Method for the automated coordination of a maintenance process for a motor vehicle

Abstract

Method for the automated coordination of a maintenance process for a motor vehicle, characterized by the following features: - Electronic reception of technical diagnostic data representing the condition of vehicle components via a data interface, - Semantic analysis of the received diagnostic data using a software agent based on a large language model, whereby the language model automatically derives a specific maintenance requirement and a list of spare parts required for maintenance as a machine-readable instruction from the technical diagnostic data, - Automated sending of electronic queries to external computer systems of workshops and spare parts suppliers using the derived instructions for determining workshop capacities and spare parts availability in real time, - Calculating an optimized maintenance appointment by algorithmically comparing the derived maintenance requirements with the determined workshop capacities and spare parts availability.

Description

[0001] The present invention relates to a computer-implemented method and a corresponding device for the automated coordination of a vehicle maintenance process. The invention further relates to a computer program and a machine-readable storage medium. The invention is particularly applicable to networked vehicles that continuously generate diagnostic data and use this data for predictive maintenance planning, scheduling appointments with a workshop, and coordinating spare parts logistics with supply chain stakeholders. State of the art

[0002] Modern vehicles collect a wealth of technical diagnostic data during operation, reflecting the condition of components, fault situations, and potential upcoming maintenance needs. In current systems, diagnostic data is typically output as fault codes or relatively simple error messages. However, further processing in practice often occurs in separate process steps and with manual interfaces. Usually, after an error message appears, a driver independently schedules a workshop appointment. The workshop then uses the available information and its internal processes to determine the likely scope of the repair and checks the availability of necessary spare parts. Procurement of the spare parts and logistical coordination with suppliers often only occur subsequently, and not infrequently only when the vehicle is already at the workshop or the appointment is imminent.This leads to delays, unnecessary downtime, and repeated queries because diagnostic data is not transmitted with sufficient precision or in a machine-readable, processable format. With the increasing complexity of networked vehicle architectures and rising demands on service quality and availability, there is therefore a need for end-to-end automation of the maintenance process, from fault detection and the derivation of specific measures to the coordination of appointments and spare parts delivery. Disclosure of the invention

[0003] One problem is that, in current technology, diagnostic data is essentially processed as codes or simple messages, without deriving a semantically interpreted, machine-readable recommendation for action that would enable automated coordination. Another problem is the lack of seamless automated communication between the vehicle, the workshop, and supply chain stakeholders, meaning that spare parts availability and delivery times are not incorporated into scheduling early on. Furthermore, dynamic planning is lacking, one that considers workshop capacity, parts logistics, and repair duration, while also incorporating user preferences and vehicle-specific contexts such as vehicle history and driving profile.

[0004] The problem described is solved according to claim 1. For this purpose, an LLM-supported agent is used that receives diagnostic data from the vehicle, preprocesses and semantically interprets it, determines a specific maintenance requirement including necessary maintenance steps, prioritization, and required spare parts, and then generates a structured maintenance request. Based on this structured maintenance request, the agent communicates via programming interfaces with workshop management systems to determine available capacities and with supply chain information systems to determine spare part availability, shipping times, and delivery options. From the responses, the agent determines possible maintenance windows, algorithmically evaluates these taking into account the estimated repair duration, workshop utilization, spare part availability, and delivery times, and selects a combination of maintenance date and delivery option.The user receives transparent appointment suggestions and status updates via a vehicle interface, a smartphone application, or a customer portal and can confirm or modify the suggestion. After confirmation, the agent automatically triggers the appointment booking and the provision of the spare parts.

[0005] This approach offers the advantage of fully automating the maintenance process for the first time, without manual intervention from the driver, across all stakeholders. Semantic interpretation of diagnostic data goes beyond simply forwarding error codes; it considers causal relationships and system connections, enabling the structured and machine-readable derivation of maintenance steps, spare parts options, and realistic repair durations. Real-time integration of the supply chain situation dynamically adjusts scheduling to availability and delivery times, reducing downtime and increasing planning reliability. Furthermore, incorporating vehicle history, driving profile, and workshop preferences allows for personalized selection of measures and appointments, improving user acceptance and service quality.

[0006] Further advantageous embodiments of the invention are specified in the dependent claims. In particular, it can be provided that the large language model is adapted or fine-tuned to suit the domain, that the structure of the maintenance request is designed as a standardized maintenance request object, and that the optimization logic weighs various objective functions against each other, such as earliest date, minimum downtime risk, and preference conformity. Embodiments of the invention

[0007] One possible embodiment of the invention will now be explained in detail. A networked vehicle continuously or event-driven provides diagnostic data that can originate from vehicle-side diagnostic sources and communication buses, including on-board diagnostics, CAN-based signals, UDS diagnostic information, or telemetry data. This data is typically semi-structured and can have different formats, timestamps, fault detections, and context parameters depending on the source. In an ingestion and preprocessing stage, the incoming diagnostic data is normalized, temporally and logically merged, and converted into a uniform internal representation. Technical fault classifications and coded signals are thereby translated into semantic descriptions, so that subsequent processing steps do not remain at the level of individual fault codes but can establish a system-wide context.

[0008] Preprocessing is followed by semantic analysis using a domain-specific, large-scale language model. This model processes the normalized diagnostic data along with optional contextual information. Contextual information can include vehicle-specific data such as maintenance history, mileage, previous repairs, environmental conditions, and driving profile, as well as user-specific data such as preferred workshops, time preferences, or communication channels. Based on this information, the language model determines the maintenance requirements and classifies the urgency, for example, by distinguishing between critical and non-critical conditions. In addition, necessary maintenance steps are derived and assigned an estimated repair time. Furthermore, required spare parts are determined, and depending on the implementation, alternative spare part options or compatible variants can also be considered.

[0009] The results of the semantic analysis are transformed into a structured maintenance request. This maintenance request contains, in machine-readable form, the identified measures, their prioritization, the estimated duration, the required spare parts, and other parameters necessary for coordination. The maintenance request is transmitted to external systems via a communication and coordination module. For this purpose, an application programming interface (API) is used to interact with a workshop management system, which provides capacity utilization, available time slots, resources, and, if applicable, qualification requirements. Simultaneously, an interface to a spare parts platform or a supply chain information system is used to query the availability of the required spare parts, possible shipping routes, expected delivery times, and logistical options in real time.

[0010] Based on feedback from external systems, a decision and optimization logic generates a set of potential maintenance windows, combining the workshop's availability with the expected delivery times of spare parts. An algorithmic evaluation weighs these maintenance windows and delivery options against each other. This evaluation considers the earliest possible appointment, the risk of extended downtime due to uncertain delivery dates, the availability of all required parts, the estimated repair duration, and alignment with customer preferences. The result is a selection of a maintenance appointment and delivery option, presented as a proposed appointment.

[0011] User interaction takes place via a suitable user interface. The appointment suggestion can be displayed in the vehicle, provided via a mobile application, or appear in a customer portal. Additionally, status messages can be generated that provide transparency regarding the progress of diagnostic data analysis, appointment scheduling, and parts provision. The user can confirm the suggestion or initiate a manual adjustment, for example, by selecting an alternative time slot or preferring a different workshop. After confirmation, the agent triggers the binding appointment booking in the workshop management system via the respective programming interfaces and initiates the spare parts provision via the supply chain information system.In a further configuration, the agent can continuously monitor the parts logistics after booking and, in the event of changes such as delivery delays or workshop capacity changes, perform further optimization and provide the user with updated suggestions or information.

[0012] The described implementation differs from the state of the art in that diagnostic data is not only processed as error codes, but is semantically interpreted and causally linked before being transformed into a structured maintenance request. Furthermore, scheduling and parts logistics are not separate and performed subsequently, but are integrated in real time and planned in a mutually dependent manner. In addition, the driver's role as a manual intermediary between the vehicle, workshop, and suppliers is reduced or eliminated, as communication and coordination are carried out autonomously by the LLM-supported agent via defined interfaces.

[0013] The invention is not limited to the embodiment described above. Rather, modifications are possible in which the data sources, the structure of the maintenance request, the connected systems, and the evaluation parameters of the optimization logic are adapted to OEM-specific architectures, different workshop networks, or various supply chain platforms. Likewise, the semantic analysis can be combined with other models or rule-based components, provided that the end-to-end coordination continues to be performed by the agent and the essential steps of the semantic derivation of the maintenance requirement, the structured generation of the request, the interface communication, and the optimized scheduling and logistics decision are provided.

Claims

[1] Method for the automated coordination of a maintenance process for a motor vehicle, characterized by the following characteristics: - Electronic reception of technical diagnostic data representing the condition of vehicle components via a data interface, - Semantic analysis of the received diagnostic data using a software agent based on a large language model, whereby the language model automatically derives a specific maintenance requirement and a list of spare parts required for maintenance as a machine-readable instruction from the technical diagnostic data, - Automated sending of electronic queries to external computer systems of workshops and spare parts suppliers using the derived instructions for determining workshop capacities and spare parts availability in real time, - Calculating an optimized maintenance appointment by algorithmically comparing the derived maintenance requirements with the determined workshop capacities and spare parts availability. [2] Method according to claim 1, characterized by the following characteristic: - The calculation of the optimized maintenance appointment is carried out using a decision logic that combines an estimated repair duration, a shipping time of the identified spare parts and a current workshop utilization as parameters in order to achieve a minimum downtime of the vehicle. [3] Device, characterized by the following characteristic: - the device is configured to carry out a method according to one of claims 1 or 2. [4] Computer program which is configured to perform all steps of a method according to claim 1 or 2. [5] Machine-readable storage medium with a computer program stored thereon according to claim 4.