Information processing device

What is AI technical title?
AI technical title is built by Patsnap AI team. It summarizes the technical point description of the patent document.
The information processing device addresses network load and data dilution issues by converting and compressing sensor data based on network conditions, ensuring real-time feature data transmission and efficient raw data storage.

JP2025165192APending Publication Date: 2025-11-04HITACHI INFORMATION & TELECOMM ENG LTD

View PDF 1 Cites 0 Cited by

Patent Information

Application Number: JP2024069153
Authority / Receiving Office: JP · JP
Patent Type: Applications
Current Assignee / Owner
Filing Date: 2024-04-22
Publication Date: 2025-11-04

AI Technical Summary

Technical Problem

Existing IoT systems face challenges in managing high-speed sampling data from sensors and PLCs, leading to excessive network communication load and data dilution when uploaded to cloud systems, while security measures are inadequate.

Method used

An information processing device with an analog-to-digital converter, processor, memory, and network interface that converts and compresses data, generates feature data, and transmits it based on network load, storing excess data for later transmission.

Benefits of technology

Enables real-time transmission of feature data and efficient storage of raw data, maintaining data accuracy and reducing network load, allowing for timely event triggering and analysis in cloud systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

Figure 2025165192000001_ABST

Patent Text Reader

Abstract

To provide an information processing device capable of collecting measurement data which are acquired from a sensor, a PLC, etc., transmitting feature amount data required for an event ignition such as a warning in real time on the basis of the measurement data, and maintaining accuracy of data analysis in a host cloud system even in a case where a data communication volume of the measurement data is large.SOLUTION: An information processing device 1 comprises an analog input IF 101, an A / D converter 102, a CPU 103, a memory 104, and a network communication IF 105. In a case where a communication load of a network is smaller than a first communication load, the CPU 103 transmits digital Raw data on which A / D conversion processing is performed to a cloud server 4 together with feature amount data which are extracted from a Raw. In a case where the communication load is equal to or more than the first communication load, the CPU transmits only the feature amount data to the could server 4, the digital Raw data are stored in the memory 104 and the transmission thereof is on standby.SELECTED DRAWING: Figure 2

Need to check novelty before this filing date? Find Prior Art

Description

[Technical Field]

[0001] The present invention relates to an information processing device. [Background technology]

[0002] As an example of a condition monitoring system that uses measurement data from sensors installed in production facilities, etc., Patent Document 1 states, "A condition monitoring system that monitors the condition of facilities, comprising: a sensor attached to the facilities; a data measurement device that receives a detection signal from the sensor and acquires measurement data from the detection signal in accordance with predetermined measurement conditions; and a data diagnosis device that receives the measurement data from the data measurement device and executes a diagnosis process to diagnose the condition of the facilities based on the measurement data, wherein the data diagnosis device has an edge application and an industrial IoT platform, and the edge application has a data collection and analysis unit that calculates feature quantities of the measurement data from the data measurement device and distributes the feature quantities to the industrial IoT platform. (Abstract)" [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2022-137947 Summary of the Invention [Problem to be solved by the invention]

[0004] When IoT technology is used to send measurement data from sensors and programmable logic controllers (PLCs) installed in production facilities and equipment to a host cloud system for analysis, the high-speed sampling performed by the sensors and PLCs can result in a huge volume of measurement data. Therefore, in consideration of the network communication load, data compression and extraction and transmission of only the necessary data are being undertaken. Therefore, a dedicated controller is used to analyze the high-speed sampling data and transmit the extracted data to a host cloud system.

[0005] However, as a result of analysis by the dedicated controller, the measurement data output from the sensor or PLC is processed into data that indicates whether or not the measurement value exceeds the threshold. As a result, the processed data does not provide information on how much the threshold has been exceeded, or how the measurement value has changed even if it is below the threshold, which poses the problem that the upper cloud system cannot analyze time-series changes in the measurement value.

[0006] Furthermore, when measurement data is sent to a cloud system, it must be encrypted to ensure security and protected from unauthorized access, which means that the connection destination and protocol must be changed to match the specifications of the cloud system to which the data is sent.

[0007] According to Patent Document 1, in the case of high-speed sampling, data processing can be performed on the edge side, but uploading the measurement data to a cloud system is not included. As a result, it does not solve the issues of consideration for the data communication load and security measures that must be taken into account when uploading measurement data from sensors, PLCs, etc. to a cloud system, and the issue of information originally contained in the measurement data being diluted as a result of analysis by a dedicated controller.

[0008] The present invention has been made to solve the above-mentioned problems, and aims to provide an information processing device that collects measurement data obtained from sensors, PLCs, etc. that are installed in production equipment and control devices to obtain operating conditions, transmits feature data necessary for triggering events such as alarms based on the measurement data in real time, and can maintain the accuracy of data analysis in a higher-level cloud system even when the data communication volume of the measurement data is large. [Means for solving the problem]

[0009] In order to solve the above problems, the present invention has the configurations described in the claims. As an example, the present invention is an information processing device comprising an analog input interface, an analog-to-digital converter, a processor, a memory, and a network communication interface, wherein the analog-to-digital converter digitally converts raw data consisting of analog data received via the analog input interface to generate digital raw data, the processor detects features included in the digital raw data to generate feature data, detects a communication load of a network to which the information processing device is connected via the network communication interface, and transmits the digital raw data together with the feature data to a cloud server if the communication load is smaller than a first communication load that allows transmission of the digital raw data, and transmits the feature data to the cloud server if the communication load is equal to or greater than the first communication load, and stores the digital raw data in the memory while waiting for transmission to the cloud server.

[0010] The present invention also provides an information processing device comprising an analog input interface, an analog-to-digital converter, a processor, a memory, and a network communication interface, wherein the analog-to-digital converter digitally converts raw data consisting of analog data received via the analog input interface to generate digital raw data, the processor detects features included in the digital raw data to generate feature data, transmits the feature data to a cloud server, and stores the digital raw data in the memory.

[0011] The present invention also provides an information processing device comprising a digital input interface, a processor, a memory, and a network communication interface, wherein the processor detects features contained in Raw data consisting of digital data received via the digital input interface to generate feature data, detects a communication load of a network to which the information processing device is connected via the network communication interface, and if the communication load is smaller than a first communication load that allows transmission of the Raw data, transmits the Raw data together with the feature data to a cloud server, and if the communication load is equal to or greater than the first communication load, transmits the feature data to the cloud server and stores the Raw data in the memory while waiting for transmission of the Raw data to the cloud server.

[0012] The present invention also provides an information processing device comprising a digital input interface, a processor, a memory, and a network communication interface, wherein the processor detects features contained in raw data consisting of digital data received via the digital input interface, generates feature data, transmits the feature data to a cloud server, and stores the raw data in the memory. [Effects of the Invention]

[0013] According to the present invention, it is possible to provide an information processing device that can collect measurement data acquired from sensors, PLCs, etc. that are installed in production equipment and control devices to acquire operating conditions, transmit feature data necessary for activating events such as alarms based on the measurement data in real time, and maintain the accuracy of data analysis in a host cloud system even when the data communication volume of the measurement data is large. Note that objects, configurations, and effects other than those described above will be made clear in the following embodiments. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a system configuration diagram using an information processing device according to an embodiment of the present invention. [Figure 2] FIG. 2 is a functional block diagram of the information processing device. [Figure 3] 10 is a flowchart showing a processing flow in the information processing device. [Figure 4] FIG. 10 is a diagram showing a comparison of digital raw data before and after compression. [Figure 5] FIG. 10 illustrates an example of a network load monitoring process. [Figure 6] FIG. 1 is a diagram illustrating a problem in a conventional example. DETAILED DESCRIPTION OF THE INVENTION

[0015] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The same components are designated by the same reference numerals throughout the drawings, and redundant explanations will be omitted.

[0016] FIG. 1 is a diagram showing the system configuration of an information processing apparatus according to this embodiment.

[0017] The information processing device 1 according to this embodiment has a function as an IoT gateway that collects and converts measurement data output by sensors, such as a vibration sensor 2a and a programmable logic controller (PLC) 2b installed in production equipment and control devices, that measure vibration, temperature, and the like and output the measurement data, and a data transmission function that transmits the measurement data to a cloud server 4 as an external device via a network 3. The information processing device 1 also has a data storage function for analog data. The cloud server 4 is an example of an external device and may be an on-premises server. The network 3 may be, for example, a public line for a cloud server or a LAN for an on-premises server.

[0018] The vibration sensor 2a transmits full waveform data of about several kilohertz to several tens of kilohertz (about 0.1 ms to 1 ms) to the information processing device 1. The full waveform data output from the vibration sensor 2a is analog data. Measurement data made up of this full waveform data is called analog raw data.

[0019] The PLC 2b collects measurement data from measuring instruments provided in various devices to be controlled and transmits the data to the information processing device 1. This measurement data is digital data.

[0020] The information processing device 1 includes an analog input interface 101, an analog-to-digital converter 102, a CPU 103 (corresponding to a processor), a memory 104, a network communication interface 105, and a digital input interface 106. As an example of implementing the information processing device 1, for example, the information processing device 1 according to this embodiment can be realized by incorporating software that realizes the functions of an IoT-GW, a transmission function, and a data storage function into a small personal computer. As an example of an OS used in the information processing device 1, Ubuntu (an example of a distribution of Linux (registered trademark)) may be used.

[0021] The analog input interface 101 and analog-to-digital converter 102 are realized by, for example, an analog-to-digital conversion board that supports high-speed sampling, or a USB-connected analog-to-digital converter. Use of this converter enables high-speed sampling of sensor data (at a cycle of about several tens of kHz).

[0022] The digital input interface 106 is realized by a data acquisition IF that can be used with a LAN interface or various field networks, and a high-speed GPIO, for example, and receives measurement data (called "digital raw data") made up of digital data from the PLC 2b.

[0023] FIG. 2 is a functional block diagram of the information processing device 1.

[0024] The information processing device 1 includes a data reception processing unit 1031, a feature data generation unit 1032, a data transmission processing unit 1033, a network load monitoring unit 1034, and a compression processing unit 1035. Each of these units is realized by the CPU 103 executing software that realizes the function of each unit. The function of each unit will be described with reference to the following flowchart.

[0025] FIG. 3 is a flowchart showing the flow of processing in the information processing device 1.

[0026] When the information processing device 1 receives analog raw data from the vibration sensor 2a via the analog input interface 101 (S01), the analog-to-digital converter 102 performs AD conversion on the received analog raw data to generate digital raw data (S02). In step S01, when the information processing device 1 receives digital raw data from the PLC 2b via the digital input interface 106 (S01), step S02 is skipped.

[0027] The data reception processing unit 1031 performs preprocessing on the digital raw data before storing the data (S03). Examples of preprocessing include protocol and data format conversion processing (e.g., format conversion processing such as MQTT) by the compression processing unit 1035, timestamp assignment processing, and channel number and sensor name information assignment processing. The channel number and sensor name information is information that identifies which channel of which vibration sensor the analog raw data or digital raw data was output from.

[0028] The compression processing unit 1035 performs conversion to MQTT format as an example of lossless compression processing. Data converted to MQTT format corresponds to raw data after compression. MQTT is a relatively lightweight text-based format and is characterized by its small header. For example, when sending measurement data every 1 msec / 1 kHz as an example of data transmission using general MQTT, if one MQTT message is inserted between each MQTT message and sent, the overhead in the header section increases, resulting in an increase in the amount of data transferred.

[0029] Therefore, the compression processing unit 1035 compresses the transmission data by transmitting the digital raw data in batches at specific time intervals. By combining the header parts into one and inserting data measured multiple times into a single MQTT message, the overhead of the transmission data can be reduced, enabling efficient data transfer.

[0030] FIG. 4 is a diagram showing a comparison of digital raw data before and after compression.

[0031] 4, the compression processing unit 1035 reduces overhead by omitting part of the header of the digital raw data, thereby compressing the digital raw data for each 1 msec / 1 kHz from approximately 544 bytes to approximately 230 bytes.

[0032] The data reception processing unit 1031 temporarily stores the compressed raw data in a data storage area secured in the memory 104 (S04). In this embodiment, the lossless compression processing by the compression processing unit 1035 has been described as part of the data pre-processing, but the data pre-processing may not include the compression processing, and the digital raw data may be temporarily stored in the memory 104. In that case, the compression processing unit 1035 may perform the lossless compression processing on the digital raw data before the raw data transmission processing in step S09, which will be described later, and transmit the compressed raw data.

[0033] The feature data generation unit 1032 compares the measurement value indicated by the compressed raw data with the warning threshold, and generates feature data indicating whether the measurement value is equal to or greater than the warning threshold. Then, the data transmission processing unit 1033 transmits the feature data to the cloud server 4 via the network 3 (S05).

[0034] The network load monitoring unit 1034 monitors the communication load of the network 3 (S06). As an example of the process of monitoring the communication load of the network, the network load monitoring unit 1034 sends a ping command to the cloud server 4 and measures the communication load based on the response time. The network load monitoring process will be described in detail later.

[0035] If the communication load is smaller than the first communication load that allows transmission of the digital raw data (or compressed raw data) (S07: Yes), the data transmission processing unit 1033 performs pre-transmission processing on the real-time digital raw data (or compressed raw data) (S08). The pre-transmission processing includes, for example, protocol / data format conversion processing to conform to the communication protocol between the cloud server 4 and the information processing device 1, and security countermeasure processing. The security countermeasure processing includes, for example, processing to guarantee the sender and data using a PKI (Public Key Infrastructure) public key cryptosystem. Furthermore, if the digital raw data has not been compressed, the compression processing unit 1035 may perform reversible compression processing on the digital raw data as pre-transmission processing.

[0036] The data reception processing unit 1031 transmits the feature data and the real-time compressed raw data to the cloud server 4 (S09). If the processing is to be continued (S10: No), the process returns to step S01, and if an end operation or end instruction is given (S10: Yes), the process ends.

[0037] On the other hand, in step S07, if the network load monitoring unit 1034 determines that the communication load is equal to or greater than the first communication load (S07: No), the data transmission processing unit 1033 transmits only the feature data and waits for transmission of the compressed raw data (S11). Thereafter, the process from receiving analog raw data or digital raw data (S01) to processing the feature data (S05) and monitoring the network communication load (S06) is performed again (S12).

[0038] If the network load monitoring unit 1034 determines that the communication load remains equal to or greater than the first communication load after waiting for transmission of the compressed raw data (S13: Yes), it transmits only the feature data, as in step S11, and waits for transmission of the compressed raw data to the cloud server 4 (S14). Thereafter, it returns to step S12 again, and performs processes from receiving analog raw data or digital raw data (S01) to processing the feature data (S05) and monitoring the communication load of the network (S06).

[0039] On the other hand, in step S13, if the network load monitoring unit 1034 determines that the communication load on the network 3 is less than the first communication load, that is, the communication load has become less than the first communication load while waiting to transmit the compressed raw data (S13: No), the data transmission processing unit 1033 executes pre-transmission processing on the compressed raw data that was waiting to be transmitted (S15) and transmits it to the cloud server 4 (S16). If the processing continues (S17: No), the process returns to step S01, and if an end operation or end instruction has been received (S17: Yes), the process ends.

[0040] FIG. 5 illustrates an example of the network load monitoring process.

[0041] 5, a ping response time of 30 milliseconds is set as the first communication load. Therefore, from time t1 to t5, the response time is less than 30 milliseconds (faster than 30 milliseconds), so feature data and compressed raw data are transmitted in real time. Here, "real time" means that the digital raw data or compressed raw data is stored in memory 104 and transmitted without waiting for transmission. Even if there is a time difference due to the processing time of the information processing device 1 between the reception of the analog raw data, the AD conversion processing or the reception of the digital raw data, and the transmission of the compressed raw data to the cloud server 4, the time difference is included in "real time" transmission as long as no transmission waiting processing is performed.

[0042] The ping response time is 30 milliseconds or more from time t5 to time t10, so the information processing device 1 determines that this is a period of high network communication load, transmits only feature data, and waits to transmit the compressed raw data.

[0043] At time t10, the ping response time becomes less than 30 milliseconds (faster than 30 milliseconds), so the information processing device 1 compresses the compressed raw data that has been waiting to be sent and then sends it to the cloud server 4. It also transmits the feature data and the compressed raw data in real time.

[0044] The effects of the information processing device 1 according to this embodiment will be described with reference to Fig. 6. Fig. 6 is a diagram illustrating the problems with the conventional example.

[0045] Conventionally, the analog raw data output from the vibration sensor 2a has been collected by the vibration analysis controller 50. The analog raw data output from the vibration sensor 2a is full-waveform analog data of about several kHz to several tens of kHz (about 0.1 ms to 1 ms), and therefore contains a large amount of information and data. Therefore, only when a predetermined threshold is exceeded, the vibration analysis controller 50 transmits a signal indicating that the threshold has been exceeded to the IoT-GW 51, and the IoT-GW 51 then transmits the signal to the cloud server 4 via the network 3. As a result, the amount of data communication between the vibration analysis controller 50, IoT-GW 51, and cloud server 4 is about 10 seconds to several tens of seconds, which is smaller than the amount of data for full-waveform data.

[0046] However, because the vibration analysis controller 50 only outputs feature amount data indicating that a predetermined threshold has been exceeded, the cloud server 4 is unable to see the deterioration of the vibration amount over time, and the amount of information contained in the data is diluted, which is a problem. Also, even if MQTT is used and the header size is small, it is difficult to send the measurement data obtained by high-speed sampling to the cloud server 4 at a frequency of about several milliseconds.

[0047] In contrast, the information processing device 1 according to the present embodiment is equipped with the functions of the vibration analysis controller 50 and the IoT-GW 51, thereby enabling it to collect full waveform data (analog raw data) measured by the vibration sensor 2a. Similarly, it can also collect digital raw data from the PLC 2b. Then, feature data is generated based on the analog raw data or digital raw data and transmitted to the cloud server 4, allowing the cloud server 4 to trigger an event, such as issuing an alarm, in real time. At the same time, the information processing device 1 monitors the communication load on the network 3, and when the communication load is relatively light, it transmits the digital raw data (or compressed raw data) to the cloud server 4, allowing the cloud server 4 to perform signal analysis processing based on the digital raw data (or compressed raw data). On the other hand, when the communication load is relatively high, the information processing device 1 temporarily suspends transmission of the digital raw data (or compressed raw data) and transmits the full waveform data to the cloud server 4 after the communication load has lightened. This allows information diluted by the feature data to be transmitted to the cloud server 4. Furthermore, transmission of feature data, which requires real-time response, can also be achieved solely by the information processing device 1.

[0048] As described above, the information processing device 1 according to this embodiment provides a data acquisition interface and data storage area constructed on the IoT-GW 51 without using a separate vibration analysis controller 50, and performs compression processing on the data to be transmitted depending on the network load status, thereby making it possible to acquire data that requires high-speed sampling of several kilohertz to several tens of kilohertz. As a result, by operating preventive maintenance and predictive maintenance applications deployed on the cloud or the IoT-GW, it is possible to perform appropriate maintenance and repair of production equipment, thereby improving quality and operating rates.

[0049] The above-described embodiment is one embodiment of the present invention, and the present invention is not limited to the above-described embodiment, and various modifications that do not depart from the spirit of the present invention are also included in the present invention.

[0050] For example, in the above-described information processing device 1, digital raw data is transmitted to the cloud server 4 in accordance with the communication load of the network, but it is also possible to transmit only feature data from the information processing device 1 to the cloud server 4 and store the digital raw data in the memory 104. This enables time series analysis using digital raw data on the edge side in an IoT environment. Another advantage is that there is no communication load on the network associated with transmitting digital raw data.

[0051] Furthermore, when the network load is high and transmission of the digital raw data is put on hold, the compression processing unit 1035 may perform lossless compression processing to generate compressed raw data, which may then be stored in the memory 104. Then, when the network load decreases, the compressed raw data may be transmitted, thereby saving the storage capacity of the memory 104.

[0052] Furthermore, lossless compression processing by the compression processing unit 1035 is not essential, and digital raw data may be transmitted and received over the network 3 if there is no problem with the network communication load. [Explanation of symbols]

[0053] 1: Information processing equipment 2a: Vibration sensor 2b:PLC 3: Network 4: Cloud server 50: Vibration analysis controller 51: IoT-GW 101: Analog input interface 102: Analog-to-digital converter 103: CPU 104: Memory 105: Network communication interface 106: Digital input interface 1031: Data receiving processing unit 1032: Feature data generation unit 1033: Data transmission processing unit 1034: Network Load Monitoring Unit 1035: Compression processing unit

Claims

1. An information processing device, an analog input interface; An analog-to-digital converter, a processor; Memory and a network communication interface; the analog-to-digital converter converts raw data, which is analog data received via the analog input interface, into digital raw data, and generates digital raw data; the processor detects features included in the digital raw data and generates feature data; detecting a communication load of a network to which the information processing device is connected via the network communication interface; If the communication load is smaller than a first communication load that allows transmission of the digital raw data, the digital raw data is transmitted to a cloud server together with the feature amount data; If the communication load is equal to or greater than the first communication load, the feature amount data is transmitted to the cloud server, and the digital raw data is stored in the memory while waiting for transmission to the cloud server.