WO2019176183A1 - Sensor control device, sensor control method, and program - Google Patents

Sensor control device, sensor control method, and program Download PDF

Info

Publication number
WO2019176183A1
WO2019176183A1 PCT/JP2018/043933 JP2018043933W WO2019176183A1 WO 2019176183 A1 WO2019176183 A1 WO 2019176183A1 JP 2018043933 W JP2018043933 W JP 2018043933W WO 2019176183 A1 WO2019176183 A1 WO 2019176183A1
Authority
WO
WIPO (PCT)
Prior art keywords
sensor
processing module
sensors
sensing data
predetermined condition
Prior art date
Application number
PCT/JP2018/043933
Other languages
French (fr)
Japanese (ja)
Inventor
修一 三角
泰司 吉川
哲二 大和
Original Assignee
オムロン株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by オムロン株式会社 filed Critical オムロン株式会社
Publication of WO2019176183A1 publication Critical patent/WO2019176183A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00—Data switching networks
    • H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]

Definitions

  • the present invention relates to a sensor control device, a sensor control method, and a program.
  • Patent Document 1 JP 2017-182596 A discloses a data collection and analysis platform.
  • sensing data is collected from various sensors connected to a network, and analysis is performed based on the collected sensing data (see Patent Document 1).
  • the processing module detects the occurrence of an event based on a plurality of collected sensing data. In such a case, when each sensor performs high-speed sampling and all the sensing data is collected, the processing module can detect the occurrence of the event with high accuracy.
  • the present invention has been made to solve such a problem, and an object of the present invention is to provide a sensor control device, a sensor control method, and a program capable of achieving both high accuracy of event detection and reduction of communication traffic. Is to provide.
  • a sensor control device controls each of a plurality of sensors.
  • Each of at least two or more sensors included in the plurality of sensors transmits sensing data to the processing module.
  • the processing module detects the occurrence of an event based on the received plurality of sensing data.
  • the sensor control device includes a determination unit and a sensor control unit.
  • the determination unit determines whether or not a detection result by any one of the two or more sensors satisfies a predetermined condition indicating that the event is highly likely to occur. When it is determined that the detection result satisfies the predetermined condition, the sensor control unit transmits sensing data having a larger amount of information to the processing module than when the detection result is determined not to satisfy the predetermined condition. Control other sensors included in the two or more sensors.
  • the sensor control device when it is detected that one of the two or more sensors transmitting sensing data to the processing module is likely to cause the event, the sensor control device includes the two or more sensors. Sensing data having a larger amount of information than normal (when the detection result does not satisfy the predetermined condition) is transmitted to the processing module. That is, according to this sensor control device, when the possibility that the event has occurred is high, the occurrence of the event is detected with high accuracy based on sensing data having a large amount of information (many), When the possibility that the event has occurred is low, the amount of data (information amount) transmitted from each sensor to the processing module is suppressed. Therefore, according to this sensor control device, it is possible to achieve both high accuracy of event detection and reduction of communication traffic.
  • the determination unit extracts some sensors from a plurality of sensors. Some of the sensors extracted by the determination unit transmit sensing data to a sensor that is determined to have a detection result satisfying a predetermined condition and a common processing module. When it is determined that the detection result satisfies the predetermined condition, the sensor control unit transmits sensing data having a larger amount of information to the processing module than when the detection result is determined not to satisfy the predetermined condition. You may control the sensor extracted by the determination part.
  • the sensor determined that the detection result satisfies the predetermined condition and the other sensor that transmits the sensing data to the common processing module are extracted by the determination unit, and the detection result satisfies the predetermined condition In this case, sensing data having a larger amount of information than usual is transmitted from the sensor extracted by the determination unit to the processing module. Therefore, according to this sensor control device, when the detection result satisfies a predetermined condition, the data transmission amount from an appropriate sensor can be increased.
  • the determination unit may extract the sensors from the plurality of sensors by referring to metadata that associates the plurality of sensors that transmit the sensing data to the common processing module.
  • this sensor control device a plurality of sensors each outputting sensing data to a common processing module are extracted by referring to the metadata. That is, according to this sensor control device, it is possible to extract a plurality of sensors each transmitting sensing data to a common processing module simply by referring to metadata.
  • a virtual sensor may be formed by a processing module and a sensor that transmits sensing data to the processing module.
  • the processing module may detect the occurrence of an event and calculate the likelihood based on time series data input to each of a plurality of input channels.
  • the processing module detects the occurrence of an event based on time series data. Therefore, according to this sensor control device, when a large amount of data is input from the actual sensor to the processing module according to the control of the sensor control device, data that is significantly different from the timing data in the time series is regarded as noise. In addition, false detection caused by noise can be suppressed.
  • the sensor control method controls each of a plurality of sensors.
  • Each of at least two or more sensors included in the plurality of sensors transmits sensing data to the processing module.
  • the processing module detects the occurrence of an event based on the received plurality of sensing data.
  • the sensor control method includes a step of determining whether or not a detection result by any one of the two or more sensors satisfies a predetermined condition indicating that the event is highly likely to occur, When it is determined that the result satisfies the predetermined condition, the two or more sensors are configured to transmit sensing data having a larger amount of information to the processing module than when the detection result is determined not to satisfy the predetermined condition. Controlling other included sensors.
  • the sensor control method when it is detected that one of the two or more sensors that transmit sensing data to the processing module is likely to cause the event, the sensor control method includes the two or more sensors. Sensing data having a larger amount of information than normal (when the detection result does not satisfy the predetermined condition) is transmitted to the processing module. That is, according to this sensor control method, when there is a high possibility that the event has occurred, the occurrence of the event is detected with high accuracy based on sensing data having a large amount of information. When there is a low possibility that data is transmitted, the amount of data transmitted from each sensor to the processing module is suppressed. Therefore, according to this sensor control method, both high accuracy of event detection and reduction of communication traffic can be achieved.
  • the program according to another aspect of the present invention is configured to cause a computer to execute processing for controlling each of a plurality of sensors.
  • Each of at least two or more sensors included in the plurality of sensors transmits sensing data to the processing module.
  • the processing module detects the occurrence of an event based on the received plurality of sensing data.
  • the program determines whether or not a detection result of any one of the two or more sensors satisfies a predetermined condition indicating that the event is highly likely to occur, and the detection result is a predetermined value.
  • the step of controlling the sensors is executed by a computer.
  • this program When this program is executed by a computer, the two or more sensors are detected when it is detected by one of the two or more sensors that transmit sensing data to the processing module that the event is highly likely to occur. Sensing data having a larger amount of information than normal (when the detection result does not satisfy the predetermined condition) is transmitted to the processing module by the other sensors included in. That is, according to this program, when there is a high possibility that the event has occurred, the occurrence of the event is detected with high accuracy based on sensing data having a large amount of information. When the possibility of being present is low, the amount of data transmitted from each sensor to the processing module is suppressed. Therefore, according to this program, both high accuracy of event detection and reduction of communication traffic can be achieved.
  • the present invention it is possible to provide a sensor control device, a sensor control method, and a program that can achieve both high accuracy of event detection and reduction of communication traffic.
  • FIG. 3 is a diagram for describing an overview of a sensor control device in the first embodiment.
  • 1 is a diagram illustrating an example of a sensor network system in Embodiment 1.
  • FIG. 3 is a diagram illustrating an example of a hardware configuration of a virtual sensor management server in the first embodiment.
  • 3 is a diagram illustrating an example of a software configuration of a virtual sensor management server according to Embodiment 1.
  • FIG. It is a figure for demonstrating the tentative example in which the processing module cannot detect generation
  • 6 is a diagram illustrating an example of a change in the amount of data transmitted from each real sensor to a processing module in Embodiment 1.
  • FIG. 1 is a diagram illustrating an example of a sensor network system in Embodiment 1.
  • FIG. 3 is a diagram illustrating an example of a hardware configuration of a virtual sensor management server in the first embodiment.
  • 3 is a diagram illustrating
  • FIG. 10 is a diagram for describing an overview of a sensor control device in a second embodiment. It is a figure which shows an example of the sensor network system in Embodiment 2.
  • FIG. 6 is a diagram illustrating an example of a hardware configuration of a virtual sensor management server in Embodiment 2.
  • FIG. 6 is a diagram illustrating an example of a software configuration of a virtual sensor management server in Embodiment 2.
  • FIG. 10 is a diagram illustrating an example of an event detection operation in the second embodiment.
  • this embodiment an embodiment according to one aspect of the present invention (hereinafter, also referred to as “this embodiment”) will be described in detail with reference to the drawings.
  • the same or corresponding portions are denoted by the same reference numerals and description thereof will not be repeated.
  • the present embodiment described below is merely an example of the present invention in all respects.
  • Various improvements and modifications can be made to the present embodiment within the scope of the present invention. That is, in carrying out the present invention, a specific configuration can be appropriately adopted according to the embodiment.
  • FIG. 1 is a diagram for describing an overview of a sensor control device (module) 120 according to the first embodiment.
  • the processing module 110 has a plurality of input ports (input channels), and sensing data output by the actual sensor 12 (an example of a sensor) is input to each input port.
  • each real sensor 12 and the processing module 110 can communicate via the internet, for example.
  • the processing module 110 is configured to detect the occurrence of an event based on a plurality of received sensing data and output a detection result. That is, a so-called virtual sensor is formed by the processing module 110 and the real sensor 12 (input sensor) that outputs sensing data to the processing module 110.
  • a virtual sensor outputs sensing results (for example, whether an event has occurred) as sensing data based on sensing data generated by the input sensor observing the target, which is different from the target observed by the input sensor. It is a sensor module. The virtual sensor will be described in detail later.
  • the processing module 110 when each real sensor 12 performs high-speed sampling of sensing data and all the sensing data is transmitted to the processing module 110, the processing module 110 is based on a large number of sensing data (a large amount of information). Thus, the occurrence of the event can be detected with high accuracy.
  • each actual sensor 12 performs high-speed sampling of sensing data
  • a large amount (for example, all) of sensing data is transmitted to the processing module 110
  • the data transmitted from each actual sensor 12 to the processing module 110 is transmitted from each actual sensor 12 to the processing module 110.
  • the amount (communication traffic) increases.
  • only a small amount of sensing data is transmitted from each real sensor 12 to the processing module 110, the detection accuracy of event occurrence decreases.
  • sensor control device 120 it is determined whether or not the detection result by one of actual sensors 12A and 12B satisfies a predetermined condition indicating that there is a high possibility that an event has occurred. Is done.
  • the sensor control device 120 determines that the detection result satisfies the predetermined condition, the sensor control device 120 detects sensing data having a larger amount of information (for example, the sampling frequency is higher than that when the detection result is determined not to satisfy the predetermined condition).
  • the other real sensor 12A, 12B is controlled so as to transmit a high (and / or a large number of quantization bits) to the processing module 110.
  • the sensor control device 120 controls the actual sensor 12B so as to transmit sensing data having a larger amount of information than usual.
  • the sensor control device 120 when there is a high possibility that an event has occurred, the occurrence of the event is detected with high accuracy based on sensing data having a large amount of information, while the event has occurred. When the possibility is low, the amount of data transmitted from each actual sensor 12 to the processing module 110 is suppressed. Therefore, according to this sensor control device 120, it is possible to achieve both high accuracy of event detection and reduction of communication traffic.
  • FIG. 2 is a diagram showing an example of the sensor network system 10 including the sensor control module (device) 120 according to the first embodiment.
  • the sensor network system 10 includes a sensor network unit 14, a virtual sensor management server 100, and an application server 300.
  • the sensor network unit 14, the virtual sensor management server 100, and the application server 300 are connected to each other via the Internet 15 so that they can communicate with each other.
  • the number of each component (virtual sensor management server 100, application server 300, sensor network adapter 11, actual sensor 12, etc.) included in the sensor network system 10 is not limited to that shown in FIG.
  • sensing data generated by the actual sensor 12 or the like can be distributed.
  • sensing data generated by the real sensor 12 can be distributed to the virtual sensor management server 100, and sensing data generated by the virtual sensor can be distributed to the application server 300.
  • the sensor network unit 14 includes, for example, a plurality of sensor network adapters 11.
  • a plurality of actual sensors 12 are connected to each of the plurality of sensor network adapters 11, and each actual sensor 12 is connected to the Internet 15 via the sensor network adapter 11.
  • the real sensor 12 is configured to obtain sensing data by observing the target.
  • the actual sensor 12 includes, for example, an image sensor (camera), a temperature sensor, a humidity sensor, an illuminance sensor, a force sensor, a sound sensor, an RFID (Radio Frequency IDentification) sensor, an infrared sensor, a posture sensor, a rainfall sensor, a radioactivity sensor, and a gas sensor. Etc.
  • the actual sensor 12 does not necessarily need to be a fixed type, and may be a mobile type such as a mobile phone, a smartphone, and a tablet.
  • each real sensor 12 does not necessarily need to be composed of a single sensor, and may be composed of a plurality of sensors.
  • the actual sensor 12 may be installed for any purpose. For example, for factory FA (Factory Automation) and production management, urban traffic control, environmental measurement such as weather, health care, crime prevention, etc. It may be installed.
  • Each real sensor 12 is configured to generate sensing data by high-speed sampling, for example.
  • all the generated sensing data is temporarily buffered.
  • the buffered sensing data is deleted sequentially from the oldest one. That is, in each real sensor 12, sensing data sampled at a high speed in a constant period is always buffered.
  • Each real sensor 12 can transmit all of the buffered sensing data to the virtual sensor management server 100 or the like, or can transmit only a part of the buffered sensing data to the virtual sensor management server 100 or the like. You can also.
  • each sensor network adapter 11 is disposed at a separate (distant) location, and each actual sensor 12 connected to each sensor network adapter 11 is disposed at the same (near) location. These arrangement locations are not limited to this.
  • Each application server 300 (300A, 300B) is configured to execute an application that uses sensing data, and is realized by, for example, a general-purpose computer.
  • the application server 300 acquires necessary sensing data via the Internet 15.
  • the virtual sensor management server 100 is a server for realizing a virtual sensor.
  • a plurality of processing modules 110 and a sensor control module 120 are realized, and a metadata DB (database) 130 is managed.
  • Each of the plurality of processing modules 110 and the sensor control module 120 is, for example, a software module.
  • the processing module 110 includes a plurality of input ports (input channels) and is configured to detect the occurrence of an event based on sensing data input to each input port.
  • the processing module 110 can switch the actual sensor 12 that outputs sensing data to the input port as necessary. For example, when the actual sensor 12 that is currently outputting sensing data to the input port fails, the processing module 110 can switch the input sensor to another actual sensor 12.
  • the processing module 110 for example, based on sensing data output from the power sensor (indicating power consumption in the home) and sensing data output from the atmospheric pressure sensor (indicating indoor air pressure). It may be configured to determine whether or not the ventilation fan is used and output the determination result. In this case, the processing module 110 and the actual sensor 12 (power sensor and atmospheric pressure sensor) can realize a virtual sensor that detects the occurrence of the event of using the ventilation fan.
  • the sensor control module 120 is configured to control transmission of sensing data from each real sensor 12 included in the sensor network unit 14 to the processing module 110.
  • the metadata DB 130 is configured to manage metadata indicating the relationship between each real sensor 12 and each processing module 110. Details of the sensor control module 120 and the metadata DB 130 will be described later.
  • FIG. 3 is a diagram illustrating an example of a hardware configuration of the virtual sensor management server 100.
  • the virtual sensor management server 100 is realized by a general-purpose computer, for example.
  • the virtual sensor management server 100 includes a control unit 180, a communication I / F (interface) 195, and a storage unit 190, and each component is electrically connected via a bus 197. Yes.
  • the control unit 180 includes a CPU (Central Processing Unit) 182, a RAM (Random Access Memory) 184, a ROM (Read Only Memory) 186, and the like, and is configured to control each component according to information processing. .
  • CPU Central Processing Unit
  • RAM Random Access Memory
  • ROM Read Only Memory
  • the communication I / F 195 is configured to communicate with external devices (for example, the application server 300 and the sensor network unit 14 (FIG. 2)) provided outside the virtual sensor management server 100 via the Internet 15. .
  • the communication I / F 195 includes, for example, a wired LAN (Local Area Network) module or a wireless LAN module.
  • the storage unit 190 is an auxiliary storage device such as a hard disk drive or a solid state drive.
  • the storage unit 190 is configured to store, for example, the metadata DB 130 and the control program 191.
  • the control program 191 is a control program for the virtual sensor management server 100 that is executed by the control unit 180.
  • the processing module 110 and the sensor control module 120 may be realized by the control unit 180 executing the control program 191.
  • the control program 191 is expanded in the RAM 184.
  • the control unit 180 controls each component by interpreting and executing the control program 191 expanded in the RAM 184 by the CPU 182.
  • the metadata DB 130 will be described in detail later.
  • FIG. 4 is a diagram illustrating an example of a software configuration of the virtual sensor management server 100.
  • each of the processing module 110 and the sensor control module 120 is realized by the control unit 180.
  • the processing module 110 detects the occurrence of an event based on sensing data input from each real sensor 12 (12C, 12D, 12E).
  • the sensing data output by the actual sensor 12F is input to another processing module 110, for example.
  • the detection result of the actual sensor 12C satisfies the predetermined condition C1
  • the detection result of the actual sensor 12D satisfies the predetermined condition C2
  • the detection result of the actual sensor 12E is predetermined within a predetermined time.
  • the condition C3 is satisfied, the occurrence of an event is detected.
  • Each actual sensor 12 basically transmits a part of the sensing data buffered to the processing module 110. That is, sensing data having a sampling frequency lower than the sampling frequency at which each real sensor 12 generates sensing data is transmitted to the processing module 110. Thereby, the communication traffic between each real sensor 12 and the processing module 110 can be suppressed.
  • each actual sensor 12 always transmits only a part (small amount) of sensing data buffered to the processing module 110. In this case, depending on the sensing data transmitted from each actual sensor 12, the processing module 110 cannot detect the occurrence of the event even though the detection result of each actual sensor 12 satisfies the predetermined condition. Can occur.
  • FIG. 5 is a diagram for explaining a tentative example in which the processing module 110 cannot detect the occurrence of an event even though the detection result of each real sensor 12 satisfies a predetermined condition.
  • the horizontal axis represents time
  • the vertical axis represents the detection results of sensors S1, S3, S4 (actual sensor 12).
  • Sensors S1, S3, and S4 correspond to sensors S1, S3, and S4 in FIG. 7 described later, and correspond to actual sensors 12C, 12D, and 12E, respectively.
  • sensing data corresponding to each plot is transmitted from each sensor (actual sensor 12) to the processing module 110.
  • each real sensor 12 always transmits only a part (small amount) of sensing data buffered to the processing module 110.
  • the predetermined condition C1 is satisfied when the detection result of the sensor S1 falls below the predetermined value Th1.
  • the predetermined condition C2 is satisfied when the detection result of the sensor S3 exceeds the predetermined value Th3.
  • the predetermined condition C3 is satisfied when the detection result of the sensor S4 exceeds the predetermined value Th4.
  • the processing module 110 considers that an event has occurred when a predetermined condition C1, C2, C3 is satisfied within a predetermined time.
  • the processing module 110 does not detect the occurrence of an event. In this case, even within a predetermined time including before and after the time t1, the detection results of the sensors S1 and S3 do not satisfy the predetermined conditions C1 and C2, respectively.
  • the processing module 110 does not detect the occurrence of an event.
  • the detection results of the sensors S1 and S3 satisfy the predetermined conditions C1 and C2, respectively, within a predetermined time including before and after the time t2. That is, in this case, the processing module 110 should originally detect the occurrence of an event.
  • the sensor control module 120 performs processing for adjusting the amount of data transmitted from each actual sensor 12 to the processing module 110. Execute. Specifically, when there is a high possibility that an event has occurred, the sensor control module 120 transmits the sensing data having a larger amount of information (higher sampling frequency) than the normal sensor to the processing module 110. 12 is controlled.
  • FIG. 6 is a diagram illustrating an example of a change in the amount of data transmitted from each real sensor 12 to the processing module 110 in the first embodiment.
  • the meaning of each axis and each plot in the drawing is the same as in FIG.
  • sensor control module 120 when it is detected at time t12 that the detection result of sensor S4 exceeds the predetermined value Th4 (the possibility that an event has occurred) is high, sensor control module 120 (FIG. 4). Controls the sensors S1, S3 to send more sensing data than normal (eg, all buffered sensing data) to the processing module 110. In this case, the detection results of the sensors S1 and S3 do not satisfy the predetermined conditions C1 and C2, respectively, within a predetermined time (from time t11 to time t13) including before and after the time t12. Therefore, even if all the sensing data from time t11 to time t13 is transmitted from the sensors S1 and S3 to the processing module 110, the processing module 110 does not detect the occurrence of an event.
  • Th4 the possibility that an event has occurred
  • the sensor control module 120 receives more sensing data than usual (for example, all buffered sensing data).
  • the sensors S1 and S3 are controlled to transmit to the processing module 110.
  • the processing module 110 detects the occurrence of an event.
  • the sensor control module 120 when there is a high possibility that an event has occurred, a larger amount of sensing data than normal is transmitted from the actual sensor 12 to the processing module 110. An event that could not be detected in the example of 5 can be detected.
  • the sensor control module 120 includes a determination unit 122 and a sensor control unit 124.
  • the determination unit 122 determines whether the detection result (sensing data) from any of the actual sensors 12C, 12D, and 12E satisfies a predetermined condition. In the first embodiment, for example, the determination unit 122 considers that there is a high possibility that an event has occurred when any of the sensing data input to the processing module 110 satisfies a predetermined condition.
  • the determination unit 122 receives sensing data from each real sensor 12 (12C, 12D, 12E).
  • the determining unit 122 refers to the metadata DB 130 to determine whether any received sensing data satisfies a predetermined condition.
  • FIG. 7 is a diagram illustrating an example of the metadata DB 130.
  • the metadata DB 130 manages metadata indicating the relationship between each real sensor 12 and each processing module 110 (FIG. 2).
  • the metadata associated with each processing module 110 indicates, for example, the input sensor of each processing module 110 and a predetermined associated with each actual sensor 12. Indicates conditions. That is, each metadata associates a plurality of real sensors 12 that transmit sensing data to the common processing module 110.
  • the input sensors of the processing module M1 are sensors marked with “ ⁇ ” in the figure, and are sensors S1, S3, S4.
  • the processing module M1 when the detection result of the sensor S1 satisfies the predetermined condition C1, the detection result of the sensor S3 satisfies the predetermined condition C2, and the detection result of the sensor S4 satisfies the predetermined condition C3, the event E1 The occurrence of is detected. That is, the determination unit 122 (FIG. 4) can recognize a predetermined condition associated with each real sensor 12 by referring to the metadata DB 130.
  • the determination unit 122 refers to the metadata DB 130 to determine the actual condition determined to satisfy the predetermined condition. Another real sensor 12 that transmits sensing data to the processing module 110 common to the sensor 12 is extracted. The determination unit 122 notifies the sensor control unit 124 of information indicating the extracted actual sensor 12. The sensor control unit 124 controls each real sensor 12 extracted by the determination unit 122 so as to transmit sensing data having a larger amount of information (higher sampling frequency) than usual to the processing module 110. Thereby, only when there is a high possibility that an event has occurred, the amount of data transmitted from the actual sensor 12 to the processing module 110 can be increased.
  • FIG. 8 is a flowchart illustrating an example of the event detection operation. The processing shown in this flowchart is executed when the control unit 180 operates as the determination unit 122, the sensor control unit 124, and the processing module 110, for example, when sensing data is input to the processing module 110.
  • control unit 180 determines whether any sensing data (detection result) input to processing module 110 satisfies a predetermined condition (step S100). If it is determined that the predetermined condition is not satisfied (NO in step S100), the amount of data transmitted from the actual sensor 12 to the processing module 110 is not increased, and the process proceeds to step S130.
  • the control unit 180 refers to the metadata DB 130 and relates to the actual sensor 12 that is determined that the sensing data satisfies the predetermined condition.
  • the actual sensor 12 is specified (step S110). That is, the control unit 180 identifies the actual sensor 12 that has been determined that the sensing data satisfies the predetermined condition and the actual sensor 12 that transmits the sensing data to the common processing module 110.
  • the control unit 180 controls the real sensor 12 specified in step S110 so as to transmit sensing data having a larger amount of information than usual (a sampling frequency is high) (step S120). Thereafter, the control unit 180 executes processing for detecting the occurrence of an event based on the received sensing data having a larger amount of information than usual (step S130).
  • the detection result by any one of the plurality of actual sensors 12 has a predetermined condition indicating that there is a high possibility that an event has occurred. It is determined whether or not it is satisfied. Then, when it is determined that the detection result satisfies the predetermined condition, the sensor control device 120 transmits sensing data having a larger amount of information to the processing module 110 than when it is determined that the detection result does not satisfy the predetermined condition. As described above, the other actual sensors 12 that transmit the sensing data to the processing module 110 are controlled.
  • the sensor control device 120 when there is a high possibility that an event has occurred, the occurrence of the event is detected with high accuracy based on sensing data having a large amount of information, while the event has occurred. When the possibility is low, the amount of data transmitted from each actual sensor 12 to the processing module 110 is suppressed. Therefore, according to this sensor control device 120, it is possible to achieve both high accuracy of event detection and reduction of communication traffic.
  • the actual sensor 12 is an example of the “sensor” of the present invention
  • the sensor control device (module) 120 is an example of the “sensor control device” of the present invention
  • the processing module 110 is the “processing” of the present invention. It is an example of a “module”.
  • the determination unit 122 is an example of the “determination unit” in the present invention
  • the sensor control unit 124 is an example of the “sensor control unit” in the present invention.
  • the output of each actual sensor 12 is input to the processing module 110.
  • the output of each actual sensor 12 is input to a processing module 110 ⁇ / b> A different from the processing module 110.
  • the processing module 110A detects occurrence of an event based on time-series data (a large amount of data (for example, sensing data)) input from a plurality of input channels, and calculates a likelihood related to the occurrence of the event. It is configured.
  • time-series data a large amount of data (for example, sensing data)
  • FIG. 9 is a diagram for explaining an outline of the sensor control device 120A according to the second embodiment.
  • the processing module 110 ⁇ / b> A has a plurality of input ports (input channels), and sensing data output by the actual sensor 12 is input to each input port.
  • Each actual sensor 12 and the processing module 110A can communicate with each other via, for example, the Internet.
  • the processing module 110A is configured to detect the occurrence of an event based on a plurality of received sensing data, calculate the likelihood, and output the detection result.
  • Sensing data input to the processing module 110A is time-series data generated continuously in time.
  • the sensor control module 120A according to the second embodiment controls the amount of sensing data (input data) input to the processing module 110A.
  • FIG. 10 is a diagram illustrating an example of the sensor network system 10A according to the second embodiment.
  • the sensor network system 10A includes a virtual sensor management server 100A.
  • the virtual sensor management server 100A is a server for realizing a virtual sensor.
  • a plurality of processing modules 110A are realized.
  • the processing module 110A is, for example, a software module.
  • the processing module 110A is a learned model generated by using a learning method based on time-series data such as LSTM (Long Short-Term Memory) or RNN (Recurrent Neural Network).
  • LSTM Long Short-Term Memory
  • RNN Recurrent Neural Network
  • the processing module 110A has a plurality of input ports (input channels), and detects occurrence of an event based on time-series data (a large amount of data (for example, sensing data)) input from the plurality of input ports. The likelihood related to the occurrence of the event is calculated.
  • the processing module 110A can switch the actual sensor 12 that outputs sensing data to the input port as necessary. For example, when the actual sensor 12 that is currently outputting sensing data to the input port fails, the processing module 110A can switch the input sensor to another actual sensor 12.
  • FIG. 11 is a diagram illustrating an example of a hardware configuration of the virtual sensor management server 100A.
  • the virtual sensor management server 100A is realized by a general-purpose computer, for example.
  • the virtual sensor management server 100A includes a control unit 180A and a storage unit 190A.
  • the control unit 180A includes a CPU 182, a RAM 184, a ROM 186, and the like, and is configured to control each component according to information processing.
  • the storage unit 190A is an auxiliary storage device such as a hard disk drive or a solid state drive.
  • the storage unit 190A is configured to store a control program 191A, for example.
  • the control program 191A is a control program of the virtual sensor management server 100A executed by the control unit 180A.
  • each processing module 110A and the sensor control module 120A may be realized by the control unit 180A executing the control program 191A.
  • the control program 191A is expanded in the RAM 184.
  • the control unit 180 ⁇ / b> A controls each component by interpreting and executing the control program 191 expanded in the RAM 184 by the CPU 182.
  • FIG. 12 is a diagram illustrating an example of a software configuration of the virtual sensor management server 100A.
  • each of the processing module 110A and the sensor control module 120A is realized by the control unit 180A.
  • the processing module 110A is a learned model generated by using a learning method based on time series data such as LSTM and RNN, and is based on sensing data input from each real sensor 12 (12C, 12D, 12E). Then, the occurrence of the event is detected and the likelihood is calculated. Since the processing module 110A handles time-series data, for example, when a large amount of data is input, data that is significantly different from the timing data before and after the time-series can be regarded as noise. Therefore, the processing module 110A can suppress erroneous detection of an event caused by noise. Note that the sensing data output by the actual sensor 12F is input to, for example, another processing module 110A.
  • the sensor control module 120A includes a determination unit 122A and a sensor control unit 124.
  • the determination unit 122A determines whether or not a detection result (sensing data) by any of the actual sensors 12C, 12D, and 12E satisfies a predetermined condition.
  • the determination unit 122A considers that there is a high possibility that an event has occurred when any of the sensing data input to the processing module 110A satisfies a predetermined condition.
  • each predetermined condition in the second embodiment does not match the event detection condition.
  • the predetermined condition in the second embodiment is managed in the metadata DB 130.
  • the determination unit 122A refers to the metadata DB 130, and the processing module 110A that is common to the actual sensor 12 that is determined to satisfy the predetermined condition.
  • the other real sensor 12 that transmits the sensing data is extracted.
  • the determination unit 122A notifies the sensor control unit 124A of information indicating the extracted actual sensor 12.
  • the sensor control unit 124A controls each real sensor 12 extracted by the determination unit 122A so as to transmit sensing data having a larger amount of information (higher sampling frequency) than usual to the processing module 110A. Thereby, only when there is a high possibility that an event has occurred, the amount of data transmitted from the actual sensor 12 to the processing module 110A can be increased.
  • FIG. 13 is a diagram illustrating an example of an event detection operation according to the second embodiment. Referring to FIG. 13, the axes of the drawing, the meaning of each plot, and the like are the same as those in FIGS.
  • sensor control module 120A (FIG. 12). Controls the sensors S1, S3 to transmit more sensing data than normal (eg, all buffered sensing data) to the processing module 110A. Thereby, a large amount of data is input to the processing module 110A. The greater the amount of data input to the processing module 110A, the higher the possibility that the input data will contain noise.
  • the processing module 110A can regard data that is significantly different from timing data before and after in time series as noise. Therefore, according to the processing module 110A, even if a large amount of input data includes noise, it is possible to appropriately detect the noise and avoid erroneous detection of event occurrence.
  • the sensor control module 120A processes more sensing data (eg, all buffered sensing data) than usual. Sensors S1 and S3 are controlled to transmit to 110A.
  • the processing module 110A detects the occurrence of an event based on the output of each of the sensors S1, S3, S4 within a predetermined time including before and after the time t26 (for example, from the time t24 to the time t27). And the likelihood is calculated. That is, according to the sensor control module 120A according to the second embodiment, when there is a high possibility that an event has occurred, a larger amount of sensing data than normal is transmitted from the actual sensor 12 to the processing module 110A. Event can be detected automatically.
  • processing module 110A detects the occurrence of an event and calculates the likelihood based on time-series data input to each of a plurality of input ports. To do.
  • processing module 110A detects the occurrence of an event and calculates the likelihood based on time-series data input to each of a plurality of input ports. To do.
  • the sensor control module 120A according to the second embodiment when a large amount of data is input from the real sensor 12 to the processing module 110A in accordance with the control of the sensor control module 120A, it is greatly different from the timing data in the time series. Since data is regarded as noise, false detection caused by noise can be suppressed.
  • a plurality of processing modules 110 and 110A are realized by the control units 180 and 180A.
  • the processing modules 110 and 110A realized by the control units 180 and 180A are not necessarily plural.
  • the number of processing modules 110 and 110A realized by the control units 180 and 180A may be one.
  • the processes performed by the virtual sensor management servers 100 and 100A may be realized by a plurality of servers and the like.
  • Sensing data having a sampling frequency higher than normal is transmitted from the other actual sensors 12 to the processing modules 110 and 110A.
  • the contents executed due to any detection result satisfying the predetermined condition are not limited to the above contents.
  • sensing data having a larger quantization bit number than usual may be transmitted from the other actual sensor 12 to the processing modules 110 and 110A.
  • control units 180 and 180A can detect occurrence of an event with higher accuracy based on sensing data having a large number of quantization bits. Further, for example, when any detection result satisfies a predetermined condition, sensing data having a sampling frequency higher than normal and a quantization bit number larger than normal is transmitted from the other real sensors 12 to the processing modules 110 and 110A. May be sent to. That is, when any detection result satisfies a predetermined condition, sensing data having a larger amount of information than usual may be transmitted from the other actual sensor 12 to the processing modules 110 and 110A.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Small-Scale Networks (AREA)
  • Arrangements For Transmission Of Measured Signals (AREA)

Abstract

Provided are a sensor control device, a sensor control method, and a program with which it is possible to both increase the accuracy of event detection and reduce communication traffic. Each of at least two or more sensors included in a plurality of sensors transmits data to a processing module. The processing module detects the occurrence of an event on the basis of a plurality of data. The sensor control device is provided with a determination unit and a sensor control unit. The determination unit determines whether or not the result of detection by any of the sensors satisfies a prescribed condition. When it is determined that the detection result satisfies the prescribed condition, the sensor control unit controls the other sensors included in the two or more sensors so as to transmit data having a large amount of information to the processing module.

Description

センサ制御装置、センサ制御方法及びプログラムSensor control device, sensor control method, and program
　本発明は、センサ制御装置、センサ制御方法及びプログラムに関する。 The present invention relates to a sensor control device, a sensor control method, and a program.
　特開２０１７－１８２５９６号公報（特許文献１）は、データ収集分析プラットフォームを開示する。このデータ収集分析プラットフォームにおいては、ネットワークに接続された種々のセンサからセンシングデータが収集され、収集されたセンシングデータに基づいて分析が行なわれる（特許文献１参照）。 JP 2017-182596 A (Patent Document 1) discloses a data collection and analysis platform. In this data collection and analysis platform, sensing data is collected from various sensors connected to a network, and analysis is performed based on the collected sensing data (see Patent Document 1).
特開２０１７－１８２５９６号公報JP 2017-182596 A
　たとえば、上記特許文献１に開示されているような、ネットワークに接続された種々のセンサからセンシングデータを収集するシステムにおいて、収集されたセンシングデータに基づいて特定のイベントが発生したか否かを検知することが考えられる。たとえば、処理モジュールが、収集された複数のセンシングデータに基づいてイベントの発生を検知する。このような場合に、各センサが高速サンプリングを行ない、全てのセンシングデータが収集されると、処理モジュールは、イベントの発生検知を高精度に行なうことができる。 For example, in a system that collects sensing data from various sensors connected to a network as disclosed in Patent Document 1, it is detected whether a specific event has occurred based on the collected sensing data. It is possible to do. For example, the processing module detects the occurrence of an event based on a plurality of collected sensing data. In such a case, when each sensor performs high-speed sampling and all the sensing data is collected, the processing module can detect the occurrence of the event with high accuracy.
　しかしながら、各センサが高速サンプリングを行なった場合に、多量のセンシングデータが処理モジュールへ送信されると、各センサから処理モジュールへ送信されるデータ量（通信トラフィック）が大きくなる。一方、各センサから処理モジュールへ少量のセンシングデータしか送信されないと、イベント発生の検知精度が低下する。 However, when a large amount of sensing data is transmitted to the processing module when each sensor performs high-speed sampling, the amount of data (communication traffic) transmitted from each sensor to the processing module increases. On the other hand, if only a small amount of sensing data is transmitted from each sensor to the processing module, the detection accuracy of event occurrence decreases.
　本発明は、このような問題を解決するためになされたものであって、その目的は、イベント検知の高精度化、及び、通信トラフィックの低減を両立可能なセンサ制御装置、センサ制御方法及びプログラムを提供することである。 The present invention has been made to solve such a problem, and an object of the present invention is to provide a sensor control device, a sensor control method, and a program capable of achieving both high accuracy of event detection and reduction of communication traffic. Is to provide.
　本発明のある局面に従うセンサ制御装置は、複数のセンサの各々を制御する。複数のセンサに含まれる少なくとも２以上のセンサの各々は、処理モジュールへセンシングデータを送信する。処理モジュールは、受信された複数のセンシングデータに基づいてイベントの発生を検知する。センサ制御装置は、判定部と、センサ制御部とを備える。判定部は、上記２以上のセンサに含まれるいずれかのセンサによる検知結果が、上記イベントが発生している可能性が高いことを示す所定条件を満たすか否かを判定する。センサ制御部は、検知結果が所定条件を満たすと判定された場合に、検知結果が所定条件を満たさないと判定された場合よりも、情報量の大きいセンシングデータを処理モジュールへ送信するように、上記２以上のセンサに含まれる他のセンサを制御する。 A sensor control device according to an aspect of the present invention controls each of a plurality of sensors. Each of at least two or more sensors included in the plurality of sensors transmits sensing data to the processing module. The processing module detects the occurrence of an event based on the received plurality of sensing data. The sensor control device includes a determination unit and a sensor control unit. The determination unit determines whether or not a detection result by any one of the two or more sensors satisfies a predetermined condition indicating that the event is highly likely to occur. When it is determined that the detection result satisfies the predetermined condition, the sensor control unit transmits sensing data having a larger amount of information to the processing module than when the detection result is determined not to satisfy the predetermined condition. Control other sensors included in the two or more sensors.
　このセンサ制御装置によれば、処理モジュールへセンシングデータを送信する２以上のセンサのいずれかによって上記イベントが発生している可能性が高いことが検知された場合に、上記２以上のセンサに含まれる他のセンサによって、通常（検知結果が所定条件を満たさない場合）よりも情報量の大きいセンシングデータが処理モジュールへ送信される。すなわち、このセンサ制御装置によれば、上記イベントが発生している可能性が高い場合には、情報量の大きい（多くの）センシングデータに基づいてイベントの発生が高精度に検知される一方、上記イベントが発生している可能性が低い場合には、各センサから処理モジュールへ送信されるデータ量（情報量）が抑制される。したがって、このセンサ制御装置によれば、イベント検知の高精度化、及び、通信トラフィックの低減を両立することができる。 According to this sensor control device, when it is detected that one of the two or more sensors transmitting sensing data to the processing module is likely to cause the event, the sensor control device includes the two or more sensors. Sensing data having a larger amount of information than normal (when the detection result does not satisfy the predetermined condition) is transmitted to the processing module. That is, according to this sensor control device, when the possibility that the event has occurred is high, the occurrence of the event is detected with high accuracy based on sensing data having a large amount of information (many), When the possibility that the event has occurred is low, the amount of data (information amount) transmitted from each sensor to the processing module is suppressed. Therefore, according to this sensor control device, it is possible to achieve both high accuracy of event detection and reduction of communication traffic.
　また、上記センサ制御装置において、判定部は、複数のセンサから一部のセンサを抽出する。判定部によって抽出される一部のセンサは、検知結果が所定条件を満たすと判定されたセンサと、共通の処理モジュールへセンシングデータを送信する。センサ制御部は、検知結果が所定条件を満たすと判定された場合に、検知結果が所定条件を満たさないと判定された場合よりも、情報量の大きいセンシングデータを処理モジュールへ送信するように、判定部によって抽出されたセンサを制御してもよい。 Further, in the sensor control device, the determination unit extracts some sensors from a plurality of sensors. Some of the sensors extracted by the determination unit transmit sensing data to a sensor that is determined to have a detection result satisfying a predetermined condition and a common processing module. When it is determined that the detection result satisfies the predetermined condition, the sensor control unit transmits sensing data having a larger amount of information to the processing module than when the detection result is determined not to satisfy the predetermined condition. You may control the sensor extracted by the determination part.
　このセンサ制御装置によれば、検知結果が所定条件を満たすと判定されたセンサと、共通の処理モジュールへセンシングデータを送信する他のセンサが判定部によって抽出され、検知結果が所定条件を満たした場合に、判定部によって抽出されたセンサから処理モジュールへ通常よりも情報量の大きいセンシングデータが送信される。したがって、このセンサ制御装置によれば、検知結果が所定条件を満たした場合に、適切なセンサからのデータ送信量を増やすことができる。 According to this sensor control apparatus, the sensor determined that the detection result satisfies the predetermined condition and the other sensor that transmits the sensing data to the common processing module are extracted by the determination unit, and the detection result satisfies the predetermined condition In this case, sensing data having a larger amount of information than usual is transmitted from the sensor extracted by the determination unit to the processing module. Therefore, according to this sensor control device, when the detection result satisfies a predetermined condition, the data transmission amount from an appropriate sensor can be increased.
　また、上記センサ制御装置において、判定部は、各々が共通の処理モジュールへセンシングデータを送信する複数のセンサを互いに関連付けるメタデータを参照することによって、複数のセンサからセンサを抽出してもよい。 In the sensor control device, the determination unit may extract the sensors from the plurality of sensors by referring to metadata that associates the plurality of sensors that transmit the sensing data to the common processing module.
　このセンサ制御装置においては、メタデータを参照することによって、各々が共通の処理モジュールへセンシングデータを出力する複数のセンサが抽出される。すなわち、このセンサ制御装置によれば、メタデータを参照するだけで、各々が共通の処理モジュールへセンシングデータを送信する複数のセンサを抽出することができる。 In this sensor control device, a plurality of sensors each outputting sensing data to a common processing module are extracted by referring to the metadata. That is, according to this sensor control device, it is possible to extract a plurality of sensors each transmitting sensing data to a common processing module simply by referring to metadata.
　また、処理モジュールと、該処理モジュールへセンシングデータを送信するセンサとによって仮想センサが形成されてもよい。 Further, a virtual sensor may be formed by a processing module and a sensor that transmits sensing data to the processing module.
　また、処理モジュールは、複数の入力チャネルの各々に入力される時系列データに基づいて、イベントの発生を検知するとともに尤度を算出してもよい。 Further, the processing module may detect the occurrence of an event and calculate the likelihood based on time series data input to each of a plurality of input channels.
　このセンサ制御装置においては、処理モジュールが時系列データに基づいてイベントの発生等を検知する。したがって、このセンサ制御装置によれば、センサ制御装置の制御に従って処理モジュールへ実センサから大量のデータが入力された場合に、時系列における前後のタイミングのデータと大きく異なるデータがノイズとみなされるため、ノイズに起因する誤検知を抑制することができる。 In this sensor control device, the processing module detects the occurrence of an event based on time series data. Therefore, according to this sensor control device, when a large amount of data is input from the actual sensor to the processing module according to the control of the sensor control device, data that is significantly different from the timing data in the time series is regarded as noise. In addition, false detection caused by noise can be suppressed.
　本発明の別の局面に従うセンサ制御方法は、複数のセンサの各々を制御する。複数のセンサに含まれる少なくとも２以上のセンサの各々は、処理モジュールへセンシングデータを送信する。処理モジュールは、受信された複数のセンシングデータに基づいてイベントの発生を検知する。センサ制御方法は、上記２以上のセンサに含まれるいずれかのセンサによる検知結果が、上記イベントが発生している可能性が高いことを示す所定条件を満たすか否かを判定するステップと、検知結果が所定条件を満たすと判定された場合に、検知結果が所定条件を満たさないと判定された場合よりも、情報量の大きいセンシングデータを処理モジュールへ送信するように、上記２以上のセンサに含まれる他のセンサを制御するステップとを含む。 The sensor control method according to another aspect of the present invention controls each of a plurality of sensors. Each of at least two or more sensors included in the plurality of sensors transmits sensing data to the processing module. The processing module detects the occurrence of an event based on the received plurality of sensing data. The sensor control method includes a step of determining whether or not a detection result by any one of the two or more sensors satisfies a predetermined condition indicating that the event is highly likely to occur, When it is determined that the result satisfies the predetermined condition, the two or more sensors are configured to transmit sensing data having a larger amount of information to the processing module than when the detection result is determined not to satisfy the predetermined condition. Controlling other included sensors.
　このセンサ制御方法によれば、処理モジュールへセンシングデータを送信する２以上のセンサのいずれかによって上記イベントが発生している可能性が高いことが検知された場合に、上記２以上のセンサに含まれる他のセンサによって、通常（検知結果が所定条件を満たさない場合）よりも情報量の大きいセンシングデータが処理モジュールへ送信される。すなわち、このセンサ制御方法によれば、上記イベントが発生している可能性が高い場合には、情報量の大きいセンシングデータに基づいてイベントの発生が高精度に検知される一方、上記イベントが発生している可能性が低い場合には、各センサから処理モジュールへ送信されるデータ量が抑制される。したがって、このセンサ制御方法によれば、イベント検知の高精度化、及び、通信トラフィックの低減を両立することができる。 According to this sensor control method, when it is detected that one of the two or more sensors that transmit sensing data to the processing module is likely to cause the event, the sensor control method includes the two or more sensors. Sensing data having a larger amount of information than normal (when the detection result does not satisfy the predetermined condition) is transmitted to the processing module. That is, according to this sensor control method, when there is a high possibility that the event has occurred, the occurrence of the event is detected with high accuracy based on sensing data having a large amount of information. When there is a low possibility that data is transmitted, the amount of data transmitted from each sensor to the processing module is suppressed. Therefore, according to this sensor control method, both high accuracy of event detection and reduction of communication traffic can be achieved.
　本発明の別の局面に従うプログラムは、複数のセンサの各々を制御する処理をコンピュータに実行させるように構成されている。複数のセンサに含まれる少なくとも２以上のセンサの各々は、処理モジュールへセンシングデータを送信する。処理モジュールは、受信された複数のセンシングデータに基づいてイベントの発生を検知する。プログラムは、２以上のセンサに含まれるいずれかのセンサによる検知結果が、上記イベントが発生している可能性が高いことを示す所定条件を満たすか否かを判定するステップと、検知結果が所定条件を満たすと判定された場合に、検知結果が所定条件を満たさないと判定された場合よりも、情報量の大きいセンシングデータを処理モジュールへ送信するように、上記２以上のセンサに含まれる他のセンサを制御するステップとをコンピュータに実行させるように構成されている。 The program according to another aspect of the present invention is configured to cause a computer to execute processing for controlling each of a plurality of sensors. Each of at least two or more sensors included in the plurality of sensors transmits sensing data to the processing module. The processing module detects the occurrence of an event based on the received plurality of sensing data. The program determines whether or not a detection result of any one of the two or more sensors satisfies a predetermined condition indicating that the event is highly likely to occur, and the detection result is a predetermined value. Other than those included in the two or more sensors so that sensing data with a larger amount of information is transmitted to the processing module when it is determined that the condition is satisfied than when the detection result is determined not to satisfy the predetermined condition. The step of controlling the sensors is executed by a computer.
　このプログラムがコンピュータによって実行されると、処理モジュールへセンシングデータを送信する２以上のセンサのいずれかによって上記イベントが発生している可能性が高いことが検知された場合に、上記２以上のセンサに含まれる他のセンサによって、通常（検知結果が所定条件を満たさない場合）よりも情報量の大きいセンシングデータが処理モジュールへ送信される。すなわち、このプログラムによれば、上記イベントが発生している可能性が高い場合には、情報量の大きいセンシングデータに基づいてイベントの発生が高精度に検知される一方、上記イベントが発生している可能性が低い場合には、各センサから処理モジュールへ送信されるデータ量が抑制される。したがって、このプログラムによれば、イベント検知の高精度化、及び、通信トラフィックの低減を両立することができる。 When this program is executed by a computer, the two or more sensors are detected when it is detected by one of the two or more sensors that transmit sensing data to the processing module that the event is highly likely to occur. Sensing data having a larger amount of information than normal (when the detection result does not satisfy the predetermined condition) is transmitted to the processing module by the other sensors included in. That is, according to this program, when there is a high possibility that the event has occurred, the occurrence of the event is detected with high accuracy based on sensing data having a large amount of information. When the possibility of being present is low, the amount of data transmitted from each sensor to the processing module is suppressed. Therefore, according to this program, both high accuracy of event detection and reduction of communication traffic can be achieved.
　本発明によれば、イベント検知の高精度化、及び、通信トラフィックの低減を両立可能なセンサ制御装置、センサ制御方法及びプログラムを提供することができる。 According to the present invention, it is possible to provide a sensor control device, a sensor control method, and a program that can achieve both high accuracy of event detection and reduction of communication traffic.
実施の形態１における、センサ制御装置の概要を説明するための図である。FIG. 3 is a diagram for describing an overview of a sensor control device in the first embodiment. 実施の形態１における、センサネットワークシステムの一例を示す図である。1 is a diagram illustrating an example of a sensor network system in Embodiment 1. FIG. 実施の形態１における、仮想センサ管理サーバのハードウェア構成の一例を示す図である。FIG. 3 is a diagram illustrating an example of a hardware configuration of a virtual sensor management server in the first embodiment. 実施の形態１における、仮想センサ管理サーバのソフトウェア構成の一例を示す図である。3 is a diagram illustrating an example of a software configuration of a virtual sensor management server according to Embodiment 1. FIG. 各実センサの検知結果が所定条件を満たしているにも拘わらず、処理モジュールがイベントの発生を検知できない仮の例を説明するための図である。It is a figure for demonstrating the tentative example in which the processing module cannot detect generation | occurrence | production of an event, although the detection result of each real sensor satisfy | fills predetermined conditions. 実施の形態１における、各実センサから処理モジュールへ送信されるデータ量の変化の一例を示す図である。6 is a diagram illustrating an example of a change in the amount of data transmitted from each real sensor to a processing module in Embodiment 1. FIG. メタデータＤＢの一例を示す図である。It is a figure which shows an example of metadata DB. イベント検知動作の一例を示すフローチャートである。It is a flowchart which shows an example of event detection operation | movement. 実施の形態２における、センサ制御装置の概要を説明するための図である。FIG. 10 is a diagram for describing an overview of a sensor control device in a second embodiment. 実施の形態２における、センサネットワークシステムの一例を示す図である。It is a figure which shows an example of the sensor network system in Embodiment 2. FIG. 実施の形態２における、仮想センサ管理サーバのハードウェア構成の一例を示す図である。6 is a diagram illustrating an example of a hardware configuration of a virtual sensor management server in Embodiment 2. FIG. 実施の形態２における、仮想センサ管理サーバのソフトウェア構成の一例を示す図である。6 is a diagram illustrating an example of a software configuration of a virtual sensor management server in Embodiment 2. FIG. 実施の形態２における、イベント検知動作の一例を示す図である。FIG. 10 is a diagram illustrating an example of an event detection operation in the second embodiment.
　以下、本発明の一側面に係る実施の形態（以下、「本実施の形態」とも称する。）について、図面を用いて詳細に説明する。なお、図中同一又は相当部分には同一符号を付してその説明は繰り返さない。また、以下で説明する本実施の形態は、あらゆる点において本発明の例示にすぎない。本実施の形態は、本発明の範囲内において、種々の改良や変更が可能である。すなわち、本発明の実施にあたっては、実施の形態に応じて具体的構成を適宜採用することができる。 Hereinafter, an embodiment according to one aspect of the present invention (hereinafter, also referred to as “this embodiment”) will be described in detail with reference to the drawings. In the drawings, the same or corresponding portions are denoted by the same reference numerals and description thereof will not be repeated. In addition, the present embodiment described below is merely an example of the present invention in all respects. Various improvements and modifications can be made to the present embodiment within the scope of the present invention. That is, in carrying out the present invention, a specific configuration can be appropriately adopted according to the embodiment.
　［１．実施の形態１］
　＜１－１．概要＞
　図１は、本実施の形態１に従うセンサ制御装置（モジュール）１２０の概要を説明するための図である。図１に示されるように、処理モジュール１１０は複数の入力ポート（入力チャネル）を有し、各入力ポートには実センサ１２（センサの一例）によって出力されたセンシングデータが入力される。なお、各実センサ１２と、処理モジュール１１０とは、たとえば、インターネットを介して通信可能である。 [1. Embodiment 1]
<1-1. Overview>
FIG. 1 is a diagram for describing an overview of a sensor control device (module) 120 according to the first embodiment. As shown in FIG. 1, the processing module 110 has a plurality of input ports (input channels), and sensing data output by the actual sensor 12 (an example of a sensor) is input to each input port. In addition, each real sensor 12 and the processing module 110 can communicate via the internet, for example.
　処理モジュール１１０は、受信された複数のセンシングデータに基づいてイベントの発生を検知し、検知結果を出力するように構成されている。すなわち、処理モジュール１１０と、処理モジュール１１０へセンシングデータを出力する実センサ１２（入力センサ）とによって、いわゆる仮想センサが形成される。 The processing module 110 is configured to detect the occurrence of an event based on a plurality of received sensing data and output a detection result. That is, a so-called virtual sensor is formed by the processing module 110 and the real sensor 12 (input sensor) that outputs sensing data to the processing module 110.
　仮想センサとは、入力センサが対象を観測することによって生成されたセンシングデータに基づいて、入力センサによって観測された対象とは異なる対象の観測結果（たとえば、イベントの発生有無）をセンシングデータとして出力するセンサモジュールである。仮想センサについては、後程詳しく説明する。 A virtual sensor outputs sensing results (for example, whether an event has occurred) as sensing data based on sensing data generated by the input sensor observing the target, which is different from the target observed by the input sensor. It is a sensor module. The virtual sensor will be described in detail later.
　このような場合に、各実センサ１２がセンシングデータの高速サンプリングを行ない、全てのセンシングデータが処理モジュール１１０へ送信されると、処理モジュール１１０は、多数の（情報量の大きい）センシングデータに基づいて、イベントの発生検知を高精度に行なうことができる。 In such a case, when each real sensor 12 performs high-speed sampling of sensing data and all the sensing data is transmitted to the processing module 110, the processing module 110 is based on a large number of sensing data (a large amount of information). Thus, the occurrence of the event can be detected with high accuracy.
　しかしながら、各実センサ１２がセンシングデータの高速サンプリングを行なった場合に、多量（たとえば、全て）のセンシングデータが処理モジュール１１０へ送信されると、各実センサ１２から処理モジュール１１０へ送信されるデータ量（通信トラフィック）が大きくなる。一方、各実センサ１２から処理モジュール１１０へ少量のセンシングデータしか送信されないと、イベント発生の検知精度が低下する。 However, when each actual sensor 12 performs high-speed sampling of sensing data, if a large amount (for example, all) of sensing data is transmitted to the processing module 110, the data transmitted from each actual sensor 12 to the processing module 110. The amount (communication traffic) increases. On the other hand, if only a small amount of sensing data is transmitted from each real sensor 12 to the processing module 110, the detection accuracy of event occurrence decreases.
　本実施の形態１に従うセンサ制御装置１２０においては、たとえば、実センサ１２Ａ，１２Ｂのいずれかによる検知結果が、イベントが発生している可能性が高いことを示す所定条件を満たすか否かが判定される。そして、センサ制御装置１２０は、検知結果が所定条件を満たすと判定された場合に、検知結果が所定条件を満たさないと判定された場合よりも、情報量の大きいセンシングデータ（たとえば、サンプリング周波数が高い、及び／又は、量子化ビット数が大きい）を処理モジュール１１０へ送信するように、実センサ１２Ａ，１２Ｂのうちの他方の実センサを制御する。たとえば、実センサ１２Ａによる検知結果が所定条件を満たすと判定された場合に、センサ制御装置１２０は、通常よりも情報量の大きいセンシングデータを送信するように実センサ１２Ｂを制御する。 In sensor control device 120 according to the first embodiment, for example, it is determined whether or not the detection result by one of actual sensors 12A and 12B satisfies a predetermined condition indicating that there is a high possibility that an event has occurred. Is done. When the sensor control device 120 determines that the detection result satisfies the predetermined condition, the sensor control device 120 detects sensing data having a larger amount of information (for example, the sampling frequency is higher than that when the detection result is determined not to satisfy the predetermined condition). The other real sensor 12A, 12B is controlled so as to transmit a high (and / or a large number of quantization bits) to the processing module 110. For example, when it is determined that the detection result by the actual sensor 12A satisfies the predetermined condition, the sensor control device 120 controls the actual sensor 12B so as to transmit sensing data having a larger amount of information than usual.
　このセンサ制御装置１２０によれば、イベントが発生している可能性が高い場合には、情報量の大きいセンシングデータに基づいてイベントの発生が高精度に検知される一方、イベントが発生している可能性が低い場合には、各実センサ１２から処理モジュール１１０へ送信されるデータ量が抑制される。したがって、このセンサ制御装置１２０によれば、イベント検知の高精度化、及び、通信トラフィックの低減を両立することができる。 According to the sensor control device 120, when there is a high possibility that an event has occurred, the occurrence of the event is detected with high accuracy based on sensing data having a large amount of information, while the event has occurred. When the possibility is low, the amount of data transmitted from each actual sensor 12 to the processing module 110 is suppressed. Therefore, according to this sensor control device 120, it is possible to achieve both high accuracy of event detection and reduction of communication traffic.
　＜１－２．構成＞
　（１－２－１．システム全体の構成）
　図２は、本実施の形態１に従うセンサ制御モジュール（装置）１２０を含むセンサネットワークシステム１０の一例を示す図である。図２の例では、センサネットワークシステム１０は、センサネットワーク部１４と、仮想センサ管理サーバ１００と、アプリケーションサーバ３００とを含む。 <1-2. Configuration>
(1-2-1. Overall system configuration)
FIG. 2 is a diagram showing an example of the sensor network system 10 including the sensor control module (device) 120 according to the first embodiment. In the example of FIG. 2, the sensor network system 10 includes a sensor network unit 14, a virtual sensor management server 100, and an application server 300.
　センサネットワーク部１４、仮想センサ管理サーバ１００及びアプリケーションサーバ３００は、インターネット１５を介して相互に通信可能に接続されている。なお、センサネットワークシステム１０に含まれる各構成要素（仮想センサ管理サーバ１００、アプリケーションサーバ３００、センサネットワークアダプタ１１及び実センサ１２等）の数は、図２に示されるものに限定されない。 The sensor network unit 14, the virtual sensor management server 100, and the application server 300 are connected to each other via the Internet 15 so that they can communicate with each other. The number of each component (virtual sensor management server 100, application server 300, sensor network adapter 11, actual sensor 12, etc.) included in the sensor network system 10 is not limited to that shown in FIG.
　センサネットワークシステム１０においては、実センサ１２等によって生成されたセンシングデータが流通可能である。たとえば、実センサ１２によって生成されたセンシングデータは仮想センサ管理サーバ１００に流通し得るし、仮想センサによって生成されたセンシングデータはアプリケーションサーバ３００に流通し得る。 In the sensor network system 10, sensing data generated by the actual sensor 12 or the like can be distributed. For example, sensing data generated by the real sensor 12 can be distributed to the virtual sensor management server 100, and sensing data generated by the virtual sensor can be distributed to the application server 300.
　センサネットワーク部１４は、たとえば、複数のセンサネットワークアダプタ１１を含む。複数のセンサネットワークアダプタ１１の各々には複数の実センサ１２が接続されており、各実センサ１２はセンサネットワークアダプタ１１を介してインターネット１５に接続されている。 The sensor network unit 14 includes, for example, a plurality of sensor network adapters 11. A plurality of actual sensors 12 are connected to each of the plurality of sensor network adapters 11, and each actual sensor 12 is connected to the Internet 15 via the sensor network adapter 11.
　実センサ１２は、対象を観測することによってセンシングデータを得るように構成されている。実センサ１２は、たとえば、画像センサ（カメラ）、温度センサ、湿度センサ、照度センサ、力センサ、音センサ、ＲＦＩＤ（Radio Frequency IDentification）センサ、赤外線センサ、姿勢センサ、降雨センサ、放射能センサ及びガスセンサ等である。また、実センサ１２は、必ずしも固設型である必要はなく、携帯電話、スマートフォン及びタブレット等の移動型であってもよい。また、各実センサ１２は、必ずしも単一のセンサで構成されている必要はなく、複数のセンサによって構成されていてもよい。また、実センサ１２は、どのような目的で設置されていてもよく、たとえば、工場におけるＦＡ（Factory Automation）及び生産管理、都市交通制御、気象等の環境計測、ヘルスケア並びに防犯等のために設置されていてもよい。 The real sensor 12 is configured to obtain sensing data by observing the target. The actual sensor 12 includes, for example, an image sensor (camera), a temperature sensor, a humidity sensor, an illuminance sensor, a force sensor, a sound sensor, an RFID (Radio Frequency IDentification) sensor, an infrared sensor, a posture sensor, a rainfall sensor, a radioactivity sensor, and a gas sensor. Etc. Moreover, the actual sensor 12 does not necessarily need to be a fixed type, and may be a mobile type such as a mobile phone, a smartphone, and a tablet. Moreover, each real sensor 12 does not necessarily need to be composed of a single sensor, and may be composed of a plurality of sensors. The actual sensor 12 may be installed for any purpose. For example, for factory FA (Factory Automation) and production management, urban traffic control, environmental measurement such as weather, health care, crime prevention, etc. It may be installed.
　各実センサ１２は、たとえば、高速サンプリングによってセンシングデータを生成するように構成されている。各実センサ１２において、生成されたセンシングデータの全ては、一旦バッファされる。バッファされているセンシングデータは、古いものから順次削除される。すなわち、各実センサ１２においては、常時一定期間において高速サンプリングされたセンシングデータがバッファされている。各実センサ１２は、バッファされているセンシングデータの全てを仮想センサ管理サーバ１００等に送信することもできるし、バッファされているセンシングデータの一部のみを仮想センサ管理サーバ１００等に送信することもできる。 Each real sensor 12 is configured to generate sensing data by high-speed sampling, for example. In each real sensor 12, all the generated sensing data is temporarily buffered. The buffered sensing data is deleted sequentially from the oldest one. That is, in each real sensor 12, sensing data sampled at a high speed in a constant period is always buffered. Each real sensor 12 can transmit all of the buffered sensing data to the virtual sensor management server 100 or the like, or can transmit only a part of the buffered sensing data to the virtual sensor management server 100 or the like. You can also.
　センサネットワーク部１４において、たとえば、各センサネットワークアダプタ１１は別々の（遠い）場所に配置され、各センサネットワークアダプタ１１に接続される各実センサ１２は同一の（近い）場所に配置されるが、これらの配置場所はこれに限定されない。 In the sensor network unit 14, for example, each sensor network adapter 11 is disposed at a separate (distant) location, and each actual sensor 12 connected to each sensor network adapter 11 is disposed at the same (near) location. These arrangement locations are not limited to this.
　各アプリケーションサーバ３００（３００Ａ，３００Ｂ）は、センシングデータを利用するアプリケーションを実行するように構成されており、たとえば、汎用のコンピュータによって実現されている。アプリケーションサーバ３００は、インターネット１５を介して必要なセンシングデータを取得する。 Each application server 300 (300A, 300B) is configured to execute an application that uses sensing data, and is realized by, for example, a general-purpose computer. The application server 300 acquires necessary sensing data via the Internet 15.
　仮想センサ管理サーバ１００は、仮想センサを実現するためのサーバである。仮想センサ管理サーバ１００においては、複数の処理モジュール１１０と、センサ制御モジュール１２０とが実現されるとともに、メタデータＤＢ（database）１３０が管理される。複数の処理モジュール１１０及びセンサ制御モジュール１２０の各々は、たとえば、ソフトウェアモジュールである。 The virtual sensor management server 100 is a server for realizing a virtual sensor. In the virtual sensor management server 100, a plurality of processing modules 110 and a sensor control module 120 are realized, and a metadata DB (database) 130 is managed. Each of the plurality of processing modules 110 and the sensor control module 120 is, for example, a software module.
　処理モジュール１１０は、複数の入力ポート（入力チャネル）を含み、各入力ポートに入力されるセンシングデータに基づいて、イベントの発生を検知するように構成されている。処理モジュール１１０は、必要に応じて入力ポートへセンシングデータを出力する実センサ１２を切り替え可能である。たとえば、現在入力ポートにセンシングデータを出力している実センサ１２が故障した場合に、処理モジュール１１０は、入力センサを他の実センサ１２に切り替えることができる。 The processing module 110 includes a plurality of input ports (input channels) and is configured to detect the occurrence of an event based on sensing data input to each input port. The processing module 110 can switch the actual sensor 12 that outputs sensing data to the input port as necessary. For example, when the actual sensor 12 that is currently outputting sensing data to the input port fails, the processing module 110 can switch the input sensor to another actual sensor 12.
　処理モジュール１１０は、たとえば、電力センサによって出力されるセンシングデータ（家庭における電力使用量を示す。）と、気圧センサによって出力されるセンシングデータ（室内の気圧を示す。）とに基づいて、室内における換気扇の使用の有無を判定し、該判定結果を出力するように構成されてもよい。この場合には、処理モジュール１１０と、実センサ１２（電力センサ及び気圧センサ）とによって、換気扇の使用というイベントの発生を検知する仮想センサを実現することができる。 The processing module 110, for example, based on sensing data output from the power sensor (indicating power consumption in the home) and sensing data output from the atmospheric pressure sensor (indicating indoor air pressure). It may be configured to determine whether or not the ventilation fan is used and output the determination result. In this case, the processing module 110 and the actual sensor 12 (power sensor and atmospheric pressure sensor) can realize a virtual sensor that detects the occurrence of the event of using the ventilation fan.
　センサ制御モジュール１２０は、センサネットワーク部１４に含まれる各実センサ１２から処理モジュール１１０へのセンシングデータの送信を制御するように構成されている。メタデータＤＢ１３０は、各実センサ１２と、各処理モジュール１１０との関係を示すメタデータを管理するように構成されている。センサ制御モジュール１２０及びメタデータＤＢ１３０の詳細については後程説明する。 The sensor control module 120 is configured to control transmission of sensing data from each real sensor 12 included in the sensor network unit 14 to the processing module 110. The metadata DB 130 is configured to manage metadata indicating the relationship between each real sensor 12 and each processing module 110. Details of the sensor control module 120 and the metadata DB 130 will be described later.
　（１－２－２．仮想センサ管理サーバのハードウェア構成）
　図３は、仮想センサ管理サーバ１００のハードウェア構成の一例を示す図である。なお、本実施の形態１において、仮想センサ管理サーバ１００は、たとえば、汎用コンピュータによって実現される。 (1-2-2. Hardware Configuration of Virtual Sensor Management Server)
FIG. 3 is a diagram illustrating an example of a hardware configuration of the virtual sensor management server 100. In the first embodiment, the virtual sensor management server 100 is realized by a general-purpose computer, for example.
　図３の例では、仮想センサ管理サーバ１００は、制御部１８０と、通信Ｉ／Ｆ（interface）１９５と、記憶部１９０とを含み、各構成は、バス１９７を介して電気的に接続されている。 In the example of FIG. 3, the virtual sensor management server 100 includes a control unit 180, a communication I / F (interface) 195, and a storage unit 190, and each component is electrically connected via a bus 197. Yes.
　制御部１８０は、ＣＰＵ（Central Processing Unit）１８２、ＲＡＭ（Random Access Memory）１８４及びＲＯＭ（Read Only Memory）１８６等を含み、情報処理に応じて各構成要素の制御を行なうように構成されている。 The control unit 180 includes a CPU (Central Processing Unit) 182, a RAM (Random Access Memory) 184, a ROM (Read Only Memory) 186, and the like, and is configured to control each component according to information processing. .
　通信Ｉ／Ｆ１９５は、インターネット１５を介して、仮想センサ管理サーバ１００の外部に設けられた外部装置（たとえば、アプリケーションサーバ３００及びセンサネットワーク部１４（図２））と通信するように構成されている。通信Ｉ／Ｆ１９５は、たとえば、有線ＬＡＮ（Local Area Network）モジュールや無線ＬＡＮモジュールで構成される。 The communication I / F 195 is configured to communicate with external devices (for example, the application server 300 and the sensor network unit 14 (FIG. 2)) provided outside the virtual sensor management server 100 via the Internet 15. . The communication I / F 195 includes, for example, a wired LAN (Local Area Network) module or a wireless LAN module.
　記憶部１９０は、たとえば、ハードディスクドライブ、ソリッドステートドライブ等の補助記憶装置である。記憶部１９０は、たとえば、メタデータＤＢ１３０と、制御プログラム１９１とを記憶するように構成されている。 The storage unit 190 is an auxiliary storage device such as a hard disk drive or a solid state drive. The storage unit 190 is configured to store, for example, the metadata DB 130 and the control program 191.
　制御プログラム１９１は、制御部１８０によって実行される仮想センサ管理サーバ１００の制御プログラムである。たとえば、制御部１８０が制御プログラム１９１を実行することによって、各処理モジュール１１０及びセンサ制御モジュール１２０が実現されてもよい。制御部１８０が制御プログラム１９１を実行する場合に、制御プログラム１９１は、ＲＡＭ１８４に展開される。そして、制御部１８０は、ＲＡＭ１８４に展開された制御プログラム１９１をＣＰＵ１８２によって解釈及び実行することにより、各構成要素を制御する。なお、メタデータＤＢ１３０については、後程詳しく説明する。 The control program 191 is a control program for the virtual sensor management server 100 that is executed by the control unit 180. For example, the processing module 110 and the sensor control module 120 may be realized by the control unit 180 executing the control program 191. When the control unit 180 executes the control program 191, the control program 191 is expanded in the RAM 184. The control unit 180 controls each component by interpreting and executing the control program 191 expanded in the RAM 184 by the CPU 182. The metadata DB 130 will be described in detail later.
　（１－２－３．仮想センサ管理サーバのソフトウェア構成）
　図４は、仮想センサ管理サーバ１００のソフトウェア構成の一例を示す図である。図４の例では、処理モジュール１１０及びセンサ制御モジュール１２０の各々が制御部１８０によって実現される。 (1-2-3. Virtual Sensor Management Server Software Configuration)
FIG. 4 is a diagram illustrating an example of a software configuration of the virtual sensor management server 100. In the example of FIG. 4, each of the processing module 110 and the sensor control module 120 is realized by the control unit 180.
　上述のように、処理モジュール１１０は、各実センサ１２（１２Ｃ，１２Ｄ，１２Ｅ）から入力されるセンシングデータに基づいて、イベントの発生を検知する。なお、実センサ１２Ｆによって出力されるセンシングデータは、たとえば、別の処理モジュール１１０へ入力される。たとえば、処理モジュール１１０は、所定時間内に、実センサ１２Ｃの検知結果が所定条件Ｃ１を満たし、かつ、実センサ１２Ｄの検知結果が所定条件Ｃ２を満たし、かつ、実センサ１２Ｅの検知結果が所定条件Ｃ３を満たす場合に、イベントの発生を検知する。 As described above, the processing module 110 detects the occurrence of an event based on sensing data input from each real sensor 12 (12C, 12D, 12E). Note that the sensing data output by the actual sensor 12F is input to another processing module 110, for example. For example, in the processing module 110, the detection result of the actual sensor 12C satisfies the predetermined condition C1, the detection result of the actual sensor 12D satisfies the predetermined condition C2, and the detection result of the actual sensor 12E is predetermined within a predetermined time. When the condition C3 is satisfied, the occurrence of an event is detected.
　各実センサ１２は、基本的には、バッファしているセンシングデータのうち一部のセンシングデータを処理モジュール１１０へ送信する。すなわち、各実センサ１２がセンシングデータを生成するサンプリング周波数よりも低いサンプリング周波数のセンシングデータが処理モジュール１１０へ送信される。これにより、各実センサ１２と処理モジュール１１０との間の通信トラフィックを抑制することができる。一方、仮に、各実センサ１２が、常時、バッファしているセンシングデータのうち一部（少量）のセンシングデータしか処理モジュール１１０へ送信しないとする。この場合には、各実センサ１２から送信されるセンシングデータ次第で、各実センサ１２の検知結果が所定条件を満たしているにも拘わらず、処理モジュール１１０がイベントの発生を検知できないという事態が生じ得る。 Each actual sensor 12 basically transmits a part of the sensing data buffered to the processing module 110. That is, sensing data having a sampling frequency lower than the sampling frequency at which each real sensor 12 generates sensing data is transmitted to the processing module 110. Thereby, the communication traffic between each real sensor 12 and the processing module 110 can be suppressed. On the other hand, assume that each actual sensor 12 always transmits only a part (small amount) of sensing data buffered to the processing module 110. In this case, depending on the sensing data transmitted from each actual sensor 12, the processing module 110 cannot detect the occurrence of the event even though the detection result of each actual sensor 12 satisfies the predetermined condition. Can occur.
　図５は、各実センサ１２の検知結果が所定条件を満たしているにも拘わらず、処理モジュール１１０がイベントの発生を検知できない仮の例を説明するための図である。図５を参照して、横軸は時間を示し、縦軸はセンサＳ１，Ｓ３，Ｓ４（実センサ１２）の各々の検知結果を示す。なお、センサＳ１，Ｓ３，Ｓ４は、後述の図７におけるセンサＳ１，Ｓ３，Ｓ４にそれぞれ対応しており、それぞれ実センサ１２Ｃ，１２Ｄ，１２Ｅに対応する。 FIG. 5 is a diagram for explaining a tentative example in which the processing module 110 cannot detect the occurrence of an event even though the detection result of each real sensor 12 satisfies a predetermined condition. Referring to FIG. 5, the horizontal axis represents time, and the vertical axis represents the detection results of sensors S1, S3, S4 (actual sensor 12). Sensors S1, S3, and S4 correspond to sensors S1, S3, and S4 in FIG. 7 described later, and correspond to actual sensors 12C, 12D, and 12E, respectively.
　各センサの検知結果のうち、各プロット（たとえば、Ｄ１，Ｄ２）に対応するセンシングデータが各センサ（実センサ１２）から処理モジュール１１０へ送信される。この仮の例においては、各実センサ１２が、常時、バッファしているセンシングデータのうち一部（少量）のセンシングデータしか処理モジュール１１０へ送信しない。 Among the detection results of each sensor, sensing data corresponding to each plot (for example, D1, D2) is transmitted from each sensor (actual sensor 12) to the processing module 110. In this temporary example, each real sensor 12 always transmits only a part (small amount) of sensing data buffered to the processing module 110.
　なお、センサＳ１の検知結果が所定値Ｔｈ１を下回った場合に、所定条件Ｃ１が満たされる。センサＳ３の検知結果が所定値Ｔｈ３を上回った場合に、所定条件Ｃ２が満たされる。センサＳ４の検知結果が所定値Ｔｈ４を上回った場合に、所定条件Ｃ３が満たされる。たとえば、処理モジュール１１０は、所定時間内に所定条件Ｃ１，Ｃ２，Ｃ３が満たされた場合に、イベントが発生したものとみなす。 It should be noted that the predetermined condition C1 is satisfied when the detection result of the sensor S1 falls below the predetermined value Th1. The predetermined condition C2 is satisfied when the detection result of the sensor S3 exceeds the predetermined value Th3. The predetermined condition C3 is satisfied when the detection result of the sensor S4 exceeds the predetermined value Th4. For example, the processing module 110 considers that an event has occurred when a predetermined condition C1, C2, C3 is satisfied within a predetermined time.
　たとえば、時刻ｔ１においては、センサＳ４の検知結果は所定値Ｔｈ４を上回るものの、センサＳ１，Ｓ３の各々の検知結果は所定条件Ｃ１，Ｃ２をそれぞれ満たさない。したがって、処理モジュール１１０は、イベントの発生を検知しない。この場合には、時刻ｔ１の前後を含む所定時間内においても、センサＳ１，Ｓ３の各々の検知結果は所定条件Ｃ１，Ｃ２をそれぞれ満たさないため、処理モジュール１１０の検知結果は妥当だといえる。 For example, at time t1, the detection result of sensor S4 exceeds a predetermined value Th4, but the detection results of sensors S1 and S3 do not satisfy the predetermined conditions C1 and C2, respectively. Therefore, the processing module 110 does not detect the occurrence of an event. In this case, even within a predetermined time including before and after the time t1, the detection results of the sensors S1 and S3 do not satisfy the predetermined conditions C1 and C2, respectively.
　一方、時刻ｔ２においても、センサＳ４の検知結果は所定値Ｔｈ４を上回るものの、センサＳ１，Ｓ３の各々の検知結果は所定条件Ｃ１，Ｃ２をそれぞれ満たさない。したがって、処理モジュール１１０は、イベントの発生を検知しない。しかしながら、この場合には、時刻ｔ２の前後を含む所定時間内において、センサＳ１，Ｓ３の各々の検知結果は、所定条件Ｃ１，Ｃ２をそれぞれ満たす。すなわち、この場合に、本来、処理モジュール１１０は、イベントの発生を検知すべきである。 On the other hand, at time t2, the detection result of sensor S4 exceeds predetermined value Th4, but the detection results of sensors S1 and S3 do not satisfy predetermined conditions C1 and C2, respectively. Therefore, the processing module 110 does not detect the occurrence of an event. However, in this case, the detection results of the sensors S1 and S3 satisfy the predetermined conditions C1 and C2, respectively, within a predetermined time including before and after the time t2. That is, in this case, the processing module 110 should originally detect the occurrence of an event.
　再び図４を参照して、このような事態を回避するために、本実施の形態１において、センサ制御モジュール１２０は、各実センサ１２が処理モジュール１１０へ送信するデータ量を調整するための処理を実行する。具体的には、センサ制御モジュール１２０は、イベントが発生している可能性が高い場合に、通常よりも情報量の大きい（サンプリング周波数の高い）センシングデータを処理モジュール１１０へ送信するように実センサ１２を制御する。 Referring to FIG. 4 again, in order to avoid such a situation, in the first embodiment, the sensor control module 120 performs processing for adjusting the amount of data transmitted from each actual sensor 12 to the processing module 110. Execute. Specifically, when there is a high possibility that an event has occurred, the sensor control module 120 transmits the sensing data having a larger amount of information (higher sampling frequency) than the normal sensor to the processing module 110. 12 is controlled.
　図６は、本実施の形態１における、各実センサ１２から処理モジュール１１０へ送信されるデータ量の変化の一例を示す図である。図６を参照して、図面の各軸や各プロットの意味等は、図５と同様である。 FIG. 6 is a diagram illustrating an example of a change in the amount of data transmitted from each real sensor 12 to the processing module 110 in the first embodiment. With reference to FIG. 6, the meaning of each axis and each plot in the drawing is the same as in FIG.
　本実施の形態１においては、時刻ｔ１２において、センサＳ４の検知結果が所定値Ｔｈ４を上回る（イベントが発生している可能性が高い）ことが検知されると、センサ制御モジュール１２０（図４）は、通常よりも多くのセンシングデータ（たとえば、バッファされている全てのセンシングデータ）を処理モジュール１１０へ送信するようにセンサＳ１，Ｓ３を制御する。この場合には、時刻ｔ１２の前後を含む所定時間内（時刻ｔ１１から時刻ｔ１３）において、センサＳ１，Ｓ３の各々の検知結果は所定条件Ｃ１，Ｃ２をそれぞれ満たさない。したがって、時刻ｔ１１から時刻ｔ１３までのすべてのセンシングデータがセンサＳ１，Ｓ３から処理モジュール１１０へ送信されても、処理モジュール１１０は、イベントの発生を検知しない。 In the first embodiment, when it is detected at time t12 that the detection result of sensor S4 exceeds the predetermined value Th4 (the possibility that an event has occurred) is high, sensor control module 120 (FIG. 4). Controls the sensors S1, S3 to send more sensing data than normal (eg, all buffered sensing data) to the processing module 110. In this case, the detection results of the sensors S1 and S3 do not satisfy the predetermined conditions C1 and C2, respectively, within a predetermined time (from time t11 to time t13) including before and after the time t12. Therefore, even if all the sensing data from time t11 to time t13 is transmitted from the sensors S1 and S3 to the processing module 110, the processing module 110 does not detect the occurrence of an event.
　一方、時刻ｔ１６において、センサＳ４の検知結果が所定値Ｔｈ４を上回ることが検知されると、センサ制御モジュール１２０は、通常よりも多くのセンシングデータ（たとえば、バッファされている全てのセンシングデータ）を処理モジュール１１０へ送信するようにセンサＳ１，Ｓ３を制御する。この場合には、時刻ｔ１６の前後を含む所定時間内（たとえば、時刻ｔ１４から時刻ｔ１７）において、センサＳ１，Ｓ３の各々の検知結果は所定条件Ｃ１，Ｃ２をそれぞれ満たす（たとえば、時刻ｔ１５）。したがって、処理モジュール１１０は、イベントの発生を検知する。すなわち、本実施の形態１に従うセンサ制御モジュール１２０によれば、イベントが発生している可能性が高い場合に通常よりも多量のセンシングデータが実センサ１２から処理モジュール１１０へ送信されるため、図５の例では検知できなかったイベントを検知することができる。 On the other hand, when it is detected that the detection result of the sensor S4 exceeds the predetermined value Th4 at time t16, the sensor control module 120 receives more sensing data than usual (for example, all buffered sensing data). The sensors S1 and S3 are controlled to transmit to the processing module 110. In this case, within a predetermined time including before and after time t16 (for example, from time t14 to time t17), the detection results of sensors S1 and S3 satisfy predetermined conditions C1 and C2, respectively (for example, time t15). Therefore, the processing module 110 detects the occurrence of an event. That is, according to the sensor control module 120 according to the first embodiment, when there is a high possibility that an event has occurred, a larger amount of sensing data than normal is transmitted from the actual sensor 12 to the processing module 110. An event that could not be detected in the example of 5 can be detected.
　再び図４を参照して、具体的には、センサ制御モジュール１２０は、判定部１２２と、センサ制御部１２４とを含む。判定部１２２は、実センサ１２Ｃ，１２Ｄ，１２Ｅのいずれかによる検知結果（センシングデータ）が所定条件を満たすか否かを判定する。本実施の形態１において、判定部１２２は、たとえば、処理モジュール１１０へ入力されるセンシングデータのいずれかが所定条件が満たす場合にイベントが発生している可能性が高いものとみなしている。 Referring to FIG. 4 again, specifically, the sensor control module 120 includes a determination unit 122 and a sensor control unit 124. The determination unit 122 determines whether the detection result (sensing data) from any of the actual sensors 12C, 12D, and 12E satisfies a predetermined condition. In the first embodiment, for example, the determination unit 122 considers that there is a high possibility that an event has occurred when any of the sensing data input to the processing module 110 satisfies a predetermined condition.
　たとえば、判定部１２２は、各実センサ１２（１２Ｃ，１２Ｄ，１２Ｅ）からセンシングデータを受信する。判定部１２２は、メタデータＤＢ１３０を参照することによって、受信されたいずれかのセンシングデータが所定条件を満たすか否かを判定する。 For example, the determination unit 122 receives sensing data from each real sensor 12 (12C, 12D, 12E). The determining unit 122 refers to the metadata DB 130 to determine whether any received sensing data satisfies a predetermined condition.
　図７は、メタデータＤＢ１３０の一例を示す図である。上述のように、メタデータＤＢ１３０は、各実センサ１２と、各処理モジュール１１０（図２）との関係を示すメタデータを管理する。図７の例において、各処理モジュール１１０（たとえば、Ｍ１，Ｍ２）に対応付けられているメタデータは、たとえば、各処理モジュール１１０の入力センサを示すとともに、各実センサ１２に対応付けられた所定条件を示す。すなわち、各メタデータは、共通の処理モジュール１１０へセンシングデータを送信する複数の実センサ１２を互いに関連付ける。 FIG. 7 is a diagram illustrating an example of the metadata DB 130. As described above, the metadata DB 130 manages metadata indicating the relationship between each real sensor 12 and each processing module 110 (FIG. 2). In the example of FIG. 7, the metadata associated with each processing module 110 (for example, M1, M2) indicates, for example, the input sensor of each processing module 110 and a predetermined associated with each actual sensor 12. Indicates conditions. That is, each metadata associates a plurality of real sensors 12 that transmit sensing data to the common processing module 110.
　たとえば、処理モジュールＭ１の入力センサは、図中「〇」が付されているセンサであり、センサＳ１，Ｓ３，Ｓ４である。処理モジュールＭ１においては、センサＳ１の検知結果が所定条件Ｃ１を満たし、かつ、センサＳ３の検知結果が所定条件Ｃ２を満たし、かつ、センサＳ４の検知結果が所定条件Ｃ３を満たす場合に、イベントＥ１の発生が検知される。すなわち、判定部１２２（図４）は、メタデータＤＢ１３０を参照することによって、各実センサ１２に対応付けられた所定条件を認識することができる。 For example, the input sensors of the processing module M1 are sensors marked with “◯” in the figure, and are sensors S1, S3, S4. In the processing module M1, when the detection result of the sensor S1 satisfies the predetermined condition C1, the detection result of the sensor S3 satisfies the predetermined condition C2, and the detection result of the sensor S4 satisfies the predetermined condition C3, the event E1 The occurrence of is detected. That is, the determination unit 122 (FIG. 4) can recognize a predetermined condition associated with each real sensor 12 by referring to the metadata DB 130.
　再び図４を参照して、受信されたいずれかのセンシングデータが所定条件を満たすと判定されると、判定部１２２は、メタデータＤＢ１３０を参照することによって、所定条件を満たしたと判定された実センサ１２と共通の処理モジュール１１０へセンシングデータを送信する他の実センサ１２を抽出する。判定部１２２は、抽出された実センサ１２を示す情報をセンサ制御部１２４に通知する。センサ制御部１２４は、通常よりも情報量の大きい（サンプリング周波数の高い）センシングデータを処理モジュール１１０へ送信するように、判定部１２２によって抽出された各実センサ１２を制御する。これにより、イベントが生じている可能性が高い場合にのみ、実センサ１２から処理モジュール１１０へ送信されるデータ量を増やすことができる。 Referring to FIG. 4 again, when it is determined that any of the received sensing data satisfies the predetermined condition, the determination unit 122 refers to the metadata DB 130 to determine the actual condition determined to satisfy the predetermined condition. Another real sensor 12 that transmits sensing data to the processing module 110 common to the sensor 12 is extracted. The determination unit 122 notifies the sensor control unit 124 of information indicating the extracted actual sensor 12. The sensor control unit 124 controls each real sensor 12 extracted by the determination unit 122 so as to transmit sensing data having a larger amount of information (higher sampling frequency) than usual to the processing module 110. Thereby, only when there is a high possibility that an event has occurred, the amount of data transmitted from the actual sensor 12 to the processing module 110 can be increased.
　＜１－３．動作＞
　図８は、イベント検知動作の一例を示すフローチャートである。このフローチャートに示される処理は、たとえば、処理モジュール１１０へセンシングデータが入力されている場合に、制御部１８０が判定部１２２、センサ制御部１２４及び処理モジュール１１０として動作することによって実行される。 <1-3. Operation>
FIG. 8 is a flowchart illustrating an example of the event detection operation. The processing shown in this flowchart is executed when the control unit 180 operates as the determination unit 122, the sensor control unit 124, and the processing module 110, for example, when sensing data is input to the processing module 110.
　図８を参照して、制御部１８０は、処理モジュール１１０へ入力されるいずれかのセンシングデータ（検知結果）が所定条件を満たすか否かを判定する（ステップＳ１００）。所定条件を満たさないと判定されると（ステップＳ１００においてＮＯ）、実センサ１２から処理モジュール１１０へ送信されるデータ量が増やされることはなく、処理はステップＳ１３０へ移行する。 Referring to FIG. 8, control unit 180 determines whether any sensing data (detection result) input to processing module 110 satisfies a predetermined condition (step S100). If it is determined that the predetermined condition is not satisfied (NO in step S100), the amount of data transmitted from the actual sensor 12 to the processing module 110 is not increased, and the process proceeds to step S130.
　一方、所定条件を満たすと判定されると（ステップＳ１００においてＹＥＳ）、制御部１８０は、メタデータＤＢ１３０を参照することによって、センシングデータが所定条件を満たすと判定された実センサ１２と、関連する実センサ１２を特定する（ステップＳ１１０）。すなわち、制御部１８０は、センシングデータが所定条件を満たすと判定された実センサ１２と、共通の処理モジュール１１０へセンシングデータを送信する実センサ１２を特定する。 On the other hand, when it is determined that the predetermined condition is satisfied (YES in step S100), the control unit 180 refers to the metadata DB 130 and relates to the actual sensor 12 that is determined that the sensing data satisfies the predetermined condition. The actual sensor 12 is specified (step S110). That is, the control unit 180 identifies the actual sensor 12 that has been determined that the sensing data satisfies the predetermined condition and the actual sensor 12 that transmits the sensing data to the common processing module 110.
　制御部１８０は、通常よりも情報量の大きい（サンプリング周波数の高い）センシングデータを送信するように、ステップＳ１１０において特定された実センサ１２を制御する（ステップＳ１２０）。その後、制御部１８０は、受信された通常よりも情報量の大きいセンシングデータに基づいて、イベント発生を検知するための処理を実行する（ステップＳ１３０）。 The control unit 180 controls the real sensor 12 specified in step S110 so as to transmit sensing data having a larger amount of information than usual (a sampling frequency is high) (step S120). Thereafter, the control unit 180 executes processing for detecting the occurrence of an event based on the received sensing data having a larger amount of information than usual (step S130).
　＜１－４．特徴＞
　以上のように、本実施の形態１に従うセンサ制御装置（モジュール）１２０においては、複数の実センサ１２のいずれかによる検知結果が、イベントが発生している可能性が高いことを示す所定条件を満たすか否かが判定される。そして、センサ制御装置１２０は、検知結果が所定条件を満たすと判定された場合に、検知結果が所定条件を満たさないと判定された場合よりも、情報量の大きいセンシングデータを処理モジュール１１０へ送信するように、処理モジュール１１０へセンシングデータを送信する他の実センサ１２を制御する。 <1-4. Features>
As described above, in the sensor control device (module) 120 according to the first embodiment, the detection result by any one of the plurality of actual sensors 12 has a predetermined condition indicating that there is a high possibility that an event has occurred. It is determined whether or not it is satisfied. Then, when it is determined that the detection result satisfies the predetermined condition, the sensor control device 120 transmits sensing data having a larger amount of information to the processing module 110 than when it is determined that the detection result does not satisfy the predetermined condition. As described above, the other actual sensors 12 that transmit the sensing data to the processing module 110 are controlled.
　このセンサ制御装置１２０によれば、イベントが発生している可能性が高い場合には、情報量の大きいセンシングデータに基づいてイベントの発生が高精度に検知される一方、イベントが発生している可能性が低い場合には、各実センサ１２から処理モジュール１１０へ送信されるデータ量が抑制される。したがって、このセンサ制御装置１２０によれば、イベント検知の高精度化、及び、通信トラフィックの低減を両立することができる。 According to the sensor control device 120, when there is a high possibility that an event has occurred, the occurrence of the event is detected with high accuracy based on sensing data having a large amount of information, while the event has occurred. When the possibility is low, the amount of data transmitted from each actual sensor 12 to the processing module 110 is suppressed. Therefore, according to this sensor control device 120, it is possible to achieve both high accuracy of event detection and reduction of communication traffic.
　なお、実センサ１２は、本発明の「センサ」の一例であり、センサ制御装置（モジュール）１２０は、本発明の「センサ制御装置」の一例であり、処理モジュール１１０は、本発明の「処理モジュール」の一例である。判定部１２２は、本発明の「判定部」の一例であり、センサ制御部１２４は、本発明の「センサ制御部」の一例である。 The actual sensor 12 is an example of the “sensor” of the present invention, the sensor control device (module) 120 is an example of the “sensor control device” of the present invention, and the processing module 110 is the “processing” of the present invention. It is an example of a “module”. The determination unit 122 is an example of the “determination unit” in the present invention, and the sensor control unit 124 is an example of the “sensor control unit” in the present invention.
　［２．実施の形態２］
　＜２－１．概要＞
　上記実施の形態１においては、各実センサ１２の出力が処理モジュール１１０へ入力された。本実施の形態２においては、各実センサ１２の出力が処理モジュール１１０とは異なる処理モジュール１１０Ａへ入力される。処理モジュール１１０Ａは、複数の入力チャネルから入力される時系列データ（大量のデータ（たとえば、センシングデータ））に基づいて、イベントの発生を検知するとともに、イベントの発生に関する尤度を算出するように構成されている。ここでは、上記実施の形態１と異なる部分を中心に説明し、共通部分については説明を繰り返さない。 [2. Second Embodiment]
<2-1. Overview>
In the first embodiment, the output of each actual sensor 12 is input to the processing module 110. In the second embodiment, the output of each actual sensor 12 is input to a processing module 110 </ b> A different from the processing module 110. The processing module 110A detects occurrence of an event based on time-series data (a large amount of data (for example, sensing data)) input from a plurality of input channels, and calculates a likelihood related to the occurrence of the event. It is configured. Here, the description will focus on the parts different from the first embodiment, and the description of the common parts will not be repeated.
　図９は、本実施の形態２における、センサ制御装置１２０Ａの概要を説明するための図である。図９に示されるように、処理モジュール１１０Ａは、複数の入力ポート（入力チャネル）を有し、各入力ポートには実センサ１２によって出力されたセンシングデータが入力される。なお、各実センサ１２と処理モジュール１１０Ａとは、たとえば、インターネットを介して通信可能である。 FIG. 9 is a diagram for explaining an outline of the sensor control device 120A according to the second embodiment. As shown in FIG. 9, the processing module 110 </ b> A has a plurality of input ports (input channels), and sensing data output by the actual sensor 12 is input to each input port. Each actual sensor 12 and the processing module 110A can communicate with each other via, for example, the Internet.
　処理モジュール１１０Ａは、受信された複数のセンシングデータに基づいてイベントの発生を検知するとともに尤度を算出し、検知結果を出力するように構成されている。処理モジュール１１０Ａに入力されるセンシングデータは、時間的に連続して生成される時系列データである。本実施の形態２に従うセンサ制御モジュール１２０Ａは、処理モジュール１１０Ａへ入力されるセンシングデータ（入力データ）の情報量を制御する。 The processing module 110A is configured to detect the occurrence of an event based on a plurality of received sensing data, calculate the likelihood, and output the detection result. Sensing data input to the processing module 110A is time-series data generated continuously in time. The sensor control module 120A according to the second embodiment controls the amount of sensing data (input data) input to the processing module 110A.
　＜２－２．構成＞
　（２－２－１．システム全体の構成）
　図１０は、本実施の形態２における、センサネットワークシステム１０Ａの一例を示す図である。図１０の例では、センサネットワークシステム１０Ａは、仮想センサ管理サーバ１００Ａを含む。仮想センサ管理サーバ１００Ａは、仮想センサを実現するためのサーバである。仮想センサ管理サーバ１００Ａにおいては、複数の処理モジュール１１０Ａが実現されている。処理モジュール１１０Ａは、たとえば、ソフトウェアモジュールである。 <2-2. Configuration>
(2-2-1. Overall system configuration)
FIG. 10 is a diagram illustrating an example of the sensor network system 10A according to the second embodiment. In the example of FIG. 10, the sensor network system 10A includes a virtual sensor management server 100A. The virtual sensor management server 100A is a server for realizing a virtual sensor. In the virtual sensor management server 100A, a plurality of processing modules 110A are realized. The processing module 110A is, for example, a software module.
　処理モジュール１１０Ａは、たとえば、ＬＳＴＭ（Long Short-Term Memory）やＲＮＮ（Recurrent Neural Network）といった、時系列データに基づく学習手法を用いることによって生成された学習済みモデルである。 The processing module 110A is a learned model generated by using a learning method based on time-series data such as LSTM (Long Short-Term Memory) or RNN (Recurrent Neural Network).
　処理モジュール１１０Ａは、複数の入力ポート（入力チャネル）を有し、複数の入力ポートから入力される時系列データ（大量のデータ（たとえば、センシングデータ））に基づいて、イベントの発生を検知するとともに、イベントの発生に関する尤度を算出するように構成されている。処理モジュール１１０Ａは、必要に応じて入力ポートへセンシングデータを出力する実センサ１２を切り替え可能である。たとえば、現在入力ポートにセンシングデータを出力している実センサ１２が故障した場合に、処理モジュール１１０Ａは、入力センサを他の実センサ１２に切り替えることができる。 The processing module 110A has a plurality of input ports (input channels), and detects occurrence of an event based on time-series data (a large amount of data (for example, sensing data)) input from the plurality of input ports. The likelihood related to the occurrence of the event is calculated. The processing module 110A can switch the actual sensor 12 that outputs sensing data to the input port as necessary. For example, when the actual sensor 12 that is currently outputting sensing data to the input port fails, the processing module 110A can switch the input sensor to another actual sensor 12.
　（２－２－２．仮想センサ管理サーバのハードウェア構成）
　図１１は、仮想センサ管理サーバ１００Ａのハードウェア構成の一例を示す図である。なお、本実施の形態２において、仮想センサ管理サーバ１００Ａは、たとえば、汎用コンピュータによって実現される。 (2-2-2. Hardware configuration of virtual sensor management server)
FIG. 11 is a diagram illustrating an example of a hardware configuration of the virtual sensor management server 100A. In the second embodiment, the virtual sensor management server 100A is realized by a general-purpose computer, for example.
　図１１の例では、仮想センサ管理サーバ１００Ａは、制御部１８０Ａと、記憶部１９０Ａとを含む。制御部１８０Ａは、ＣＰＵ１８２、ＲＡＭ１８４及びＲＯＭ１８６等を含み、情報処理に応じて各構成要素の制御を行なうように構成されている。記憶部１９０Ａは、たとえば、ハードディスクドライブ、ソリッドステートドライブ等の補助記憶装置である。記憶部１９０Ａは、たとえば、制御プログラム１９１Ａを記憶するように構成されている。 In the example of FIG. 11, the virtual sensor management server 100A includes a control unit 180A and a storage unit 190A. The control unit 180A includes a CPU 182, a RAM 184, a ROM 186, and the like, and is configured to control each component according to information processing. The storage unit 190A is an auxiliary storage device such as a hard disk drive or a solid state drive. The storage unit 190A is configured to store a control program 191A, for example.
　制御プログラム１９１Ａは、制御部１８０Ａによって実行される仮想センサ管理サーバ１００Ａの制御プログラムである。たとえば、制御部１８０Ａが制御プログラム１９１Ａを実行することによって、各処理モジュール１１０Ａ及びセンサ制御モジュール１２０Ａが実現されてもよい。制御部１８０Ａが制御プログラム１９１Ａを実行する場合に、制御プログラム１９１Ａは、ＲＡＭ１８４に展開される。そして、制御部１８０Ａは、ＲＡＭ１８４に展開された制御プログラム１９１をＣＰＵ１８２によって解釈及び実行することにより、各構成要素を制御する。 The control program 191A is a control program of the virtual sensor management server 100A executed by the control unit 180A. For example, each processing module 110A and the sensor control module 120A may be realized by the control unit 180A executing the control program 191A. When the control unit 180A executes the control program 191A, the control program 191A is expanded in the RAM 184. The control unit 180 </ b> A controls each component by interpreting and executing the control program 191 expanded in the RAM 184 by the CPU 182.
　（２－２－３．仮想センサ管理サーバのソフトウェア構成）
　図１２は、仮想センサ管理サーバ１００Ａのソフトウェア構成の一例を示す図である。図１２の例では、処理モジュール１１０Ａ及びセンサ制御モジュール１２０Ａの各々が制御部１８０Ａによって実現される。 (2-2-3. Virtual Sensor Management Server Software Configuration)
FIG. 12 is a diagram illustrating an example of a software configuration of the virtual sensor management server 100A. In the example of FIG. 12, each of the processing module 110A and the sensor control module 120A is realized by the control unit 180A.
　処理モジュール１１０Ａは、ＬＳＴＭやＲＮＮといった、時系列データに基づく学習手法を用いることによって生成された学習済みモデルであって、各実センサ１２（１２Ｃ，１２Ｄ，１２Ｅ）から入力されるセンシングデータに基づいて、イベントの発生を検知するとともに尤度を算出する。処理モジュール１１０Ａは、時系列データを扱うため、たとえば、大量のデータが入力された場合に、時系列における前後のタイミングのデータと大きく異なるデータをノイズとみなすことができる。したがって、処理モジュール１１０Ａは、ノイズに起因するイベントの誤検知を抑制することができる。なお、実センサ１２Ｆによって出力されるセンシングデータは、たとえば、別の処理モジュール１１０Ａへ入力される。 The processing module 110A is a learned model generated by using a learning method based on time series data such as LSTM and RNN, and is based on sensing data input from each real sensor 12 (12C, 12D, 12E). Then, the occurrence of the event is detected and the likelihood is calculated. Since the processing module 110A handles time-series data, for example, when a large amount of data is input, data that is significantly different from the timing data before and after the time-series can be regarded as noise. Therefore, the processing module 110A can suppress erroneous detection of an event caused by noise. Note that the sensing data output by the actual sensor 12F is input to, for example, another processing module 110A.
　センサ制御モジュール１２０Ａは、判定部１２２Ａと、センサ制御部１２４とを含む。判定部１２２Ａは、実センサ１２Ｃ，１２Ｄ，１２Ｅのいずれかによる検知結果（センシングデータ）が所定条件を満たすか否かを判定する。本実施の形態２において、判定部１２２Ａは、たとえば、処理モジュール１１０Ａへ入力されるセンシングデータのいずれかが所定条件が満たす場合にイベントが発生している可能性が高いものとみなしている。なお、実施の形態1とは異なり、本実施の形態２における各所定条件は、イベント検知の条件とは一致していない。本実施の形態２における所定条件は、メタデータＤＢ１３０において管理されている。 The sensor control module 120A includes a determination unit 122A and a sensor control unit 124. The determination unit 122A determines whether or not a detection result (sensing data) by any of the actual sensors 12C, 12D, and 12E satisfies a predetermined condition. In the second embodiment, for example, the determination unit 122A considers that there is a high possibility that an event has occurred when any of the sensing data input to the processing module 110A satisfies a predetermined condition. Unlike the first embodiment, each predetermined condition in the second embodiment does not match the event detection condition. The predetermined condition in the second embodiment is managed in the metadata DB 130.
　受信されたいずれかのセンシングデータが所定条件を満たすと判定されると、判定部１２２Ａは、メタデータＤＢ１３０を参照することによって、所定条件を満たしたと判定された実センサ１２と共通の処理モジュール１１０Ａへセンシングデータを送信する他の実センサ１２を抽出する。判定部１２２Ａは、抽出された実センサ１２を示す情報をセンサ制御部１２４Ａに通知する。センサ制御部１２４Ａは、通常よりも情報量の大きい（サンプリング周波数の高い）センシングデータを処理モジュール１１０Ａへ送信するように、判定部１２２Ａによって抽出された各実センサ１２を制御する。これにより、イベントが生じている可能性が高い場合にのみ、実センサ１２から処理モジュール１１０Ａへ送信されるデータ量を増やすことができる。 If it is determined that any one of the received sensing data satisfies the predetermined condition, the determination unit 122A refers to the metadata DB 130, and the processing module 110A that is common to the actual sensor 12 that is determined to satisfy the predetermined condition. The other real sensor 12 that transmits the sensing data is extracted. The determination unit 122A notifies the sensor control unit 124A of information indicating the extracted actual sensor 12. The sensor control unit 124A controls each real sensor 12 extracted by the determination unit 122A so as to transmit sensing data having a larger amount of information (higher sampling frequency) than usual to the processing module 110A. Thereby, only when there is a high possibility that an event has occurred, the amount of data transmitted from the actual sensor 12 to the processing module 110A can be increased.
　＜２－３．動作＞
　図１３は、本実施の形態２における、イベント検知動作の一例を示す図である。図１３を参照して、図面の各軸や各プロットの意味等は、図５，６と同様である。 <2-3. Operation>
FIG. 13 is a diagram illustrating an example of an event detection operation according to the second embodiment. Referring to FIG. 13, the axes of the drawing, the meaning of each plot, and the like are the same as those in FIGS.
　本実施の形態２においては、時刻ｔ２２において、センサＳ４の検知結果が所定値Ｔｈ１４を上回る（イベントが発生している可能性が高い）ことが検知されると、センサ制御モジュール１２０Ａ（図１２）は、通常よりも多くのセンシングデータ（たとえば、バッファされている全てのセンシングデータ）を処理モジュール１１０Ａへ送信するようにセンサＳ１，Ｓ３を制御する。これにより、大量のデータが処理モジュール１１０Ａへ入力される。処理モジュール１１０Ａへ入力されるデータのデータ量が多ければ多い程、入力データにノイズが含まれる可能性が高くなる。 In the second embodiment, when it is detected at time t22 that the detection result of sensor S4 exceeds a predetermined value Th14 (the possibility that an event has occurred) is high, sensor control module 120A (FIG. 12). Controls the sensors S1, S3 to transmit more sensing data than normal (eg, all buffered sensing data) to the processing module 110A. Thereby, a large amount of data is input to the processing module 110A. The greater the amount of data input to the processing module 110A, the higher the possibility that the input data will contain noise.
　しかしながら、上述のように、処理モジュール１１０Ａは、大量のデータが入力された場合に、時系列における前後のタイミングのデータと大きく異なるデータをノイズとみなすことができる。したがって、処理モジュール１１０Ａによれば、大量の入力データにノイズが含まれていたとしても、該ノイズを適切に検知し、イベント発生の誤検知を回避することができる。 However, as described above, when a large amount of data is input, the processing module 110A can regard data that is significantly different from timing data before and after in time series as noise. Therefore, according to the processing module 110A, even if a large amount of input data includes noise, it is possible to appropriately detect the noise and avoid erroneous detection of event occurrence.
　時刻ｔ２６において、センサＳ４の検知結果が所定値Ｔｈ１４を上回ることが検知されると、センサ制御モジュール１２０Ａは、通常よりも多くのセンシングデータ（たとえば、バッファされている全てのセンシングデータ）を処理モジュール１１０Ａへ送信するようにセンサＳ１，Ｓ３を制御する。この場合には、たとえば、時刻ｔ２６の前後を含む所定時間内（たとえば、時刻ｔ２４から時刻ｔ２７）におけるセンサＳ１，Ｓ３，Ｓ４の各々の出力に基づいて、処理モジュール１１０Ａは、イベントの発生を検知するとともに尤度を算出する。すなわち、本実施の形態２に従うセンサ制御モジュール１２０Ａによれば、イベントが発生している可能性が高い場合に通常よりも多量のセンシングデータが実センサ１２から処理モジュール１１０Ａへ送信されるため、効率的にイベントを検知することができる。 When it is detected that the detection result of the sensor S4 exceeds the predetermined value Th14 at time t26, the sensor control module 120A processes more sensing data (eg, all buffered sensing data) than usual. Sensors S1 and S3 are controlled to transmit to 110A. In this case, for example, the processing module 110A detects the occurrence of an event based on the output of each of the sensors S1, S3, S4 within a predetermined time including before and after the time t26 (for example, from the time t24 to the time t27). And the likelihood is calculated. That is, according to the sensor control module 120A according to the second embodiment, when there is a high possibility that an event has occurred, a larger amount of sensing data than normal is transmitted from the actual sensor 12 to the processing module 110A. Event can be detected automatically.
　＜２－４．特徴＞
　以上のように、本実施の形態２に従うセンサ制御モジュール１２０Ａにおいて、処理モジュール１１０Ａは、複数の入力ポートの各々に入力される時系列データに基づいて、イベントの発生を検知するとともに尤度を算出する。本実施の形態２に従うセンサ制御モジュール１２０Ａによれば、センサ制御モジュール１２０Ａの制御に従って処理モジュール１１０Ａへ実センサ１２から大量のデータが入力された場合に、時系列における前後のタイミングのデータと大きく異なるデータがノイズとみなされるため、ノイズに起因する誤検知を抑制することができる。 <2-4. Features>
As described above, in sensor control module 120A according to the second embodiment, processing module 110A detects the occurrence of an event and calculates the likelihood based on time-series data input to each of a plurality of input ports. To do. According to the sensor control module 120A according to the second embodiment, when a large amount of data is input from the real sensor 12 to the processing module 110A in accordance with the control of the sensor control module 120A, it is greatly different from the timing data in the time series. Since data is regarded as noise, false detection caused by noise can be suppressed.
　［３．変形例］
　以上、実施の形態１，２について説明したが、本発明は、上記実施の形態１，２に限定されるものではなく、その趣旨を逸脱しない限りにおいて、種々の変更が可能である。以下、変形例について説明する。但し、以下の変形例は適宜組合せ可能である。 [3. Modified example]
Although the first and second embodiments have been described above, the present invention is not limited to the first and second embodiments, and various modifications can be made without departing from the gist thereof. Hereinafter, modified examples will be described. However, the following modifications can be combined as appropriate.
　＜３－１＞
　上記実施の形態１，２においては、制御部１８０，１８０Ａによって複数の処理モジュール１１０，１１０Ａが実現された。しかしながら、制御部１８０，１８０Ａによって実現される処理モジュール１１０，１１０Ａは、必ずしも複数である必要はない。制御部１８０，１８０Ａによって実現される処理モジュール１１０，１１０Ａは、１つであってもよい。 <3-1>
In the first and second embodiments, a plurality of processing modules 110 and 110A are realized by the control units 180 and 180A. However, the processing modules 110 and 110A realized by the control units 180 and 180A are not necessarily plural. The number of processing modules 110 and 110A realized by the control units 180 and 180A may be one.
　＜３－２＞
　また、上記実施の形態１，２において、仮想センサ管理サーバ１００，１００Ａによって行なわれた処理は、複数のサーバ等によって実現されてもよい。 <3-2>
In the first and second embodiments, the processes performed by the virtual sensor management servers 100 and 100A may be realized by a plurality of servers and the like.
　＜３－３＞
　また、上記実施の形態１，２においては、共通の処理モジュール１１０，１１０Ａへセンシングデータを出力する複数の実センサ１２のうちいずれかの実センサ１２による検知結果が所定条件を満たした場合に、通常よりも高いサンプリング周波数のセンシングデータが他の実センサ１２から処理モジュール１１０，１１０Ａへ送信された。しかしながら、いずれかの検知結果が所定条件を満たすことに起因して実行される内容は、上記内容に限られない。たとえば、いずれかの検知結果が所定条件を満たした場合に、通常よりも大きい量子化ビット数のセンシングデータが他の実センサ１２から処理モジュール１１０，１１０Ａへ送信されてもよい。この場合であっても、制御部１８０，１８０Ａは、大きい量子化ビット数のセンシングデータに基づいて、イベントの発生検知をより高精度に行なうことができる。また、たとえば、いずれかの検知結果が所定条件を満たした場合に、通常よりも高いサンプリング周波数、かつ、通常よりも大きい量子化ビット数のセンシングデータが他の実センサ１２から処理モジュール１１０，１１０Ａへ送信されてもよい。すなわち、いずれかの検知結果が所定条件を満たした場合に、通常よりも情報量が大きいセンシングデータが他の実センサ１２から処理モジュール１１０，１１０Ａへ送信されればよい。 <3-3>
In the first and second embodiments, when the detection result by one of the plurality of actual sensors 12 that outputs sensing data to the common processing modules 110 and 110A satisfies a predetermined condition, Sensing data having a sampling frequency higher than normal is transmitted from the other actual sensors 12 to the processing modules 110 and 110A. However, the contents executed due to any detection result satisfying the predetermined condition are not limited to the above contents. For example, when one of the detection results satisfies a predetermined condition, sensing data having a larger quantization bit number than usual may be transmitted from the other actual sensor 12 to the processing modules 110 and 110A. Even in this case, the control units 180 and 180A can detect occurrence of an event with higher accuracy based on sensing data having a large number of quantization bits. Further, for example, when any detection result satisfies a predetermined condition, sensing data having a sampling frequency higher than normal and a quantization bit number larger than normal is transmitted from the other real sensors 12 to the processing modules 110 and 110A. May be sent to. That is, when any detection result satisfies a predetermined condition, sensing data having a larger amount of information than usual may be transmitted from the other actual sensor 12 to the processing modules 110 and 110A.
　１０，１０Ａ　センサネットワークシステム、１１　センサネットワークアダプタ、１２　実センサ、１４　センサネットワーク部、１５　インターネット、１００，１００Ａ　仮想センサ管理サーバ、１１０，１１０Ａ　処理モジュール、１２０，１２０Ａ　センサ制御モジュール、１２２，１２２Ａ　判定部、１２４　センサ制御部、１３０　メタデータＤＢ、１８０，１８０Ａ　制御部、１８２　ＣＰＵ、１８４　ＲＡＭ、１８６　ＲＯＭ、１９０，１９０Ａ　記憶部、１９１，１９１Ａ　制御プログラム、１９５　通信Ｉ／Ｆ、１９７　バス、３００　アプリケーションサーバ。 10, 10A sensor network system, 11 sensor network adapter, 12 real sensors, 14 sensor network unit, 15 internet, 100, 100A virtual sensor management server, 110, 110A processing module, 120, 120A sensor control module, 122, 122A determination unit , 124 sensor control unit, 130 metadata DB, 180, 180A control unit, 182 CPU, 184 RAM, 186 ROM, 190, 190A storage unit, 191, 191A control program, 195 communication I / F, 197 bus, 300 application server .

Claims (7)

  1. 　複数のセンサの各々を制御するセンサ制御装置であって、
    　前記複数のセンサに含まれる少なくとも２以上のセンサの各々は、処理モジュールへセンシングデータを送信し、
    　前記処理モジュールは、受信された複数のセンシングデータに基づいてイベントの発生を検知し、
    　前記センサ制御装置は、
    　前記２以上のセンサに含まれるいずれかのセンサによる検知結果が、前記イベントが発生している可能性が高いことを示す所定条件を満たすか否かを判定する判定部と、
    　前記検知結果が前記所定条件を満たすと判定された場合に、前記検知結果が前記所定条件を満たさないと判定された場合よりも、情報量の大きいセンシングデータを前記処理モジュールへ送信するように、前記２以上のセンサに含まれる他のセンサを制御するセンサ制御部とを備える、センサ制御装置。     A sensor control device that controls each of a plurality of sensors,
    Each of at least two or more sensors included in the plurality of sensors transmits sensing data to the processing module;
    The processing module detects the occurrence of an event based on a plurality of received sensing data,
    The sensor control device includes:
    A determination unit that determines whether a detection result by any of the two or more sensors includes a predetermined condition indicating that the event is highly likely to occur;
    When it is determined that the detection result satisfies the predetermined condition, the sensing data having a larger amount of information is transmitted to the processing module than when the detection result is determined not to satisfy the predetermined condition. And a sensor control unit that controls other sensors included in the two or more sensors.
  2. 　前記判定部は、前記複数のセンサから一部のセンサを抽出し、
    　前記一部のセンサは、検知結果が前記所定条件を満たすと判定されたセンサと、共通の前記処理モジュールへセンシングデータを送信し、
    　前記センサ制御部は、前記検知結果が前記所定条件を満たすと判定された場合に、前記検知結果が前記所定条件を満たさないと判定された場合よりも、情報量の大きいセンシングデータを前記処理モジュールへ送信するように、前記判定部によって抽出されたセンサを制御する、請求項１に記載のセンサ制御装置。     The determination unit extracts some sensors from the plurality of sensors,
    The some sensors transmit sensing data to the processing module in common with the sensor determined that the detection result satisfies the predetermined condition,
    When the detection result is determined to satisfy the predetermined condition, the sensor control unit outputs sensing data having a larger amount of information than when the detection result is determined not to satisfy the predetermined condition. The sensor control device according to claim 1, wherein the sensor extracted by the determination unit is controlled so as to be transmitted to the sensor.
  3. 　前記判定部は、各々が共通の前記処理モジュールへセンシングデータを送信する複数のセンサを互いに関連付けるメタデータを参照することによって、前記複数のセンサからセンサを抽出する、請求項２に記載のセンサ制御装置。 The sensor control according to claim 2, wherein the determination unit extracts a sensor from the plurality of sensors by referring to metadata that associates the plurality of sensors each transmitting sensing data to the common processing module. apparatus.
  4. 　前記処理モジュールと、前記処理モジュールへセンシングデータを送信するセンサとによって仮想センサが形成される、請求項１から請求項３のいずれか１項に記載のセンサ制御装置。 The sensor control device according to any one of claims 1 to 3, wherein a virtual sensor is formed by the processing module and a sensor that transmits sensing data to the processing module.
  5. 　前記処理モジュールは、複数の入力チャネルの各々に入力される時系列データに基づいて、前記イベントの発生を検知するとともに尤度を算出する、請求項１から請求項４のいずれか１項に記載のセンサ制御装置。 5. The processing module according to claim 1, wherein the processing module detects occurrence of the event and calculates a likelihood based on time-series data input to each of a plurality of input channels. Sensor control device.
  6. 　複数のセンサの各々を制御するセンサ制御方法であって、
    　前記複数のセンサに含まれる少なくとも２以上のセンサの各々は、処理モジュールへセンシングデータを送信し、
    　前記処理モジュールは、受信された複数のセンシングデータに基づいてイベントの発生を検知し、
    　前記センサ制御方法は、
    　前記２以上のセンサに含まれるいずれかのセンサによる検知結果が、前記イベントが発生している可能性が高いことを示す所定条件を満たすか否かを判定するステップと、
    　前記検知結果が前記所定条件を満たすと判定された場合に、前記検知結果が前記所定条件を満たさないと判定された場合よりも、情報量の大きいセンシングデータを前記処理モジュールへ送信するように、前記２以上のセンサに含まれる他のセンサを制御するステップとを含む、センサ制御方法。     A sensor control method for controlling each of a plurality of sensors,
    Each of at least two or more sensors included in the plurality of sensors transmits sensing data to the processing module;
    The processing module detects the occurrence of an event based on a plurality of received sensing data,
    The sensor control method includes:
    Determining whether a detection result by any of the two or more sensors includes a predetermined condition indicating that the event is highly likely to occur;
    When it is determined that the detection result satisfies the predetermined condition, the sensing data having a larger amount of information is transmitted to the processing module than when the detection result is determined not to satisfy the predetermined condition. Controlling another sensor included in the two or more sensors.
  7. 　複数のセンサの各々を制御する処理をコンピュータに実行させるように構成されたプログラムであって、
    　前記複数のセンサに含まれる少なくとも２以上のセンサの各々は、処理モジュールへセンシングデータを送信し、
    　前記処理モジュールは、受信された複数のセンシングデータに基づいてイベントの発生を検知し、
    　前記プログラムは、
    　前記２以上のセンサに含まれるいずれかのセンサによる検知結果が、前記イベントが発生している可能性が高いことを示す所定条件を満たすか否かを判定するステップと、
    　前記検知結果が前記所定条件を満たすと判定された場合に、前記検知結果が前記所定条件を満たさないと判定された場合よりも、情報量の大きいセンシングデータを前記処理モジュールへ送信するように、前記２以上のセンサに含まれる他のセンサを制御するステップとを前記コンピュータに実行させるように構成されている、プログラム。     A program configured to cause a computer to execute processing for controlling each of a plurality of sensors,
    Each of at least two or more sensors included in the plurality of sensors transmits sensing data to the processing module;
    The processing module detects the occurrence of an event based on a plurality of received sensing data,
    The program is
    Determining whether a detection result by any of the two or more sensors includes a predetermined condition indicating that the event is highly likely to occur;
    When it is determined that the detection result satisfies the predetermined condition, the sensing data having a larger amount of information is transmitted to the processing module than when the detection result is determined not to satisfy the predetermined condition. A program configured to cause the computer to execute a step of controlling another sensor included in the two or more sensors.
PCT/JP2018/043933 2018-03-14 2018-11-29 Sensor control device, sensor control method, and program WO2019176183A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2018-047261 2018-03-14
JP2018047261A JP6512326B1 (en) 2018-03-14 2018-03-14 Sensor control device, sensor control method and program

Publications (1)

Publication Number Publication Date
WO2019176183A1 true WO2019176183A1 (en) 2019-09-19

Family

ID=66530715

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2018/043933 WO2019176183A1 (en) 2018-03-14 2018-11-29 Sensor control device, sensor control method, and program

Country Status (2)

Country Link
JP (1) JP6512326B1 (en)
WO (1) WO2019176183A1 (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014045242A (en) * 2012-08-24 2014-03-13 Oki Electric Ind Co Ltd Virtual sensor creating device, virtual sensor creating method and program
JP2016166802A (en) * 2015-03-10 2016-09-15 株式会社東芝 Position detecting system, position detecting device, and position detecting method
WO2016206110A1 (en) * 2015-06-26 2016-12-29 Intel Corporation Methods and apparatus to adaptively manage data collection devices in distributed computing systems

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5753286B1 (en) * 2014-02-05 2015-07-22 株式会社日立パワーソリューションズ Information processing apparatus, diagnostic method, and program

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014045242A (en) * 2012-08-24 2014-03-13 Oki Electric Ind Co Ltd Virtual sensor creating device, virtual sensor creating method and program
JP2016166802A (en) * 2015-03-10 2016-09-15 株式会社東芝 Position detecting system, position detecting device, and position detecting method
WO2016206110A1 (en) * 2015-06-26 2016-12-29 Intel Corporation Methods and apparatus to adaptively manage data collection devices in distributed computing systems

Also Published As

Publication number Publication date
JP6512326B1 (en) 2019-05-15
JP2019161511A (en) 2019-09-19

Similar Documents

Publication Publication Date Title
US11108575B2 (en) Training models for IOT devices
WO2020199926A1 (en) Image recognition network model training method, image recognition method and device
US20210235543A1 (en) Split predictions for iot devices
US20240070388A1 (en) Lexical analyzer for a neuro-linguistic behavior recognition system
US20190037040A1 (en) Model tiering for iot device clusters
WO2020188794A1 (en) Video system, imaging device, and video processing device
JP2021064099A (en) Information processing apparatus, information processing system, learning model generation method, anomaly prediction method, and program
CN114007231B (en) Heterogeneous unmanned aerial vehicle data unloading method and device, electronic equipment and storage medium
CN104050241A (en) Remote-operating accepting system, remote operating system and program
EP3832557B1 (en) Determination apparatus, determination method, determination program, and related system, method, program and computer readable recording medium
WO2019176183A1 (en) Sensor control device, sensor control method, and program
JP2020162765A (en) Recognition system and recognition method
WO2019167368A1 (en) Suitability determination device, suitability determination method, and program
JP6477943B1 (en) Metadata generation apparatus, metadata generation method and program
CN117749836B (en) Internet of things terminal monitoring method and system based on artificial intelligence
JP6501007B1 (en) OUTPUT MANAGEMENT DEVICE, OUTPUT MANAGEMENT METHOD, AND PROGRAM
KR102550982B1 (en) Surveillance information processing apparatus for integrated operating of surveillance terminal
US20240135149A1 (en) Anomaly Detection System for Embedded Devices
EP3659321B1 (en) Tiered data processing for iot device clusters
US8719482B2 (en) Electronic device with bus sharing function
WO2023058433A1 (en) Learning device, learning method, sensing device, and data collection method
US11422914B2 (en) Metadata generation apparatus, metadata generation method, and program
CN111602123B (en) Pretreatment determination device, pretreatment determination method, and program
JP6468377B1 (en) Output management apparatus, output management method and program
WO2019159488A1 (en) Data processing device, and data processing method and program

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18909658

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18909658

Country of ref document: EP

Kind code of ref document: A1