Computer-implemented method for performing an update campaign

Abstract

The invention relates to a computer-implemented method for performing an update campaign for a vehicle (1), wherein an update server (2) wirelessly distributes an update package (3) to the vehicle (1), the update package (3) comprises an associated software update for each of at least two in-vehicle computing units (4), an installation sequence for the computing units (4) is defined, the computing units (4) are updated according to the installation sequence, and the vehicle (1) subsequently transmits an installation report (5) to an analysis point (6). The method according to the invention is characterized in that - the installation report (5) comprises at least the installation sequence in the particular vehicle (1), the update duration (tinstallation) of each computing unit (4), and a data transmission rate (7) of the communication channel (K1, K2, K3) by which each particular computing unit (4) is connected to a communication network (8) of the vehicle (1); - at least two different vehicles (1) of a vehicle fleet perform the update campaign; - the analysis point (6) examines the installation reports (5) from all vehicles (1); and - the analysis point (6) uses the installation reports (5) as a basis for determining, according to at least respective update packages (3), the installation sequence that results in the shortest total update duration (tmin) for the installation of the complete update package (3).

Description

[0001] Computer-implemented method for conducting an update campaign

[0002] The invention relates to a computer-implemented method for carrying out an update campaign for a vehicle according to the type defined in the preamble of claim 1.

[0003] With increasing digitalization, the proportion of computer systems in vehicles such as cars is also growing. For example, processing units integrated in vehicles take on the role of a central on-board computer. This may require updating the software of such a processing unit, for example to correct errors, implement new functions and / or close security gaps. Modern vehicles have a telecommunications unit via which the corresponding processing units can be connected to the Internet. For this purpose, the telecommunications unit can use a cellular connection or a Wi-Fi connection. This enables software updates to be introduced wirelessly into the vehicle, allowing the processing units to be updated quickly and flexibly. This process is also known as an "over-the-air update" (OTA).

[0004] Typically, a vehicle manufacturer combines several software changes into a single update campaign. In such an update campaign, update packages are sent out that contain software updates for one or more in-vehicle computing units. This can also update one or more different software components of a single computing unit. The term "computing unit" refers to the underlying hardware, i.e., a computer or computer system such as a system-on-a-chip (SoC), a control unit, or the like. The term "software component" refers to the software executed by the computing unit, or parts thereof, such as individual program code sections or subprograms intended for a specific task, or the like.For example, updates for several programs running on a single processing unit can be included in the update package.

[0005] For many vehicle functions, such as autonomous driving, multiple computing units and / or software components must work together and are therefore coordinated with each other. In the case of a software update, all software components of a function usually have to be updated together to take software dependencies into account.

[0006] Software updates can also be implemented as so-called differential updates, in which only the changes to the program code are replaced or supplemented. An update package can also include, for example, a new firmware version, an update for an application program, or even changed parameter values, for example, of a characteristic curve for a control program. Installing firmware on a computing unit is also referred to as "flashing."

[0007] The overall update duration in the vehicle should be as short as possible. This means that individual vehicle functions or even the usability of the vehicle itself may be restricted or even unavailable during the installation of software updates. In many cases, the computing units cannot be updated during ferry operation, meaning that in a vehicle with a combustion engine, the power for the update must be supplied from the starter battery. In a battery-electric vehicle, this is also the case during ferry operation, with the traction battery serving as the energy storage device. Since battery capacity in vehicles is limited depending on the drive type, updates must be performed as quickly as possible.

[0008] To provide the customer with better planning for a software update during which the vehicle may only be used to a limited extent, a forecast of the update duration is announced, in particular including a statement regarding the update sequence. For example, the update sequence is determined depending on the available data transmission rate of a particular fieldbus via which the computing units are connected to the vehicle's communications network, as well as randomly measured values ​​of the installation duration in a test system. However, during actual installation in the vehicle, unexpected influences may arise, such as different environmental conditions, unexpected load on individual computer systems, and the like. This may result in the installation time in the field being longer than in the test environment.In this case, the total update time can even be longer with a parallel update sequence of individual processing units than with a sequential update sequence. The key factor here is that such boundary conditions are unpredictable, since in a complex system such as a vehicle, these boundary conditions are also interdependent. A further small database of measured values ​​exacerbates the problem of predictability and the influence of boundary conditions.

[0009] Measures must be taken to account for this. For example, generous time buffers could be provided so that the installation takes place during a vehicle break. However, the vehicle would then be unavailable for an unexpected period of time during this break. The update sequence can also be processed only partially or in multiple steps. However, this requires that all updates are only fully installed after a considerable time, which is generally not feasible due to functional security reasons.

[0010] A method for scheduling an update campaign for so-called wireless updates depending on the update size and the data transfer rate of a bus channel is known, for example, from US Pat. No. 10,042,629 B2. A priority level is assigned to the individual processing units to be updated.

[0011] Furthermore, EP 3662 364 B1 discloses a system for transmitting at least one update package for at least one control unit of a motor vehicle. The document describes an in-vehicle download manager that determines the update sequence of the computing units in the vehicle. Dynamic update campaigns can be executed taking vehicle information into account. This allows the content of an update package to be tailored to a specific vehicle. After an installation attempt, status information about the success of an update installation can be transmitted to a server.

[0012] The present invention is based on the object of providing an improved computer-implemented method for carrying out an update campaign for a vehicle, the application of which shortens the total downtime of the vehicles of a vehicle fleet due to the installation of software updates, compared to known solutions.

[0013] According to the invention, this object is achieved by a computer-implemented method for carrying out an update campaign for a vehicle having the features of claim 1. Advantageous embodiments and further developments emerge from the dependent claims.

[0014] A generic computer-implemented method for carrying out an update campaign for a vehicle, wherein an update server wirelessly distributes an update package to the vehicle, the update package comprises a software update for at least two vehicle-internal computing units, an installation sequence for the computing units is determined, the computing units are updated according to the installation sequence and the vehicle subsequently transmits an installation report to an analysis center, is further developed according to the invention in that

[0015] - the installation report includes at least the installation sequence in the respective vehicle, the update duration of each computing unit and a data transmission rate of the communication channel via which the respective computing unit is connected to a communication network of the vehicle;

[0016] - at least two different vehicles in a fleet carry out the update campaign;

[0017] - the analysis centre examines the installation reports of all vehicles; and

[0018] - the analysis center, specifically for each update campaign, determines from the installation reports the installation sequence that results in the shortest total update time for installing the complete update package.

[0019] The computer-implemented method according to the invention therefore proposes evaluating a summary of the installation history of the update package in a respective vehicle by a central location and comparing the installation sequence of the individual software updates of the respective computing units with the total update duration, so that the installation sequence that results in the shortest total update duration in the individual case can be found. The computer-implemented method according to the invention therefore proposes not only performing an algorithmic calculation of the minimum update duration of a vehicle before the campaign, but also establishing a continuous improvement in the calculation accuracy through analysis of the vehicle data. The vehicle data is categorized and entered into a database in relation to one another.

[0020] To do this, the update campaign is first carried out according to the usual pattern, orchestrated by the update server, for a subset of the vehicles in the fleet. The installation sequence for a particular vehicle can be determined externally, in particular by the update server, or internally, for example, by a download or installation manager running on an on-board computing unit. Thus, the installation sequence will typically differ across the individual vehicles in the fleet. This makes it possible to identify an optimized installation sequence by analyzing the installation reports. This sequence is characterized by the shortest overall update duration compared to the other installation sequences.

[0021] Depending on the constraints prevailing during the installation of the update packages in the respective vehicles, a first installation sequence in a first vehicle may result in a shorter overall update time than in a second vehicle, while a second installation sequence in the second vehicle may result in a shorter overall update time than in the first vehicle. The analysis center evaluates the installation reports of the vehicles in the fleet and is thus able to identify patterns. Proven computer-based or mathematical data analysis methods can be used for this purpose. For example, the "optimal" installation sequence can be determined based on statistical variables.For example, it can be checked which installation sequence results in the shortest overall update time for the majority of vehicles in the fleet and this installation sequence can be defined as the optimal installation sequence.

[0022] The installation sequence can generally stipulate that all of the vehicle's computing units are updated sequentially or in parallel, or even that several computing units are updated sequentially and other computing units are updated in parallel. The analysis center determines the optimal installation sequence for different update campaigns and thus for different update packages individually. For example, an update package specifies the installation of a specific combination of software components on a specific selection of computing units in the vehicle. To obtain meaningful results, the corresponding update campaigns must be examined individually. The number of configuration combinations for a vehicle results in an exponential growth in vehicle variants, which means that each vehicle update for each vehicle must be considered individually.

[0023] The vehicle manufacturer can use the knowledge gained by carrying out the computer-implemented method according to the invention, for example, to adapt the architecture of the communication network of its vehicles so that update campaigns in future series can be carried out more quickly and thus more efficiently and effectively due to a shortened overall update time.

[0024] To communicate with the update server or the analysis center, the vehicle includes appropriately configured communication devices, such as a telecommunications unit. The telecommunications unit also forms an in-vehicle computing unit and is connected to the vehicle's communications network.

[0025] When examining the installation reports, the analysis center sorts out anomalies in the installation report and enters the data from the installation report into a database.

[0026] An advantageous development of the method according to the invention provides that the installation sequence discovered by the analysis center is sent to the update server for adapting the update campaign, so that the update server defines the installation sequence discovered by the analysis center as the optimized installation sequence for the update package. The findings already obtained by the analysis center during the update campaign can be used to adapt the update campaign while it is still being carried out. Thus, an initial set of vehicles in the fleet completes the installation of the update package, which is used for the corresponding analysis by the analysis center.This allows the analysis center to determine the shortest overall update duration based on previous data and adjust the installation sequence for a second set of vehicles in the fleet.

[0027] Various methods can be used to determine the first and second sets of vehicles in the vehicle fleet. For example, the first and second sets of vehicles in the vehicle fleet can be specified by a central location, such as the update server or the analysis center. A wide variety of criteria can be considered to assign vehicles to a particular set, such as the software and / or hardware architecture of the vehicles. Additional sets of vehicles in the vehicle fleet, such as a third, fourth, fifth, or even more, can also be specified.For example, after updating the second set of vehicles according to the optimized installation sequence, the third set of vehicles in the fleet can again perform the update campaign according to the usual pattern, so that even more insights can be gained regarding an even more optimized installation sequence, which insights can then be used to update the fourth set of vehicles, and so on.

[0028] The first set of vehicles can also be determined automatically and freely. For example, the vehicle manufacturer can specify that a certain number of vehicles in the fleet, say 1,000 vehicles, should run the update campaign according to the usual pattern. Once the specified number is reached, a cut is made, and the second set of vehicles, for example, all remaining vehicles in the fleet, are then updated according to the optimized installation sequence determined by the analysis center.

[0029] According to a further advantageous embodiment of the method according to the invention, the update server is used as the analysis point. This simplifies the system structure of the components involved in implementing the method according to the invention.

[0030] A further advantageous embodiment of the method according to the invention further provides that the data transmission rate of each communication channel via which a computing unit to be updated is connected to the communication network is included in the installation report as a continuous or time-discrete signal, at least for the entire update duration. This makes it possible to track fluctuations in the respective data transmission rate during the actual installation process. This enables an even more comprehensive and thus more detailed analysis of the installation process during the entire update duration. The installation report contains at least the data transmission rate of the communication channels via which the computing units to be updated are connected to the communication network.However, the data transmission rates of additional communication channels, preferably all communication channels of the communication network, can also be included in the installation report. This allows further factors influencing the overall update duration to be identified, such as the sending and / or receiving of data packets by computing units that are not part of the update campaign.

[0031] According to a further advantageous embodiment of the method according to the invention, the update duration of a respective computing unit is divided into individual update subsections, wherein in particular at least one of the following update subsections is included:

[0032] - the time required to receive the software update via the

[0033] communication channel;

[0034] - the time required to switch to a bootloader;

[0035] - the time required to delete a memory;

[0036] - the time required to rewrite the memory; and / or

[0037] - a restart period.

[0038] By breaking down the update duration of the respective computing units into update subsections, even more detailed insights into factors influencing the overall update duration can be gained. This allows the analysis center to identify which update subsections are particularly time-consuming and which update subsections can be performed more quickly.

[0039] The update subsections listed above are considered particularly relevant in this regard. A bootloader is a program that loads an operating system. In particular, during an update campaign, the memory is written with a new operating system, for example, in the form of firmware. The processing unit is thus "flashed."

[0040] A further advantageous embodiment of the method according to the invention further provides that the analysis center uses artificial intelligence, in particular in the form of an artificial neural network, to examine the installation reports. Artificial intelligence is particularly powerful in identifying characteristic features in large amounts of data. Thus, the use of artificial intelligence is particularly suitable for examining installation reports. This is particularly interesting for detecting and correcting anomalies and for combining individual data that may reveal dependencies.

[0041] According to a further advantageous embodiment of the method according to the invention, the analysis center groups the vehicles in the vehicle fleet based on the hardware and / or software architecture. The vehicles in the vehicle fleet differ in terms of the computer systems installed and the general vehicle configuration itself, for example, different special equipment. Different vehicles can, for example, have different sensor systems, which allows the provision of completely different driver assistance systems. Such driver assistance systems may require additional control units for data processing. Accordingly, these additional control units can affect the installation process during the installation of an update package and the overall update duration.

[0042] Equivalent hardware components, such as a processing unit of a certain type, can be installed in different vehicles, but run different programs. These programs can also be "the same" but in a different program version. Depending on the nature of a particular assistance system, the same software can also be implemented on different hardware systems and versions. Grouping vehicles according to their hardware and / or software architecture thus allows for even more reliable pattern recognition and thus the identification of other factors influencing the overall update time.

[0043] Proven grouping algorithms, such as the k-means algorithm, can be used to group vehicles. The k-means algorithm is characterized by its simplicity, robustness, and reliability. Grouping vehicles can also be referred to as clustering. K1-based methods can also be used for this.

[0044] A further advantageous embodiment of the method according to the invention further provides that the analysis center additionally considers the assignment of a vehicle to a specific vehicle group when determining an optimized installation sequence. The analysis center can therefore determine several different optimized installation sequences for one and the same update campaign or one and the same update package, depending on the hardware and / or software architecture of the vehicle. For example, a first and a second software component of a first and a second processing unit need to be updated. The first and second processing units can, for example, be installed in a vehicle of a first and a second type.The vehicle of the first type can additionally have a third and fourth processing unit with a third and fourth software component, while the vehicles of the second type additionally comprise, for example, a fifth and sixth processing unit with a fifth and sixth software component. The vehicles of the first and second type therefore have in common the first and second processing unit with the first and second software components that need to be updated. Thus, the vehicles of the first and second types are equally affected by the update campaign. According to the invention, the analysis center is now able to differentiate between the vehicles of the first and second type, i.e. based on the hardware and / or software architecture of the vehicle. The third, fourth, fifth and sixth processing units as well as the corresponding software components serve as distinguishing features for this purpose.The analysis center can thus determine a first optimized installation sequence for the vehicles of the first type and a different optimized installation sequence for the vehicles of the second type.

[0045] According to a further advantageous embodiment of the method according to the invention, at least one vehicle sensor value recorded during the total update duration is included as an additional parameter in the installation report, and the analysis center additionally considers characteristics of vehicle sensor values ​​when determining an optimized installation sequence. This enables even more differentiated identification of influencing factors on the total update duration. The vehicle can record a wide variety of measured values ​​using a wide variety of vehicle sensors. These include, for example, the ambient temperature of the vehicle, an oil temperature, a battery charge level of an electrical energy storage device in the vehicle, status information of a vehicle subsystem, a wheel speed, and the like. Here, too, artificial intelligence is particularly suitable for identifying characteristic patterns.

[0046] Analogous to considering the vehicle architecture, this procedure also allows the determination of different optimized installation sequences for different vehicles in an update campaign, however, not depending on the vehicle design, but taking into account time-dependent boundary conditions. Both additional influencing factors are preferably taken into account. Thus, a further advantageous embodiment of the method according to the invention further provides that a vehicle intended to carry out the update campaign transmits a current vehicle sensor value corresponding to at least one vehicle sensor value to the update server, and the update server distributes an update package to the vehicle with an installation sequence specifically tailored to the characteristics of the received vehicle sensor value.

[0047] The vehicles in the fleet participating in the update campaign can transmit currently measured vehicle sensor values ​​or predicted vehicle sensor values ​​for a future time horizon to the update server, which then selects an installation sequence optimally tailored to the respective constraints and distributes it to the respective vehicle. This ensures that the optimal installation sequence for the respective application is implemented in the vehicle, ultimately allowing the update campaign to be carried out particularly quickly, thus shortening the overall update time in the vehicle.

[0048] Further advantageous embodiments of the computer-implemented method according to the invention for carrying out an update campaign for vehicles also emerge from the exemplary embodiments which are described in more detail below with reference to the figures.

[0049] Showing:

[0050] Fig. 1 is a schematic representation of a communication network of a vehicle;

[0051] Fig. 2 two diagrams showing the update history of the on-vehicle computing units for two vehicles with different installation sequences;

[0052] Fig. 3 is a schematic flow diagram of a computer-implemented method according to the invention for carrying out an update campaign for the vehicles of a vehicle fleet; and

[0053] Fig. 4 shows a schematic system structure of the actors involved in carrying out the method according to the invention.

[0054] Figure 1 shows a vehicle 1 comprising several computing units 4 that communicate with one another via a communications network 8. For the sake of clarity, not all similar elements are provided with reference numerals. The computing units 4, designated Gateway 1 and Gateway 2, are corresponding gateways, for example in the form of a hub or switch. The computing units 4, designated ECU 1 to ECU 6, are control units (Electronic Control Units). In the exemplary embodiment shown in Figure 1, the communications network 8 comprises three communications lines 8.1, 8.2, and 8.3, wherein the respective communications lines 8.1 - 8.3 can be divided into several communications channels as required. The communications network 8 can also be referred to or understood as a bus system. For example, the communications line 8 is1 is an Ethernet data line with an exemplary data transfer rate of 100 Mbit / s. The communication line 8.2 can be, for example, a FlexRay data line with a data transfer rate of 10 Mbit / s, for example. The communication line 8.3 can be, for example, a CAN bus with a data transfer rate of 500 kbit / s, for example.

[0055] The vehicle 1 further comprises a telecommunications unit 9 for establishing a wireless communication connection to the update server 2. The telecommunications unit 9 can, as indicated in Figure 1 by a dashed box, be designed externally to the gateway 1, or can also form the gateway 1 itself.

[0056] The vehicle 1 receives an update package 3, shown in Figure 3, from an update server 2, comprising a software update for several of the computing units 4. This allows individual code components to be newly introduced into the vehicle 1, existing code components to be deleted, and / or existing code components to be replaced with new ones. This does not necessarily require software updates for all computing units 4 of the vehicle 1. However, the method according to the invention is aimed at update campaigns in which at least two different computing units 4, i.e., two computing units 4 that are connected to the communication network 8 via separate communication channels, are updated.

[0057] In this case, it is necessary to determine an installation sequence for the computing units 4, which specifies the order, in particular parallel and / or sequential, in which the respective computing units 4 are to receive the respective software update via the individual communication channels.

[0058] As Figure 2 shows, the installation sequence can affect the total update duration t, i.e. the time required to install all software updates contained in the update package 3 on the computing units 4.

[0059] Figure 2a) shows a diagram of the update process in a first vehicle with a first installation sequence, and Figure 2b) shows the update process for a second vehicle with a second installation sequence. Time is plotted on the abscissa, and the bandwidth or data transmission rate 7 of a respective communication channel K1-KN, divided between the communication lines 8.1 and 8.3, is plotted on the ordinate. Due to the high data transmission rates 7 of the first communication line 8.1, the respective software updates for the computing units 4 "ECLI5" and "ECLI6" as well as "ECLI1" - "ECLI3" can be transmitted simultaneously via the first to third channels K1-K3 of the first communication line 8.1.The software updates can be transferred directly to the processing units 4 "ECLI5" and "ECLI6," while the software updates for the processing units 4 "ECII1" - "ECLI3" must first be transmitted to the gateway 2, which connects the first communication line 8.1 to the third communication line 8.3. The gateway 2 then forwards the respective software updates to the processing units 4 "ECII1" - "ECLI3."

[0060] The representation in Figure 2 is idealized, so that the respective horizontally extending bars have the same thickness in the vertical direction. Taking into account the actual data transmission rate 7, however, the bars of the first communication line 8.1 would be thicker in the vertical direction and the bars of the third communication line 8.3 would be thinner. The extent of each bar on the abscissa corresponds to an update duration tinstaiiation of the respective software update on the respective computing unit 4 (entered as an example for the computing unit 4 “ECU1”). The update duration tinstaiiation includes in particular the time required to receive the software update via the respective communication channel K1-K3, the time required to switch to a bootloader, the time required to delete a memory, the time required to rewrite the memory and / or a restart duration.In the simplest case, the update duration tinstaiition advantageously describes at least the time required to receive the software update via the communication channel K1-K3.

[0061] At point t, which indicates the total update duration, the complete update package 3 was installed in vehicle 1.

[0062] In Figure 2 b), the software update for the computing unit 4 "ECU3" was not transmitted via the first channel K1, but via the second channel K2 of the third communication line 8.3. This leads to a shortened total update time t m in. To illustrate the time difference, the total update time t from Figure 2a) is shown in Figure 2b) as t max. Figure 2 illustrates that the total update duration t for different vehicles 1, in particular with the same or at least similar hardware and / or software architecture, can differ solely by choosing a different installation sequence.

[0063] This situation is used for a computer-implemented method according to the invention for carrying out an update campaign for the vehicles 1 of a vehicle fleet, the sequence of which is illustrated in Figure 3. At the end of the total update period t, the vehicles 1 compile an installation report 5, which includes at least the installation sequence in the respective vehicle 1, the update period t of each computing unit 4, and a data transmission rate 7 of the respective communication channels K1-KN during the installation of the software updates. The total update period t can be reconstructed from the installation sequence and the respective update period t. The installation report 5 is then transmitted from the vehicle 1 to an analysis center 6 for examination.The analysis center 6 identifies patterns within the installation reports 5 transmitted by a large number of vehicles 1 of the vehicle fleet and is thereby able to determine the installation sequence that results in the shortest total update duration t for the respective update package 3 of the update campaign, in particular taking into account the software and / or hardware vehicle architecture and / or currently existing boundary conditions. min conditional. This information is then transmitted by the analysis center 6 to the update server 2, which then adapts the update package 3. Thus, the optimized installation sequence discovered by the analysis center 6 is used to update the other vehicles 1 in the fleet. This reduces the downtime for the other vehicles 1 in the fleet. This increases the reliability of the vehicles 1 in the fleet, extrapolated to the entire fleet.

[0064] The system structure of the actors involved in the method according to the invention is shown again in greater detail in Figure 4. The analysis center 6 comprises a first, second, and third database 10.1, 10.2, 10.3. The first database 10.1 serves to store the installation reports 5, the second database 10.2 describes the hardware and / or software vehicle architecture of the vehicles 1 in the vehicle fleet, and the third database 10.3 describes an assignment of the updated vehicles to the respective vehicle architectures. The installation reports 5 or the vehicle architectures are sorted depending on a unique vehicle identifier, such as the vehicle identification number.

[0065] The analysis unit 6 can read the first and second databases 10.1, 10.2 and, using proven grouping or clustering algorithms, classify vehicles 1 based on vehicle architecture, compare them with the values ​​contained in the installation reports 5, and thereby draw conclusions about the optimized installation sequence. The information obtained from this is stored in the database 10.3.

[0066] The analysis center 6 analyzes the information contained in the first, second, and third databases 10.1, 10.2, 10.3 and is thus able to determine the optimized installation sequence applicable to the respective vehicle architecture. This information is forwarded to the update server 2, which transmits appropriately adapted update packages 3 to the vehicles 1.

[0067] In the exemplary embodiments shown in the figures, the analysis point 6 and the update server 2 are implemented separately from each other. However, the functions of the update server 2 and the analysis point 6 can also be integrated into a common computing device, in particular in the form of a server or server network. In other words, the update server 2 can also function as the analysis point 6, or the analysis point 6 can function as the update server 2.

[0068] Optionally, in addition to the vehicle architecture, a vehicle sensor value available from a particular vehicle during the installation of the software updates can also be considered as a further factor influencing the total update duration t. However, this is not shown in the figures.

[0069] Using the method according to the invention, the total update duration t for individual vehicles in the fleet can be reduced compared to the standard procedure for installing software updates. This can also lead to a lower number of aborts during installation, which ultimately improves the reliability of installing the update packages 3 or executing the update campaign. Furthermore, unexpected influencing factors can be detected and taken into account.

Claims

Patent claims 1. A computer-implemented method for carrying out an update campaign for a vehicle (1), wherein an update server (2) distributes an update package (3) wirelessly to the vehicle (1), the update package (3) comprises a software update for at least two in-vehicle computing units (4), an installation sequence for the computing units (4) is determined, the computing units (4) are updated according to the installation sequence, and the vehicle (1) subsequently transmits an installation report (5) to an analysis center (6), characterized in that - the installation report (5) contains at least the installation sequence in the respective vehicle (1), the update duration (t instaiiation) of each computing unit (4) and a data transmission rate (7) of the communication channel (K1, K2, K3) via which the respective computing unit (4) is connected to a communication network (8) of the vehicle (1); - at least two different vehicles (1) of a vehicle fleet carry out the update campaign; - the analysis center (6) examines the installation reports (5) of all vehicles (1); and - the analysis center (6), specifically for each update campaign, determines from the installation reports (5) the installation sequence that results in the shortest total update duration (t m in) to install the full update package (3).

2. Method according to claim 1, characterized in that the installation sequence found by the analysis point (6) is sent to the update server (2) for adapting the update campaign, so that the update server (2) uses the installation sequence found by the analysis point (6) as an optimized installation sequence for the Update package (3).

3. Method according to claim 1 or 2, characterized in that the update server (2) is used as an analysis point (6).

4. Method according to one of claims 1 to 3, characterized in that the data transmission rate (7) of each communication channel (K1, K2, K3) via which a computing unit (4) to be updated is connected to the communication network (8) is included in the installation report (5) as a continuous or time-discrete signal at least for the total update duration (t).

5. Method according to one of claims 1 to 4, characterized in that the update period (tinstaiiation) of a respective computing unit (4) is divided into individual update subsections, wherein in particular at least one of the following update subsections is included: - the time required to receive the software update via the communication channel (K1, K2, K3); - the time it takes to switch to a bootloader; - the time required to delete a memory; - the time required to rewrite the memory; and / or - a restart period.

6. Method according to one of claims 1 to 5, characterized in that the analysis center (6) uses artificial intelligence, in particular in the form of an artificial neural network, to examine the installation reports (5).

7. Method according to one of claims 1 to 6, characterized in that the analysis center (6) groups the vehicles (1) of the vehicle fleet based on the hardware and / or software vehicle architecture.

8. The method according to claim 7, characterized in that the analysis point (6) additionally takes into account the assignment of a vehicle (1) to a specific vehicle group when determining an optimized installation sequence.

9. Method according to one of claims 1 to 8, characterized in that at least one vehicle sensor value recorded during the total update period (t) is included as an additional parameter in the installation report (5) and the analysis point (6) additionally takes into account characteristics of vehicle sensor values ​​when determining an optimized installation sequence.

10. The method according to claim 9, characterized in that a vehicle (1) intended to carry out the update campaign transmits a current vehicle sensor value corresponding to the at least one vehicle sensor value to the update server (2) and the update server (2) distributes an update package (3) with an installation sequence specifically tailored to the characteristic of the received vehicle sensor value to the vehicle (1).

Images