Automation application service detection for configuring an industrial communication network

Abstract

An automated method for configuring an industrial communication network (IN) comprising the steps of: - deploying an application (A) deployable to the industrial communication network (IN) to a digital twin (DT), wherein the digital twin is set up to: represent the application, represent a digital representation of an industrial communication system comprising the industrial communication network (IN), devices (D) connected to the industrial communication network, and a process, and represent physical interactions in the industrial communication system, - running the application (A) on the digital twin (DT), wherein - data of the communication traffic of the application is collected, - based on the collected communication traffic of the application, endpoints of the application and their corresponding communication characteristics in the industrial communication network are initially detected and classified, - the initially detected communication characteristics of the communication links used by the application are changed until a limit of a predefined application performance quality is reached, and - from this, determined communication characteristic network parameters for the communication services required by the application are derived, - and configuring the industrial communication network based on the determined network parameters. The invention further discloses a corresponding apparatus with a digital twin (DT).

Description

Technical Field

[0001] The present invention relates to an automated method and apparatus for configuring industrial communication networks using digital twins. Background Technology

[0002] (Communication) networks are the medium for transmitting data. However, networks vary based on the amount of data being transmitted. Industrial communication networks refer to networks that handle large-scale data transmissions. This means that they allow various devices to be connected across large spaces and enable communication between them by allowing the transmission of huge blocks of data.

[0003] Industrial applications typically have very specific requirements for industrial communication systems, such as determinism, reliability, and many more. These requirements are highly process-specific and use-case-specific, and often cannot be easily "guessed," but rather require careful consideration with a significant amount of manual work. Especially for application deployment and system integration, they are often impossible to fully capture or easily described with simple values. Inaccurate or incomplete descriptions of requirements, or even incorrect parameter settings, will lead to unreliable operation of industrial applications. Another "extreme" but typical setup is the parameterization of systems with overly high requirements, resulting in the over-allocation of system resources.

[0004] The following small example illustrates a question that arises in response to a typical inquiry: When does the device interface need to send data packets so that arrival at time X, determined by the TSN (Time-Sensitive Networking) controller, is complied with? Typical tasks and problems are: - Application or application endpoint A is a sensor and wants to send packets to application or application endpoint B using a transmission interval of time X. - Application B or application endpoint B is the controlling application and wants to send packets to actuator C every Y nanoseconds—within a given period of the application. Ideally, endpoint B is intended to send a packet after it has been received from endpoint A. - The underlying network must serve both requests, but the request to send a packet after receiving a packet from endpoint A is not taken into account.

[0005] Current solutions require applications or configuration tools to send service requests to a control entity (i.e., the network controller). Creating such service requests is not easily possible in automated and production environments because industrial applications have a very specific set of requirements—completely different from what we know from the IT world.

[0006] To create such communication requests, application developers typically provide mechanisms for capturing these requests (e.g., through user input), allowing the application to create and send service requests. Automation of this process is currently unavailable. The current process is even more complex and error-prone.

[0007] When formulating requirements related to communication services, application developers must consider timing constraints caused by factors such as network interfaces, network stacks, and operating systems, which are often difficult or impossible for them. In particular, consider the development of portable apps, where the destination hardware is variable, and ultimately, the operating system is variable. Summary of the Invention

[0008] The purpose of this invention is to provide an automated detection solution for application services in industrial communication networks.

[0009] To achieve this objective, the present invention provides a solution according to the independent claims. Advantageous embodiments are provided in the dependent claims.

[0010] According to one aspect of the invention, an automated support method for detecting application services and their parameters in such industrial communication systems is introduced. The proposed method also provides network controller information related to the type of application sending data packets. Typically, the network controller (e.g., a TSN controller) does not automatically know how to handle and parameterize the data traffic of an application. The proposed method will detect the timing of data packets from the communication endpoints of the industrial application (e.g., sensors, actuators, applications). This parameter detection can be used in all types of setups, such as systems following a sense-analysis-action paradigm, and in systems with asynchronous communication.

[0011] The mechanism described in this invention allows for the automated evaluation of application requirements, which must be performed by the industrial communication system. Therefore, application developers and system integrators can easily deploy applications in industrial communication environments with reduced risk of system misconfiguration.

[0012] This invention claims an automated method for configuring an industrial communication network, comprising the following steps: - Deploy applications that are deployable to the industrial communication network onto a digital twin, wherein the digital twin is configured to: represent the application; represent a digital representation of an industrial communication system including the industrial communication network, devices connected to the industrial communication network, and processes; and represent physical interactions within the industrial communication system. - The application runs on the digital twin, where - Collect data on the communication traffic volume of the application. - Based on the collected communication traffic volume of the applications, the endpoints of the applications and their corresponding communication characteristics in the industrial communication network are initially detected and classified. - Change the initially detected communication characteristics of the communication link used by the application until a predefined limit for application performance quality is reached, and - Derive the network parameters for the determined communication characteristics required by the application from these parameters. - and configure the industrial communication network based on the determined network parameters.

[0013] This means that communication characteristics are altered until application performance degrades or fails to function properly at a predefined threshold required by the demanding communication services, thereby finding the limits of the communication characteristics that the application can tolerate without degradation of process quality.

[0014] In other words, the application's communication is "pushed to the limit," and the digital twin will notify the user once the application fails (e.g., the control loop is no longer able to control the physical processes in the digital twin, and the results of the processes are outside the required operational envelope).

[0015] The advantage of this invention is that it assesses the behavior, network, and communication effects of applications in a digital twin environment, taking into account the host system on which the application endpoints run, as well as other applications running in the system that result in data traffic.

[0016] In a further embodiment of the method, the communication characteristics include latency, jitter, and packet loss.

[0017] In a further embodiment of the method, the application performance quality includes predefined quality parameters and / or predefined functional envelopes.

[0018] In a further embodiment of the method, altering the communication characteristics includes: adding additional delays on the link; creating packet loss; and / or realigning the transmission times of different streams relative to each other by adding additional communication services that require resources from the industrial communication network or rearranging network arrangements.

[0019] In a further embodiment of the method, the device includes a sensor and / or an actuator.

[0020] In a further embodiment of the method, the application performance quality (i.e., QoS parameters of the communication flow) is continuously monitored, and an alert is generated if it deviates from a predefined minimum requirement.

[0021] In a further embodiment of the method, state feedback data from the actual machine of the application is used to modify the communication characteristics.

[0022] Whenever an additional application or component is added to the factory using an industrial communication network, or when an existing application or component is modified or removed, the planning and design phase may need to be restarted.

[0023] The present invention further claims an apparatus comprising an industrial communication network and one or more computing devices, wherein the one or more computing devices include a digital twin, a network information provider and monitoring module, a network traffic classification module, and a network controller module. -The digital twin is configured to run a deployed application that is deployable to the industrial communication network, and represents the application, representing a digital representation of an industrial communication system including the industrial communication network, devices connected to the industrial communication network, and processes, as well as representing the physical interactions within the industrial communication system. -The network information provider and monitoring module are configured to retrieve network data traffic characteristics from the communication network, its digital representation, or the simulated or analogized network. It can retrieve measurement data from network devices and application endpoints (such as hosts) or virtualized computing environments (such as virtual machines or software containers). Network data traffic characteristics can be obtained through active or passive monitoring. Typical network data traffic characteristics include end-to-end (e2e) latency of data packets exchanged between application endpoints, bandwidth used by e2e data streams, packet jitter within the data stream, packet loss, packet size, and many more. -The network traffic classification module is configured to classify the endpoint types of the industrial communication network and the detected application data traffic. Endpoint classification can be based on its data traffic characteristics, a well-known area in traditional Internet traffic engineering. However, these traditional approaches have their limitations, and the components of this invention are designed to also utilize more recently developed approaches that employ artificial intelligence and machine learning, such as Thuy TTNguyen, GA (FOURTHQUARTER 2008), A Survey of Techniques for Internet Traffic Classification using Machine Learning, IEEE COMMUNICATIONS SURVEYS & TUTORIALS, VOL.10, NO.4, to capture the specific characteristics of industrial communication data traffic and take into account the specific needs of endpoints. -The network controller module is configured to embed data streams into network devices in the digital twin and the industrial communication network, and is configured to detect active topology, including network device attributes and capabilities. Network device configuration is typically accomplished based on communication service requests from terminal stations (via the user network interface UNI) or via the northbound interface of the central management unit. The industrial communication network mentioned above is a real network configured to operate and function in an industrial plant. Here, real-world applications rely on data from sensors and other sources to control real industrial processes and actuators. -The network traffic classification module is configured to: perform initial detection and classification of the application's endpoints and their corresponding communication characteristics in the industrial communication network based on the collected communication traffic of the application monitored by the network information provider and the monitoring module. -The digital twin is configured to alter the initially detected communication characteristics of the communication link used by the application until a predefined limit for application performance quality is reached. -The network controller module is configured to derive determined network parameters for the communication services required by the application and to configure the industrial communication network based on the determined network parameters.

[0024] In a further embodiment of the device, the communication characteristics include latency, jitter, and packet loss.

[0025] In a further embodiment of the device, the application performance quality includes predefined quality parameters and / or predefined functional envelopes.

[0026] In a further embodiment of the device, altering the communication characteristics includes: adding additional delays to the link; creating packet loss; and / or realigning the transmission times of different streams relative to each other by adding additional communication services that require resources from the industrial communication network or rearranging the communication network arrangements.

[0027] In a further embodiment of the device, the device includes sensors and / or actuators.

[0028] In a further embodiment of the device, the network information provider and monitoring module are configured to: continuously monitor the application performance quality; and generate alerts when deviations from predefined minimum requirements occur.

[0029] In a further embodiment of the device, the digital twin is configured to use state feedback data from a real machine that is specific to the application to modify the communication characteristics.

[0030] Further advantages and benefits of the invention will become apparent upon careful reading of the detailed description with appropriate reference to the accompanying drawings. Attached Figure Description

[0031] Figure 1 A block diagram of the device is shown, and Figure 2 A flowchart illustrating the configuration of an industrial communication network is shown. Detailed Implementation

[0032] As an embodiment of the present invention Figure 1 A block diagram of the communication device is shown. The device has three main components: a network information provider and monitoring module (NIP), a network traffic classification module (NTC), and a network controller module (NC), which interacts with other components in the industrial control system, such as Application A, the digital twin DT, and the industrial communication network IN itself. If the digital twin DT environment is unavailable and the real system is in the testing and configuration phase, the real system can, of course, be used for training. All modules and the digital twin DT can be implemented in one or more computing devices (CDs).

[0033] The Network Information Provider and Monitoring Module (NIP) is a software-implemented component that retrieves network data traffic characteristics from a communication network, a digital representation of that communication network, or a simulated or analogized network. It can retrieve measurement data from network devices and application endpoints (such as hosts) or virtualized computing environments (such as virtual machines or software containers). Network data traffic characteristics can be obtained through active or passive monitoring. Typical network data traffic characteristics include end-to-end (e2e) latency of data packets exchanged between application endpoints, bandwidth used by the e2e data stream, packet jitter within the data stream, packet loss, packet size, and many other characteristics.

[0034] Network Traffic Classification Module (NTC): Industrial communication systems contain many different types of devices (D) (e.g., sensors, actuators, industrial PCs) and different applications (A) (e.g., SCADA software, control applications, monitoring data streams). The NTC can classify endpoint types and classify the detected application data traffic. Endpoint classification can be based on its data traffic characteristics and is well-known in traditional Internet data traffic engineering. However, these traditional approaches have their limitations, and the components in this invention are designed to also use more recently developed approaches that utilize artificial intelligence and machine learning, such as Thuy TTNguyen, GA (FOURTH QUARTER 2008), A Survey of Techniques for Internet Traffic Classification using Machine Learning, IEEE COMMUNICATIONS SURVEYS & TUTORIALS, VOL.10, NO.4, to capture the specific characteristics of industrial communication data traffic and take into account the specific needs of endpoints.

[0035] The Network Controller Module (NC) can configure network devices in the Digital Twin (DT) and industrial communication networks to embed data streams, etc. To achieve this, it can detect active topology, including device attributes and capabilities. Network device configuration is typically based on communication service requests from end stations (via the User Network Interface (UNI)) or from the central management unit's northbound interface.

[0036] Application A refers to any type of application A that exists or should be deployed in the industrial communication network IN. Applications with stringent, though not explicitly defined, requirements related to the quality of service for their communication relationships, such as time-sensitive applications, are of particular interest to this invention.

[0037] Digital twin (DT) represents a complete industrial system, including industrial communication networks, digital representations of equipment (e.g., sensors, actuators), and digital representations of applications, processes, and physical interactions.

[0038] Industrial communication networks (IN) are real networks used to operate and manage industrial plants. Here, real applications control real industrial processes and actuators based on data from sensors and other sources.

[0039] Figure 2 The diagram illustrates the creation of an industrial communication network, in which sensors, actuators, control processes, and applications can be structured into four phases, each containing the main functions of a key component.

[0040] Planning and Design Phase In the initial phase, Application A, which should be deployed in the Industrial Communication Network (IN), is first deployed in the Digital Twin (DT). After Application A is deployed and started in the Digital Twin (DT), the Network Information Provider (NIP) and Monitoring Module (NIP) can begin collecting communication data for Application A. This data may have already been used for initial classification of application endpoint types and corresponding data traffic volumes. The planning and design phase may be restarted whenever additional Application A or components are added to the plant using the Industrial Communication Network (IN), or when existing Application A or components are modified or removed.

[0041] Detection and analysis phase Now, the Network Traffic Classification (NTC) module will identify the type of data flow and endpoint. Here, different approaches can be used, such as Fannia Pacheco, EE (2018), Towards the Deployment of Machine Learning Solutions in Network Traffic Classification: A Systematic Survey. Communications Surveys and Tutorials, IEEE Communications Society, 21(2), S.pp. 1988-2014.

[0042] The classification results are passed to the software-implemented Network Controller Module (NC), enabling the NC to receive relevant data traffic characteristics from the Network Information Provider (NIP) and Monitoring Module (NIP) in the same way it receives communication service requests according to existing technologies. Since evaluating only data from application A operating in an ideal environment is insufficient to create communication patterns useful in the real world, the NC now begins to alter the communication characteristics of the communication links used by applications in the digital twin (DT).

[0043] Changing communication characteristics includes: adding additional delays to the link; creating packet loss; (virtually) adding additional communication services that require resources from the network or rearranging network arrangements to re-align the transmission times of different streams relative to each other. Therefore, the communication of application A is "pushed to its limits," and once application A fails (e.g., the control loop is no longer able to control the physical processes in the digital twin, and the results of the processes are outside the required operational envelope), the digital twin DT will notify the network controller module NC.

[0044] An optional extension to this assessment could be the use of real machines; for example, adding delays to communication in a real system could cause the surface quality after milling to no longer meet requirements. This real-world data could then be used by the network controller module (NC). By using this status feedback from the application, the NC module clearly understands the application's requirements for one or more communication channels—even when the application doesn't explicitly require them. The determined parameter range can then be used to further schedule tasks for the network controller.

[0045] Configuration and deployment phase Following the detection and analysis phase, the Network Controller Module (NC) knows the upper bound (“its limits”) of latency, jitter, and packet loss that the application can tolerate without process quality degradation. From this, it can derive the parameters for the communication services requested by Application A. Typically, these parameters will be significantly lower than those observed by monitoring the application’s “undisturbed” communication during the planning and design phases. The NC then configures the physical industrial communication network (IN), including the end stations, based on the determined parameters. It can then notify Application A or the application deployment framework (e.g., Kubernetes) of the successful execution of their requested communication relationships, thereby triggering production operations.

[0046] Operational phase During operation, real application A exchanges data via a physical communication network. From the detection and analysis phase, the network controller module NC knows the actual QoS requirements for each communication service of application A. In a further embodiment, the network information provider and monitoring module NIP continuously monitor the QoS parameters of the relevant flows and issue an alert to the network controller module NC if the parameters deviate from the defined minimum requirements.

[0047] Traditional approaches require detailed specifications of the application's communication requirements so that the network controller can configure and embed the corresponding data traffic flows. Specifying application communication requirements is a highly complex task, as it involves knowledge of the application's workflow, the physical processes involved, and interfaces with other systems in the production environment. This knowledge is typically not available to a single person but must be consolidated by experts from different domains, such as application developers, subject matter experts for production processes, production operators, and system integrators.

[0048] The present invention provides automated support for describing the requirements of application A, and uses this information to automatically derive the communication quality parameters of application A required for the network controller to properly embed data service flows.

[0049] The described method must address applications following the "sense-analyze-action" paradigm for process control, where actions based on received sensor data readings are given in equal cycles—for example, receiving sensor data at the beginning of the cycle, then calculating the response, and sending the response command to the process actuator at the end of the cycle. However, the described approach can also be used for parallel processing of sensor readings and control commands—for example, where the cycle is not the primary criterion for sending and receiving, and the control response is calculated once sensor data readings are received (and this does not necessarily occur at the beginning of the cycle).

[0050] Although the invention has been explained with respect to its preferred embodiments as mentioned above, it should be understood that many other possible modifications and variations can be made without departing from the scope of the invention. Therefore, it is conceivable that one or more of the appended claims will cover such modifications and variations that fall within the true scope of the invention.

[0051] List of reference numerals in the attached figures Application A CD computing device D device DT Digital Twin IN Industrial Communication Network NC Network Controller NIP Network Information Provider and Monitoring Module NTC Network Traffic Classification Module

Claims

1. An automated method for configuring an industrial communication network (IN), comprising the following steps: - Deploy an application (A) that is deployable to the industrial communication network (IN) onto a digital twin (DT), wherein the digital twin is configured to: represent the application, represent a digital representation of an industrial communication system including the industrial communication network (IN), devices (D) connected to the industrial communication network, and processes, and represent the physical interactions within the industrial communication system. - The application (A) is run on the digital twin (DT), where - Collect data on the communication traffic volume of the application. - Based on the collected communication traffic volume of the applications, the application endpoints and their corresponding communication characteristics in the industrial communication network are initially detected and classified. - Change the initially detected communication characteristics of the communication link used by the application until a predefined limit for application performance quality is reached, and - Derive the network parameters for the determined communication characteristics required by the application from these parameters. - And configure the industrial communication network based on the determined network parameters.

2. The method according to claim 1, wherein Communication characteristics include latency, jitter, and packet loss.

3. The method according to claim 1 or 2, wherein The application performance quality includes predefined quality parameters and / or predefined functional envelopes.

4. The method according to claim 1 or 2, wherein changing the communication characteristics includes: Add additional delay to the link; Group creation failed; And / or by adding additional communication services that require resources from the industrial communication network or by rearranging the communication network arrangements, the transmission times of different streams can be realigned relative to each other.

5. The method according to claim 1 or 2, wherein the device comprises a sensor and / or an actuator.

6. The method of claim 1 or 2, wherein the application performance quality is continuously monitored, and an alarm is generated if the performance deviates from a predefined minimum requirement.

7. The method of claim 6, wherein the application performance quality includes QoS parameters of the communication stream.

8. The method according to claim 1 or 2, wherein the state feedback data of the real machine dedicated to the application is used to modify the communication characteristics.

9. An apparatus comprising an industrial communication network (IN) and one or more computing devices (CD), The one or more computing devices (CDs) mentioned above include a digital twin (DT), a network information provider and monitoring module (NTP), a network traffic classification module (NTC), and a network controller module (NC). - Wherein the digital twin (DT) is configured to run a deployed application (A) that is deployable to the industrial communication network, representing the application, representing a digital representation of an industrial communication system including the industrial communication network, devices (D) connected to the industrial communication network, and processes, and representing the physical interactions within the industrial communication system. - The Network Information Provider and Monitoring Module (NTP) is configured to retrieve network data traffic characteristics from a communication network, a digital representation of that communication network, or a simulated or analogized network. - The Network Traffic Classification (NTC) module is configured to classify the endpoint types of the industrial communication network and the detected application data traffic. - The network controller module (NC) is configured to configure network devices in the digital twin and the industrial communication network to embed data streams, and is configured to detect active topology, including network device attributes and capabilities. - The industrial communication network (IN) mentioned above is a real network configured to operate and manage industrial plants. - The Network Traffic Classification (NTC) module is configured to: perform initial detection and classification of the application's endpoints and their corresponding communication characteristics in the industrial communication network based on the collected communication traffic of the application monitored by the network information provider and the monitoring module. - The digital twin (DT) is configured to alter the initially detected communication characteristics of the communication link used by the application until a predefined limit for application performance quality is reached. - The network controller module (NC) is configured to derive determined network parameters for the communication services required by the application and to configure the industrial communication network based on the determined network parameters.

10. The apparatus of claim 9, wherein the communication characteristics include delay, jitter, and packet loss.

11. The apparatus according to claim 9 or 10, wherein The application performance quality includes predefined quality parameters and / or predefined functional envelopes.

12. The apparatus according to claim 9 or 10, wherein Changing the communication characteristics includes: Add additional delay to the link; Group creation failed; And / or by adding additional communication services that require resources from the industrial communication network or by rearranging network arrangements, the transmission times of different streams can be realigned relative to each other.

13. The apparatus according to claim 9 or 10, wherein The device includes sensors and / or actuators.

14. The apparatus according to claim 9 or 10, wherein The Network Information Provider and Monitoring Module (NTP) is configured to: continuously monitor the performance quality of the application; and generate alerts when deviations from predefined minimum requirements occur.

15. The apparatus according to claim 9 or 10, wherein The digital twin (DT) is configured to use state feedback data from a real machine that is dedicated to the application to modify the communication characteristics.

Images