupants and the vehicle.

[0008] The invention can provide an AI-based vehicle assistance system that recognizes social dynamics and can actively moderate mediation processes. The invention can include an LLM-based multi-agent system that analyzes the preferences of all occupants, automatically detects conflicts, and—similar to a digital mediator—proposes or autonomously implements a fair and comprehensible compromise. The vehicle assistance system can include, as a first core module, a conflict detection module that can analyze user interactions, voice commands, HMI inputs, and stored preferences, and recognize conflicting comfort or usage requirements, such as "turn it colder" vs. "turn it warmer."As a second core module, the vehicle assistance system can include an LLM-supported mediation module. This module uses a social agent system to simulate the perspectives of the occupants, apply moderation and conflict resolution strategies, and generate a contextual compromise, such as a moderate temperature, playlist mix, or a middle ground on the route. A third core system within the vehicle assistance system can be a communication module for transparent communication. This module can provide explainable recommendations or measures, comprehensible to all occupants, and optionally initiate a dialogic vote, such as, "Would you both like a middle ground?" Typical input dimensions that can lead to conflicts between occupants include the climate (e.g., temperature, airflow, seat heating), audio settings (e.g., music style, volume), and navigation (especially route, waypoints, and lighting and / or seat configurations).

[0009] The vehicle assistance system according to the invention can create a social AI within the vehicle that recognizes and moderates conflicts. Furthermore, a multi-agent architecture can be provided that represents the roles of the occupants and simulates compromises. It can provide LLM-based conflict resolution, which may be inspired in particular by professional mediation techniques. It can achieve fairness optimization for all occupants instead of focusing on driver preference. Explainable and transparent decision-making criteria can be provided, which increases the acceptance of the decision. The vehicle assistance system can also be extended to new dimensions of conflict, such as entertainment, seating position, and privacy.

[0010] The technical structure of the vehicle assistance system according to the invention can include a sensor data and user interface, for example, to capture HMI inputs, e.g., control elements, touchscreens, to detect speech signals and natural language inputs, to recognize profile and seat position data of the occupants, and / or to retrieve historical usage data. The vehicle assistance system can also include a conflict detection module that can perform an LLM-based analysis of competing intentions and / or classify different conflict types, wherein the conflict types can include a temperature conflict, an entertainment conflict, a driver / passenger route conflict, and / or an unspoken conflict recognizable by usage patterns.Furthermore, the vehicle assistance system can include a multi-agent mediator that can simulate each occupant as an individual agent with its own preferences, and in particular a superior meta-agent that can perform compromise finding. Compromise finding can, in particular, include Pareto optimization and prioritization depending on the occupant's role, e.g., driver, child, passenger, etc., as well as apply strategies such as suggestion generation, finding a fair middle ground, and rotating control. The vehicle assistance system according to the invention can include compromise generation in which a LLM (Limited Liaison Manager) proposes measures for conflict resolution and / or implements them automatically.For example, a temperature can be set to an average of different preferences, adaptive volume and / or a dynamically calculated alternative route can be implemented, and / or playlist mixing and / or a "fair share DJ mode" can be used. Furthermore, an additional layer of explanatory functionality can be provided, including short, easily understandable explanations in the HMI or via voice assistant, such as "I have set the temperature to 21°C to accommodate both preferences." System integration can utilize the vehicle's HMI, infotainment system, route planning system, over-the-air updates, and / or machine learning models. This enables the establishment of a social, conflict-resolving driver assistance system that completely reimagines the comfort interactions of multiple occupants.

[0011] In particular, the communication module is designed to output the generated compromise to the inmates. This allows for feedback on the generated compromise to be provided to the inmates.

[0012] Preferably, the communication module is designed to output a justification for the generated compromise, created by the AI ​​model, to the occupants. This justification makes the generated compromise more transparent and understandable for the occupants.

[0013] The conflict detection module is particularly well-designed to detect and assign the settings for the support system to the respective occupant based on user interactions via HMI input and / or voice commands and / or stored preferences. This assignment facilitates conflict resolution within the AI ​​module.

[0014] In particular, the AI ​​module is designed to apply pre-defined and / or trained moderation and conflict resolution strategies when generating a compromise based on the inmates' perspectives. This allows the AI ​​module to be particularly effective in generating the compromise.

[0015] Preferably, the AI ​​module is designed to allow the inmates to vote on the compromise. This allows the inmates to participate transparently in a final decision on the generated compromise.

[0016] Particularly desirable is the inclusion of at least one assistance system, a climate control system, an audio system, a navigation system, a lighting system, a seat configuration system, an entertainment system, and / or a privacy system. In such assistance systems, AI-based compromise finding can lead to a particularly high level of comfort for the occupants.

[0017] In particular, the AI ​​module features a separate AI agent for each inmate to handle the respective perspective of that inmate. This facilitates a discussion of the different inmates' viewpoints within the AI ​​module.

[0018] Preferably, the AI ​​module is designed to prioritize the settings of different inmates based on their role, subjective preferences, and / or perspectives. This allows for better consideration of the varying needs of different inmates.

[0019] Another aspect concerns a motor vehicle with a driver assistance system, which, as described above, can be trained and further developed to operate support systems. AI-based resolution of conflicts in the setting preferences of different occupants enables a better adaptation of a motor vehicle's support systems to the differing preferences of multiple occupants. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] DE 10 2023 133 909 A1

[0002]

Claims

Vehicle assistance system for operating support systems of a motor vehicle, comprising a conflict detection module for detecting conflicting setting specifications of different occupants of the motor vehicle for the support system, a conflict resolution module for generating a compromise between the conflicting setting specifications, wherein the conflict resolution module includes an AI module for simulating the occupants' perspectives and / or communicating with the occupants to detect the occupants' perspectives, and a communication module for controlling the support system with the generated compromise. Vehicle assistance system according to claim 1, wherein the communication module is configured to output the generated compromise to the occupants. Vehicle assistance system according to claim 1 or 2, wherein the communication module is configured to output to the occupants a justification for the generated compromise generated by the AI ​​model. Vehicle assistance system according to one of claims 1 to 3, wherein the conflict detection module is configured to detect the setting specifications for the support system through user interactions with an HMI input and / or voice commands and / or stored preferences of the respective occupants and to assign them to the respective occupant. Vehicle assistance system according to one of claims 1 to 4, wherein the AI ​​module is designed to apply stored and / or trained moderation and conflict resolution strategies when generating the compromise based on the views of the occupants. Vehicle assistance system according to one of claims 1 to 5, wherein the AI ​​module is designed to perform a vote by the occupants for generating the compromise. Vehicle assistance system according to one of claims 1 to 6, wherein the at least one assistance system is a climate control system and / or an audio system and / or a navigation system and / or a lighting system and / or a seat configuration system and / or an entertainment system and / or a privacy system. Vehicle assistance system according to one of claims 1 to 7, wherein the AI ​​module has a separate AI agent for each occupant to handle the respective viewpoint of the assigned occupant. Vehicle assistance system according to one of claims 1 to 8, wherein the AI ​​module prioritizes the setting preferences of different occupants differently based on a role of the respective occupant and / or the subjective preference of the respective occupant and / or the viewpoint of the respective occupant. Motor vehicle with a vehicle assistance system according to one of claims 1 to 9 for operating support systems.