Method and data processing network for processing sensor data
Patent Information
- Application Number
- EP2022839829
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2022-01-20
- Filing Date
- 2022-12-21
- Publication Date
- 2024-11-27
AI Technical Summary
Current systems for processing sensor data in autonomous driving applications face challenges in achieving reproducibility and high performance due to limited computing capacity and fluctuating execution times, leading to difficulties in implementing safety measures and maintaining low latency.
A method that prioritizes data processing tasks by dividing sensors into resource groups, determining and establishing priorities based on specific parameters, and activating data processing modules accordingly, allowing for efficient resource utilization and improved reproducibility.
This approach enhances resource utilization, reduces latency, and ensures high-quality output data by combining data-driven and time-driven execution methods, enabling exact reproducibility and efficient processing of sensor data in autonomous driving systems.
Smart Images

Figure 1.1
Abstract
Description
[0001] Description
[0002] title
[0003] Method and data processing network for processing sensor data
[0004] State of the art
[0005] Systems for driver assistance or automated driving consist of many individual software units, which can usually be described with graphs in terms of data flow. These software units (often called runnables, nodes, or data processing components) are characterized by the fact that a set of input data is processed and a set of output data is generated from it.
[0006] In the aforementioned systems, input data from sensors such as radar or video is processed in a graph of data processing components that visualizes the data flow in a static view.
[0007] The various software units typically form a complex data processing network with which sensor data is processed to perform actions based on the sensor data. Such actions can, for example, be control tasks within the context of autonomous vehicle operation. Data processing in the data processing network typically comprises a plurality of interdependent data processing steps or data processing tasks that are executed with the data processing components.
[0008] In such a data processing network, the execution of a data processing task depends on a corresponding condition, which can include stimuli such as time steps or the arrival of data. The control flow that determines the execution of the data processing components is usually derived from the data flow. There are data-driven approaches in which the execution of the data processing components is based on the data flow.
[0009] In addition, there are approaches that apply time-driven execution of data processing components. In recent years, this approach has been enhanced with concepts of worst-case execution time (the longest possible execution time).
[0010] In a strictly data-driven approach, the execution of a data processing component is triggered by the arrival of a data packet. Sending data packets during the execution of a data processing component can, with the appropriate graph, lead to the immediate execution of a dependent data processing component. It is even conceivable that a data processing component could be executed multiple times in parallel if new data packets arrive while a data processing component is still executing. Such a system has low latency but a high number of possible states.
[0011] Particularly when processing sensor data for autonomous ferry applications, there are often limitations regarding the available computing capacity for uniform processing of all available sensor data.
[0012] Disclosure of the invention
[0013] Against this background, a data processing method is described here which aims to achieve the reproducibility of data processing while maintaining high performance.
[0014] This describes a method for processing sensor data in a vehicle, comprising a data processing network comprising a plurality of data processing modules, each of which comprises at least one data processing component, wherein each data processing component is configured for a defined data processing task for processing the sensor data, wherein each data processing module receives the sensor data and / or output data of further data processing modules as input data and generates output data, which in turn are network output data of the data processing network and / or input data of further data processing modules, wherein the at least two sensors are divided into at least two different resource groups, wherein the method comprises the following steps:
[0015] A) Receiving at least one parameter by means of which at least one priority of sensors of a resource group can be determined;
[0016] B) setting at least one priority for at least one resource group based on the at least one parameter;
[0017] C) Activating data processing modules according to the defined priorities.
[0018] It is particularly preferred if the following steps are carried out for at least one data processing module after step C): a) receiving a set of input data for carrying out the data processing tasks in the at least one data processing component of the respective data processing module; b) receiving a stimulus for activating the at least one data processing component of the data processing module; c) if the set of input data was received in step a) and the stimulus was received in step b): activating the at least one data processing component of the data processing module and carrying out the data processing task for which the data processing component is configured, using the respective input data to generate output data; and d) providing the output data for further data processing and / or as network output data.
[0019] Implementing the method according to the resource groups and steps A) to C) enables an improvement in the use of resources in data processing in a data processing network in a vehicle.
[0020] The at least one parameter received in step A) is any parameter that can be used to establish a technically reasonable priority for a specific resource group. The types of parameters that can generally be used here are explained below. A parameter that can be used to establish a priority for a resource group refers, in particular, to a parameter that can be used to establish a priority with which the sensor data from sensors within the respective resource group should be processed.
[0021] In step A), a parameter can in principle originate from any data source. For example, such a parameter can originate from a specification external to the process. However, it is also possible for such a parameter to arise from the process itself, for example if such a parameter was determined during previous executions of the process within the data processing network. Another possibility is that the sensor data of the at least two sensors from the resource groups are fed into further data processing in parallel to the process described here, and such parameters are determined in this data processing, possibly in a brief preliminary evaluation of the sensor data to estimate the priority with which the sensor data should be processed.
[0022] In step B), at least one priority is then determined for at least one resource group based on the at least one parameter. Preferably, a priority is determined for each resource group. This step also includes, in particular, a weighting of the respective parameters. For example, it may be the case that the priorities of the two resource groups are both very high based on the parameters received in step A). In this case, it may not be of any use to set both resource groups to a very high priority. In such a situation, it is often more advantageous to assign both priorities to an equally high average priority value. Such an approach can be technically implemented, for example, in such a way that all of the priorities determined in step B) together are smaller than a total priority sum.Another suitable procedure that can be implemented in step B) is, for example, to downgrade the priority of other resource groups instead of, or possibly in addition to, increasing the priority of resource groups. Activating the data processing modules in step C) based on the defined priorities preferably includes all data processing modules that are necessary for the desired evaluation of the respective sensor data from the sensors in the resource group. In particular, this includes not only data processing modules that directly process sensor data as input data, but also data processing modules that process output data from other data processing modules as input data, which is the result of processing the sensor data from the sensors of the respective high-priority resource groups.
[0023] In this context, it is also possible for sensor data from sensors in other resource groups to be processed with lower priority. For example, it is possible for data processing modules that primarily process sensor data from a specific resource group to occasionally be additionally supplied with sensor data from sensors in another resource group because the graph of the data processing network dictates this, for example, because this is helpful for a comprehensive evaluation of the sensor data collected by the sensors.
[0024] The data processing in the individual data processing modules according to steps a) to d) forms a preferred, subordinate process to the higher-level process steps A) to C) for determining and utilizing priorities for the individual data processing tasks. Both processes (according to steps A) to C) and steps a) to d)) work together to ensure particularly good utilization of the available computing capacity / hardware while simultaneously ensuring high reliability and reproducibility of the data processing.
[0025] By processing the data in the individual data processing modules according to steps a) to d), the problem that predictability and reproducibility are very difficult to achieve with a conventional approach is solved. This complicates the implementation of safety measures, such as a software lockstep, in which the same software is executed simultaneously on two p-processors. Even recomputing a recorded driving situation as faithfully as possible is difficult, since a different runtime behavior and thus potentially a different result must be expected.
[0026] The basis of the method described here according to steps a) to d) is that a plurality of data processing components of a data processing network are each combined into so-called data processing modules, thus creating an additional higher-level structure of the data processing network. At the level of this structure, input data and output data of the individual data processing components are combined within the module, and the data flow is controlled or monitored by the data processing network at this level. This structure is preferably also used to utilize the priorities defined in step B) for the (preferential) activation of the respective data processing modules according to the defined priorities (step C)).
[0027] In the automotive industry, with its high level of control engineering, execution in time slices has always been dominant (e.g., 10ms, 20ms, 100ms tasks). However, additional problems arise, particularly when multi-core systems are used as the hardware for executing data processing with such data processing networks. Fluctuating runtimes of the data processing components of a data processing network occur particularly on multi-core systems. Such fluctuating runtimes make the assignment of output data from one data processing component as input data for other data processing components difficult or even impossible to predict. This predictability can be improved if necessary—for example, with concepts of the maximum possible execution time / processing time. However, such concepts impair the utilization of the hardware (of the multi-core system). The hardware must be dimensioned considerably larger.
[0028] This is where the described (higher-level) method offers synergistic effects. By focusing on data processing tasks and data processing modules with particularly high priority, the utilization of the available hardware can be further improved and used more intensively to ensure particularly high quality output data from the data processing network.
[0029] Particularly in highly complex driver assistance or automated driving applications, the amount of data to be processed from sensors is drastically increased. However, the required response time of such systems is comparable to or even lower than that of traditional driver assistance applications. This means that more and more computationally intensive calculations must be performed in a longer processing chain within a comparable time period. This leads to unacceptable latency in existing time-driven approaches, as the additional latency due to the transitions between the individual time slices accumulates throughout the entire chain.
[0030] The presented method, according to process steps a) to d), enables the combination of approaches from data-driven execution with achievements from the time-driven execution of data processing components. This allows data processing networks to be operated in such a way that lower latency (better performance) occurs than with purely time-driven systems and better reproducibility than with strictly data-driven systems. This makes it possible to meet the high latency requirements in a system for automated driving while simultaneously having a system that enables execution in software lockstep and exact reproducibility in recompute. For step b), a suitable stimulus can be defined that determines the execution of the data processing in step c), which then uses the input data received in step a).The output data is then made available for subsequent processing steps in step d). If the respective data processing module is the last data processing module in a data processing network, the data can also be referred to as network output data or system output data, which then, for example, simultaneously serves as input data for a controller that processes this data or considers it for an application.
[0031] The described method enables both time- and data-driven execution of data processing tasks. The actual start of data processing occurs when the stimulus is received in step b). The data becomes visible to the data processing components upon arrival of the stimulus. Temporarily related data structures are thus transferred together between data processing modules. Data is provided to the data processing components based on the data processing modules provided in a higher-level structure. The structure of the higher-level data processing modules and the fact that data is provided at this level significantly reduce the number of system states of the entire data processing network.
[0032] It is particularly preferred if the following step is carried out after step d): e) providing a validation data set consisting of the set of input data, the stimulus and / or the output data for validating the execution of the at least one data processing task with the at least one data processing component.
[0033] Step e) is to be understood here as part of the series of process steps a) to d) which are carried out after step C) for each data processing module.
[0034] It is preferred if the validation data set additionally contains at least one time information which enables a time of the stimulus and / or time information about the processing of the at least one data processing task.
[0035] Such time information can be obtained, for example, by recording the start and end of the execution of data processing tasks. This allows the processed input data and the generated output data to be displayed on a common logical timeline, enabling test calculations of the described process.
[0036] It is possible for a data processing module that processes output data from another data processing module as input data to start processing when a stimulus occurs as an activation. Thus, compared to the time-driven approach, the additional latency caused by WCET and until the start of the time slice of the receiving data processing module is avoided. The method described here achieves reproducibility in the execution of the individual data processing tasks because the information regarding the respective processed input data is reproducible.
[0037] In the presented method, a data processing module always has a frozen view of the world or the input data being processed during execution (execution of step c). The input data does not change during execution of the method. This is achieved by collecting the incoming input data in step a), and by passing the data in a controlled manner and atomically in logical time. To ensure consistent processing of the output data of a data processing module with respect to its execution by other data processing modules, the output data is preferably also collected.
[0038] It is particularly preferred if, in order to activate the data processing module in step C), the stimulus (9) received in step b) is determined with a priority defined in step B).
[0039] Steps C) and b) preferably jointly relate to the interaction of the higher-level method according to steps A) to C) and the lower-level method according to steps a) to d), which is carried out for each data processing module. The priorities are preferably used to control the transmission of the stimuli for activating the respective data processing modules. With the help of the priorities, for example, delays can be specifically caused by data processing modules with a lower priority in order to create additional computing capacity for data processing modules with a higher priority. It is also possible, for example, to specifically increase the (time-controlled) frequency with which stimuli are transmitted for certain data processing modules.
[0040] It is preferred if at least one timer is used to generate the stimulus used in step b), which timer specifies a time frame for regularly repeating the execution of the data processing tasks with the data processing components. The timer can, for example, be a corresponding component on a piece of hardware on which the data processing network is operated and which emits a timer signal for each data processing module at regular intervals, which forms the stimulus and triggers execution of the method. Such a time frame can be modified using the priorities in order to realize execution of the data processing tasks according to the respective priorities.
[0041] It is also preferred if at least one availability signal indicating the availability of data is used to generate the stimulus used in step b). This availability signal can, for example, have been generated by a previous execution of the described method in another data processing module. If it can be determined from priorities that data processing with certain data processing modules is less important in a particular situation despite the availability of the data, then such an availability signal can, for example, be delayed with the priorities in order to create computing capacity for higher-priority data processing tasks.
[0042] It is particularly preferred if the stimulus is formed by a combination of a timer and an availability signal. Whenever new data is indicated via an availability signal, the data processing module is prepared to respond to the timer. Only if both the timer and the (at least one) availability signal indicate that data processing is to begin does the data processing take place in the data processing components of the respective data processing module (step c). Preferably, the priorities of the respective resource groups are also incorporated into the formation of a stimulus with a combination of a timer and an availability signal.
[0043] It is particularly preferred if step a) is carried out with an input data reception module of the data processing module, which has an input memory for temporarily storing incomplete input data and which performs a completeness check of the set of input data. It is also preferred if step d) is carried out with an output data provision module, which has an output memory in which incomplete output data is temporarily stored.
[0044] Preferably, the data processing module has special gates (input data reception module = input gate and output data provision module = output gate) to carry out the data processing task.
[0045] Through these gates, a controlling unit can control the data flow between the data processing modules. If the controlling unit synchronizes the start and end of the execution of the data processing modules with the passing of data through the gates, it can control which data processing module is executed when and with which data. To decide whether to start a data processing module, the aforementioned stimulus is evaluated.
[0046] It is particularly preferred if the method is carried out on a data processing system on which several threads can be executed in parallel or pseudo-parallel, wherein each data processing module is assigned to a thread, wherein the activation of the data processing modules in step C) is carried out by activating the respective threads with the respective priority.
[0047] It is also preferred if the data processing system provides an operating system in which the described method is carried out, wherein system priorities for threads can be assigned in the operating system, wherein in order to activate the respective threads with the respective priority in step C), the priority is converted into system priorities.
[0048] The threads are preferably specified or predefined by an operating system on which the described method is implemented or on which the described data processing network is configured to implement the described method. Such an operating system can, for example, provide a fixed number of threads or a variable number of threads (adjustable as needed). In a preferred embodiment, the data processing tasks corresponding to a data processing module are assigned to respective threads for execution when they are to be executed regularly. In the threads, these data processing tasks are then processed according to the priorities.
[0049] Typically, an operating system specifies a priority system that allows the priority of processing data processing tasks within the operating system to be controlled. However, such a priority system is generally not well-suited to implementing the priority management presented here for data processing tasks for processing sensor data based on prioritized resource groups. For example, there are priority systems in operating systems that are primarily designed to assign high priority to individual, particularly important tasks and to initiate their accelerated (or early) processing. The system described here often has a somewhat different purpose.For example, faster processing of sensor data from sensors in a specific resource group with the goal of generating decision-relevant information for controlling an autonomous driving process as early as possible / as quickly as possible is often more easily achieved if a significant number of less relevant data processing tasks are downgraded in terms of priority than if a number of data processing tasks are upgraded. For this reason, such requirements can often only be implemented to a limited extent using the priority system provided by an otherwise very suitable operating system. For this reason, a priority compiler or priority converter is proposed here, which converts the priorities defined in step B) into system priorities, which are then used to execute the data processing tasks or the data processing modules in the operating system threads.
[0050] It is further advantageous if, by defining the priorities in step B), the computing capacity of a data processing system on which the described method is carried out is concentrated on processing data from sensors in one or more resource groups, namely on those sensors or resource groups to which a high priority was assigned in step B).
[0051] By concentrating computing power on sensor data from sensors in specific resource groups, it is possible, in particular, to process sensor data from such sensors with reduced processing time, increased processing frequency (more frequent checking and processing of updated sensor data from these sensors), and / or with increased accuracy (more in-depth analysis of the respective sensor data). Preferably, the processing of sensor data from sensors in other resource groups, for which lower (or lower) priorities were defined in step B), is not completely suppressed.
[0052] Furthermore, the method is advantageously carried out on a data processing system and the priorities are determined in step B) taking into account at least one utilization parameter, wherein the at least one utilization parameter takes into account the utilization of the data processing system with the processing of sensor data from sensors from one or more specific resource groups.
[0053] Such a utilization parameter can be used as a higher-level parameter in determining priorities in step B). In a large data processing system, for example, comprising a plurality of microprocessors, several utilization parameters can also be used in parallel.
[0054] The method is particularly preferred if the priorities in step B) are determined depending on the driving situation.
[0055] The method is further advantageous if the priorities in step B) are set such that data from sensors are preferentially processed where increased attention is advantageous depending on the driving situation. A variety of effective method implementations are possible here. In this context, it is preferably advantageous if the parameters received in step A) allow a conclusion to be drawn about the respective driving situation. Such parameters can, for example, be determined directly from the vehicle's bus system. For example, it is possible to determine a corner to the right or left using a general steering angle. Other information that can be obtained via a bus system includes, for example, speed, reversing, etc.In other embodiments, it is also possible for such parameters to be obtained, for example, from previous executions of the method described here. For example, using the described method, it can be determined from sensor data that a large number of pedestrians are in the vicinity of the vehicle. Such information can then be subsequently used to determine priorities for the described method. Another group of potentially usable information includes, for example, geographical information, such as whether the vehicle is currently on a motorway or a country road, or time information, which could, for example, allow conclusions to be drawn about increased traffic volume or similar information.
[0056] Determining which sensors or resource groups should receive increased attention depending on the driving situation in order to set priorities can be done in step B) using many different approaches. This can be done, for example, using characteristic maps or networks in which knowledge about the need for increased attention is stored. However, these can also be quite simple decisions. If one resource group includes sensors on the left side of a vehicle while another resource group includes sensors on the right side of a vehicle, concentrating the available computing power on sensors on the right side of the vehicle can be advantageous when the vehicle is making a right turn. Any more complex cases can be covered using the described method.
[0057] Preferably, the provision or division of the
[0058] Sensors / data sources in resource groups also allow priorities to control how many data processing modules can be activated simultaneously to process sensor data from sensors in a resource group. If a stimulus is triggered for a module, the processing of this data processing module or the associated data processing task can be postponed until a "free resource" is available for this resource group again.
[0059] For this purpose, a parameter is preferably provided that defines the maximum number of data processing modules permitted for processing. This parameter is preferably dynamically adjustable (depending on the driving situation).
[0060] Also to be described here is a data processing network for carrying out the described method with at least one data processing module having an input data reception module to which an input memory is assigned and an output data provision module to which an output memory is assigned, and further comprising at least one data processing component for carrying out a data processing task based on the input data in the input memory and for generating output data for storage in the output memory.
[0061] The data processing network also comprises, in particular, a priority module for determining priorities for the activation of individual data processing modules according to steps A) to C).
[0062] In addition, a data processing network comprising a plurality of such data processing modules is to be described
[0063] The explanations of the method provided above are transferable and applicable to the data processing module and the data processing network.
[0064] The method is explained in more detail below with reference to the figures. They show: Fig. 1: a vehicle in which a described data processing network is used and in which the described method is applied;
[0065] Fig. 2: a data processing module for carrying out the described method;
[0066] Fig. 3: a further embodiment of a data processing module for carrying out the described method;
[0067] Fig. 4: a type of data processing with a described method, shown on a timeline;
[0068] Fig. 5: another type of data processing with a described method, shown on a timeline, and
[0069] Fig. 6: yet another type of data processing with a described method, shown on a timeline.
[0070] Fig. 1 schematically illustrates a vehicle 1. The vehicle 1 is, in particular, a road vehicle, for example, a passenger car or a truck. The vehicle 1 has sensors 23 for acquiring information, such as environmental data from the surroundings of the vehicle 1, which can be used by various driver assistance systems. Such systems can be, for example, active or passive safety systems or systems for autonomous (or semi-autonomous) ferry operation. Such systems are illustrated in Fig. 1 as controller 20.
[0071] In order for the controller 20 to process the data from the sensors 23, this data must be prepared. For this purpose, the vehicle 1 has a data processing network 4. The sensors 23 form data sources 13 for this data processing network 4, which provide sensor data 3 to the data processing network 4. The controller 20 forms an output data receiver 21 for this data processing network 4, which receives network output data 14 from the data processing network 4. The data processing network 4 consists of a plurality of data processing modules 5, each of which consists of (one or more) data processing components 6.The data processing network 4 with the data processing modules 5 and the data processing components 6 is preferably implemented on hardware 25, which in particular has a data memory 27 on which the data processing network 4 can store data 2 and which can further provide various other hardware functions for the data processing network 4, for example a timer 10.
[0072] The vehicle 1 according to Fig. 1 has sensors 23 or data sources 13 that are assigned to two different resource groups 28. In a simple example, these can be, for example, sensors 23 or data sources 13 relating to the left side of the vehicle 1 in a first resource group 28 and sensors 23 or data sources 13 relating to the right side of the vehicle 1 in a second resource group 28. Using the described method, the needs-dependent prioritization of the sensor data 3 of the sensors 23 or data sources 13 from the respective resource group can be carried out. The data processing network 4 preferably has a prioritization module 29 for this purpose.
[0073] 2 and 3 show various embodiments of data processing modules 5 for the described data processing network 4. The described method is carried out at the level of each individual data processing module 5. Each data processing module 5 has an input data reception module 15 with an input memory 16 for receiving input data 7 and an output data provision module 17 with an output memory 18 for providing output data 8, as well as a data processing component structure 26 with data processing components 6 for data processing from the input data 7 to the output data 8. The input data reception module 15 can have an interface for receiving a stimulus 9 with which the data processing with the data processing components 6 can be initiated.The output data provision module 17 may have an interface for emitting an availability signal 12 when the data processing with the data processing components 6 is completed and the output data 8 is available.
[0074] Method steps a) and b) of the described method relate to the reception of input data 7 and are primarily carried out by the input data reception module 15. Method steps d) and e) of the described method relate to the provision of output data 8 and are primarily carried out by the output data provision module 17. The actual data processing takes place in step c) in the data processing components 6, which form a data processing component structure 26 of the respective data processing module 5.
[0075] The embodiment of the data processing module 5 according to Fig. 3 has a feedback 22 as a special feature. Fig. 1 shows an example of a data processing network 4 which ultimately processes the data from the data source 13 directly deterministically and without feedback from the data source 13 (here the sensors 23) to the output data receiver 21 (the controller 20). This is sufficient for many applications. However, the feedback 22 shown in Fig. 3 also shows how non-directly deterministic data processing can be enabled in data processing networks 4 constructed in this way, namely by means of such feedback 22. Output data 8 can be routed via such feedback 22 back to the input data reception module 15 of the same data processing module 5 in order to then process it in subsequent data processing with the data processing components 6 of the data processing module 5.In this way, the data processing network 4 can be taught a memory for data. In the embodiment according to Fig. 3, another special feature is shown: the output data 8 consists of partial data sets 19 that are available at different times during data processing with the data processing components 6. The output data provision module 17 can be configured to output the availability signal 12 for all output data 8 (partial data sets 19) in one go. Figs. 4, 5, and 6 show various types of data processing that can be performed with the described data processing network 4 or with the described method.
[0076] The illustrations show different data processing modules 5,1,
[0077] 5.2 and / or 5.3, wherein the duration of the execution of the data processing operations with the data processing components 6 belonging to the data processing modules 5 is plotted as a bar on the timeline 24. The data sources 13 (here, sensors 23) with which data 2 (here, sensor data 3) can be introduced into the data processing network 4 are shown schematically at the top.
[0078] The individual data processing modules 5.1, 5.2, and / or 5.3 are shown repeatedly in succession across the timeline 24. This illustrates that the individual data processing modules 5.1, 5.2, and / or 5.3 are each executed repeatedly, each time accessing different input data. The execution of one of the data processing modules 5.1, 5.2, and / or 5.3 begins when a stimulus 9 is present.
[0079] Figures 4, 5 and 6 each show different types of stimuli 9 that can be used here.
[0080] Fig. 4 shows that only availability signals 12 are used as stimuli 9, which indicate the availability of data for carrying out specific data processing tasks with data processing modules 5.1, 5.2 and / or
[0081] 5.3. The process depicted in Fig. 4 is quite uncontrolled in terms of data processing. As soon as new data is available, the execution of the respective data processing modules 5.1, 5.2, and / or 5.3 begins. While this achieves high processing speed, it also greatly reduces traceability of which data was processed by the respective data processing module 5.1, 5.2, and / or 5.3. This is especially true because the duration of data processing with the data processing modules 5.1, 5.2, and / or 5.3 cannot be accurately predicted, and thus there is little or no reproducibility as to which data processing module 5.1, 5.2, and / or 5.3 responds to which input data. "Reproducibility" here refers to the reproducibility of data processing with the data processing network 4.In this context, low reproducibility means that a very high level of effort must be made to reproduce the data processing with the individual data processing modules 5.1, 5.2 and / or 5.3, as is necessary, for example, for debugging tasks or for tasks to ensure the correctness of data processing in redundant systems.
[0082] Fig. 5 shows that signals from a timer 10 are used as stimuli 9, which follow a fixed time grid 11. By using such stimuli 9 to trigger data processing with the data processing modules 5.1, 5.2 and / or 5.3, it can be achieved that it is always known exactly with which input data the data processing with the data processing modules 5.1, 5.2 and / or 5.3 starts. This achieves a high level of reproducibility in the sense mentioned above - at the same time, however, the performance of the data processing network 4 is drastically reduced. The performance of the data processing network 4 here means the ability of the data processing network 4 to be operated with the least possible hardware resources. This is because it must always be ensured that the data processing with a data processing module (e.g. data processing module 5.1) is completed before a further data processing module is started, which processes output data 8 of the first data processing module as input data (e.g.
[0083] Data processing module 5.2). Because the duration of data processing with data processing modules 5.1, 5.2, and / or 5.3 cannot be accurately predicted, the time frame must be adequately designed based on the longest possible execution time.
[0084] The embodiments shown in Fig. 6 combine aspects of the embodiments shown in Figs. 4 and 5 to achieve high reproducibility on the one hand and good performance on the other. Depending on the task, the data processing network 4 operates with availability signals 12 or timers 10 as stimulus 9.
Claims
Claims: A method for processing sensor data (3) from at least two sensors (13, 23) in a vehicle (1), comprising a data processing network (4) comprising a plurality of data processing modules (5), each of which includes at least one data processing component (6), wherein each data processing component (6) is configured for a defined data processing task for processing the sensor data (3), wherein each data processing module (5) receives as input data (7) the sensor data (3) and / or output data (8) of further data processing modules (5) and generates output data (8) which are network output data (14) of the data processing network (4) and / or input data (7) of further data processing modules (5), wherein the at least two sensors (13, 23) are divided into at least two different resource groups (28), wherein the method comprises the following steps: A) Receiving at least one parameter from which at least one priority of a resource group (28) can be determined; B) Establishing at least one priority for at least one resource group (28) based on at least one parameter; C) Activating data processing modules (5) according to the specified priorities. Method according to claim 1, wherein for at least one Data processing module (5) after step C) the following steps are performed: a) Receiving a set of input data (7) to perform the data processing tasks in the at least one data processing component (6) of the respective data processing module (5); b) Receiving a stimulus (9) to activate the at least one data processing component (6) of the data processing module (5); CORRECTED SHEET (RULE 91) ISA / EP c) If in step a) the set of input data (7) and in step b) the stimulus (9) was received: activating the at least one data processing component (6) of the data processing module (5) and performing the data processing task for which the data processing component (6) is configured with the respective input data (7) to generate output data (8); and d) providing the output data (8) for further data processing and / or as network output data (14), method according to claim 2, wherein to activate the The data processing module (5) in step C) determines the stimulus (9) received in step B) with a priority defined in step B). A method according to one of the preceding claims, wherein at least one timer (10) is used to generate the stimulus (9) used in step B), which specifies a time grid (11) for the regular repetition of the execution of the data processing tasks with the data processing components (6). A method according to claim 2, wherein at least one availability signal (12) is used to generate the stimulus (9) used in step B), which indicates the availability of data (2).A method according to any one of the preceding claims, wherein step a) is performed with an input data reception module (15) of the data processing module (5), which has an input memory (16) for temporarily storing incomplete input data (7) and which performs a completeness check of the set of input data (7). A method according to any one of the preceding claims, wherein step d) is performed with an output data provision module (17), which has an output memory (18) in which incomplete output data (8) is temporarily stored. CORRECTED SHEET (RULE 91) ISA / EP Method according to one of the preceding claims, wherein the method is carried out on a data processing system on which several threads can be executed in parallel or pseudo-parallel, wherein each data processing module (5) is assigned to a thread, wherein the activation of the data processing modules (5) in step c) is carried out by activating the respective threads with the respective priority.Method according to one of the preceding claims, wherein the data processing system provides an operating system in which the described method is carried out, wherein system priorities for threads can be assigned in the operating system, wherein, in order to activate the respective threads with the respective priority in step c), a conversion of the priority into system priorities takes place. Method according to one of the preceding claims, wherein, by setting the priorities, the computing capacity of a data processing system on which the described method is executed is concentrated on the processing of data from sensors (13, 23) in one or more resource groups (28).A method according to any of the preceding claims, wherein the method is carried out on a data processing system and the setting of priorities in step B) takes into account at least one utilization parameter, wherein the at least one utilization parameter considers the utilization of the data processing system with the processing of sensor data (3) from sensors (13, 23) from one or more specific resource groups (28). A method according to any of the preceding claims, wherein in step d) an availability signal (12) is additionally generated, from which it is apparent that output data (8) have been made available for further processing. A method according to any of the preceding claims, wherein the priorities in step B) are set depending on the driving situation. CORRECTED SHEET (RULE 91) ISA / EP The method of claim 13, wherein the priorities in step A) are set such that data from sensors where increased attention is advantageous depending on the driving situation are preferentially processed. A data processing network (4) for carrying out the method of any of the preceding claims, comprising at least one data processing module (5) comprising an input data receiving module (15) to which an input memory (16) is assigned and an output data provision module (17) to which an output memory (18) is assigned, and further comprising at least one data processing component (6) for carrying out a data processing task based on the input data (7) in the input memory (16) and for generating output data (8) for storage in the output memory (18). CORRECTED SHEET (RULE 91) ISA / EP
