JP6132327B1 - Operation management support system, operation management method, server device, and program - Google Patents

Abstract

[PROBLEMS] To incorporate a machine learning method when analyzing sleepiness, appropriately predicting a driver's dozing timing based on information such as a driving route, and issuing a predetermined warning to indicate the state of safe driving. Support operation management to ensure. An operation management support system according to the present invention includes a storage unit (master DB) 214 that stores a predetermined master table, a pre-analysis unit 211c that performs pre-analysis on measurement data, and the master table for the measurement data. Proper operation based on a driver's brain wave, comprising: a primary analysis unit 211d that performs a primary analysis with reference; and a secondary analysis unit 211e that performs a secondary analysis with reference to the master table for the measurement data. It is characterized by providing management support. [Selection] Figure 7

Description

  The present invention relates to an operation management support system, an operation management method, and a server for supporting operation management of a vehicle and the like based on measured brain waves, pulse, blood flow, and the like (hereinafter simply referred to as an electroencephalogram) of the driver of the vehicle and the like. The present invention relates to an apparatus and a program.

  2. Description of the Related Art Conventionally, for safety measures for preventing automobile accidents, human-centered safety measures have been provided in the form of software using an electroencephalogram sensor. For example, in the case of corporate services, the realization of services for medium- and long-distance, late-night flight trucks and high-speed, chartered bus operators is encouraging. It is envied to mitigate risk.

  As this type of technology, for example, Patent Document 1 discloses a sleepiness estimation rule update device that calculates an activity index of an autonomic nerve from RRI data and estimates sleepiness using this.

  Further, in Patent Document 2, an electroencephalogram acquisition apparatus that acquires a driver's brain wave, an electroencephalogram calculation ECU that calculates an attention amount for driving from the acquired electroencephalogram, and a driver visually confirms an object according to the attention amount. A driving support system including a driving support ECU that issues a warning for prompting is disclosed.

  Patent Document 3 discloses a dozing prevention device that detects driver drowsiness, determines signs of dozing, and emits a warning sound based on the determination result. For sleepiness detection, information about the driver's brain waves is used.

  Furthermore, in Patent Document 4, an in-vehicle smartphone that monitors the operation status and a detection terminal can be communicated with each other via Bluetooth (registered trademark), and if the operation status starts falling asleep, a warning is given and the monitoring terminal is connected via the in-vehicle smartphone. In addition, an operating vehicle management system for reporting an emergency state is disclosed.

JP, 2006-120062, A Japanese Patent Laying-Open No. 2015-095111 JP 2013-232143 A JP2013-120409A

  However, all of the techniques according to Patent Documents 1 to 4 determine signs such as dozing, and perform predetermined warnings based on the results, and log information and attribute information related to driving by the driver. Based on this, the timing of the warning has not been predicted, and only a subsequent warning based on the measurement or determination result has been urged.

  The present invention has been made in view of such problems, and incorporates a machine learning method when analyzing sleepiness, and appropriately predicts the timing of a driver's sleep based on information such as a driving route. The purpose is to support the operation management so as to ensure a safe driving state by issuing a predetermined warning or the like.

In order to solve the above problems, an operation management support system according to an aspect of the present invention includes baseline data that is brain wave data acquired before driving, brain wave data acquired during driving, and real-time acquired data that is GPS data. It is an operation management support system that receives an input of measurement data and performs analysis, and extracts only the storage unit that stores a predetermined master table and the detection pattern weighting data that matches the attribute of the measurement data from the master table, From the extracted detection pattern weight data, the one that matches the baseline data is left, and the drowsiness detection time zone related to the detection pattern weight data with the largest weight among the last remaining detection pattern weight data is used as the analysis result. Primary analysis sleepiness detection pattern from the pre-analysis unit to output and the master table Extraction is performed to check whether the primary analysis drowsiness detection pattern data matches the real-time acquired data, and when it matches, the sleepiness detection time zone related to the primary analysis sleepiness detection pattern data is output as an analysis result 1 Only the detection pattern weighting data that matches the attribute of the measurement data is extracted from the master table, and from the extracted detection pattern weighting data, the one that matches the real-time acquisition data is left, and finally A secondary analysis unit that outputs, as an analysis result, a sleepiness detection time zone related to the detection pattern weighting data having the largest weight among the remaining detection pattern weighting data, and a driver based on the sleepiness detection time zone related to the output It is characterized by supporting proper operation management based on the EEG of the brain.

  According to the present invention, a machine learning method is adopted when analyzing sleepiness, and the driver's dozing timing is appropriately predicted based on information such as a driving route, and a safe warning is issued by issuing a predetermined warning or the like. It is possible to provide an operation management support system, an operation management method, a server device, and a program that support operation management so as to ensure the state of the system.

It is a lineblock diagram of the operation management support system concerning one embodiment of the present invention. It is a figure which shows a measurement data image. It is a figure which shows an attribute classification master image. It is a figure which shows a measurement data extraction image. It is a figure which shows a drowsiness detection pattern candidate master image. It is a figure which shows a detection pattern weighting data image. It is a detailed block diagram of a server apparatus. It is a flowchart which shows the process sequence of prior measurement by the operation management assistance system which concerns on one Embodiment of this invention. It is a figure which shows the screen transition at the time of prior measurement. It is a flowchart which shows the process sequence in driving | operation by the operation management assistance system which concerns on one Embodiment of this invention. It is a figure which shows the screen transition during a driving | operation. It is a figure which shows a management screen. It is a figure which shows a management screen. It is a flowchart which shows the process sequence of detection pattern weighting indexing. It is a flowchart which shows the process sequence of prior measurement analysis. It is a flowchart which shows the process sequence of the primary analysis during a driving | operation. It is a flowchart which shows the process sequence of the secondary analysis during a driving | operation.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

  In FIG. 1, the structure of the operation management assistance system which concerns on one Embodiment of this invention is shown and demonstrated.

  As shown in FIG. 1, the operation management support system 1 includes a driver-side system 10, a cloud-based system 20, and an administrator-side system 30.

  The driver-side system 10 includes an electroencephalogram sensor unit 11 and a mobile terminal 12 such as a smartphone. Although not shown here, the electroencephalogram sensor unit 11 includes an electroencephalogram sensor, a radio communication unit, a control unit, and the like, and communication with the mobile terminal 12 is realized via the radio communication unit. An application program that can be downloaded by an ASP service or the like is mounted on the mobile terminal 12.

  The system 20 on the cloud includes a web server device 21 and a database server device 22. The web server device 21 executes driver operation management based on data sent from the mobile terminal 12 of the driver side system 10, and the DB of the database server device 22 is referred to as appropriate. Here, two server apparatuses 21 and 22 that are physically separated are shown, but it is needless to say that they may be physically integrated and logically separated.

  On the other hand, the administrator's system 30 is composed of information terminals 31 and 32 such as personal computers and smartphones, and can access the web server device 21 on the cloud and appropriately check the operation management status of the driver. It has become.

  In such a configuration, on the driver side, the brain wave of the driver is measured using an application program installed in the brain wave sensor unit 11 and the portable terminal 12. EEG data is analyzed in real time to evaluate “brain activity”. When the evaluation falls below a certain standard, a warning sound or a voice message is reproduced on the portable terminal 12 side. This warning sound or voice message is merely an example of an output mode for alerting the driver, and it is needless to say that it can be dealt with by reproducing a broad sound including general music (music). Further, in addition to such sound reproduction, or instead of sound reproduction, the electroencephalograph itself may be vibrated to act on the driver's tactile sense and urge multifaceted attention.

  The driver's brain activity, evaluation, and GPS information are periodically transmitted to the web server device 21 on the cloud. If the evaluation is bad, a warning mail is transmitted from the cloud to the manager terminals 31 and 32 of the operation manager. The administrator can check the position and status of the driver from the web screen. The administrator can instruct the application program of the driver's portable terminal 12 to play the voice message.

  Here, the usage scene of this system will be described in detail by dividing it into “Before driving”, “During driving”, and “After driving”. By incorporating all three of these into the business before driving, during driving, and after driving, the service is used to the maximum and the risk of car accidents is reduced.

・ Before driving The driver performs a simple measurement for 1 to 3 minutes before departure. The result can be confirmed by the administrator at the administrator terminals 31 and 32 as well. The purpose of this simple measurement is to make the driver highly aware of health care. Dangers inherent in the driver are detected by brain activity and evaluated objectively. Daily care is important because it cannot be repaired by appearance or words. On the other hand, the fact that the administrator checks each time also leads to an improvement in awareness. The operation plan for the day can be adjusted according to the evaluation of simple measurement to reduce the risk of accidents. If the evaluation is very bad, the manager can check the driver's mind and body again, considering the replacement of the driver.

・ Driving The driver starts real-time measurement at the start of driving. At the same time, a departure report is sent to the manager. If the evaluation falls below a certain standard during traveling, a warning sound or a voice message is reproduced on the portable terminal 12 side by the function of the application program. Be aware of the driver at an early stage so that the state of distraction and shaking will not last for a long time. Further, even when approaching the cautionary point, a warning sound can be sounded from the portable terminal 12 side and a voice message can be reproduced by the function of the application program. It will be the location where the administrator has registered the location based on the location where the accident occurred in the past and the driver's diary. On the other hand, the administrator can check the brain activity on the web together with the driver's location information. It is possible to instruct the application program to play a warning sound or a voice message according to the situation. For example, like a message saying "Please stop the bus and call me". The aim is for the administrator to detect changes in the driver and give instructions to increase safety. If you make a phone call while driving, it is assumed that attention will be lost and pressure will be put on, so email notification by the function of the application program is appropriate.

・ After driving The driver finishes real-time measurement. At the same time, a break or arrival report is sent to the manager. You can see the evaluation and transition of brain activity while driving in a graph. Compared with past evaluations and self-evaluations, have them become aware of the next driving improvement. Register the drive diary and inform the administrator of information for safety measures. The administrator terminals 31 and 32 report whether the route / rest was appropriate or whether there was a dangerous road. On the other hand, the administrator uses the administrator terminals 31 and 32 to look at the drive diary and evaluation and register a location, and review the route and break. A warning level can be set for each driver. If the driver does not continue to concentrate, take measures such as changing to short- to medium-distance driving or taking many breaks.

  Next, the outline of each table configuration is as follows.

  On the database server device 22 side on the cloud, various data are held in a table format so that they can be accessed from the web server device 21.

  First, the “master table” includes a user master (user ID, user name, date of birth, gender and skill level ID), course master (course ID, course name, departure place, arrival place, waypoint 1, 2,... N. ).

  Next, the “transaction table” includes a measurement header (measurement ID, user ID, course ID, operation status), measurement section (section ID, measurement ID, sequence, section start date / time, section end date / time, prior measurement result, drowsiness detection time. ), Baseline (section ID, count (seconds), electroencephalogram data), real time (section ID, count (seconds), GPS position data, electroencephalogram data), alert (alert ID, section ID, occurrence time, alert type, level) ).

  “Aggregation table” includes detection pattern weighting (attribute classification ID, number of attribute classification, drowsiness detection time zone, number of drowsiness detection time zone, drowsiness detection pattern candidate ID, success number, success rate, failure number, failure rate, weight, adoption Or not).

  The “master table” includes a time zone master (time zone ID, start time, end time), skill level master (skill level ID, label, skill level, license color), sleepiness detection pattern candidate master (sleepiness detection pattern candidate). ID, sleepiness detection time zone, electroencephalogram band judgment pattern, parameter), attribute classification master (attribute classification ID, user ID, course ID, time zone ID, age group, gender, skill level ID, aggregation target flag), primary Analysis drowsiness detection pattern master (primary analysis drowsiness detection pattern ID, drowsiness detection time zone, brain wave band determination pattern, parameter), and secondary analysis drowsiness detection pattern master (secondary analysis drowsiness detection pattern ID, drowsiness detection time zone, EEG band determination pattern, parameter).

  Hereinafter, a processing procedure from measurement to control will be described on the premise of the table configuration described above.

  The image of the measurement data by this system is as shown in FIG.

  That is, the measurement data is roughly divided into a measurement header, measurement section data, baseline data, real-time acquisition data, and alert data. The measurement header includes a measurement ID, a user ID, a course ID, and operation status data. The measurement section data includes a section ID and a sequence, a section start date and time, a section end date and time, a preliminary measurement result, a drowsiness detection time, and the like. Baseline data is brain wave data. The real-time acquisition data is brain wave data and GPS position data. The alert data includes alert ID, alert occurrence time, alert type, and alert level data.

  For example, for the measurement ID “103”, the user ID is 5 (Fujishima Shiroro), the course ID is C (Kanazawa / Tokyo), the operation status is driving, the section ID is 5, the sequence is 1, the section start date is 2016/07/24, section end date and time is blank, pre-measurement result is 2: caution, no drowsiness detection time. Further, for the alert data, the alert ID is 1104, the occurrence time is 2016/07/24 23:00, the alert type is driving, and the level is 2: caution.

  The image of the attribute classification master is as shown in FIG.

That is, in this embodiment, grouping is performed according to the pattern of conditions registered in the attribute classification master. The contents of the item number classification are as follows.
0 items: All data is targeted 1 item: 1 item is linked to one value, all data is 1 group 2 items: 2 items is linked to 2 values, all data is 1 group 3 items : All data associated with three values of three items is one group. Here, the classification by the number of three items consisting of an arbitrary combination of user ID, course ID, and time zone ID is illustrated. However, this is merely an example, and it is needless to say that grouping by a larger number of items can be comprehensively performed according to the type and type of IDs to be combined.

  In the example of FIG. 3, in the attribute classification master, the item number classification, user ID, course ID, time zone ID, age group, gender, skill level ID, and aggregation target flag are associated with the attribute classification ID. Are managed respectively.

  Here, FIG. 4 illustrates an image of measurement data extraction.

  As shown in FIG. 4A, after one piece of measurement data is extracted, measurement data is extracted based on the above conditions as shown in FIGS. 4B and 4C. Sort by detection time and group by sleepiness detection time zone.

  The image of the sleepiness detection pattern candidate master is as shown in FIG.

As shown in the figure, from drowsiness detection pattern candidate master, the parameters extracted for each determination pattern of bandwidth brainwave, examining the number of detected erroneous speed detection from each of sleepiness detection time period.

  FIG. 5A shows an example of a drowsiness detection pattern candidate master. A drowsiness detection pattern candidate ID, a drowsiness detection time zone, an electroencephalogram band determination pattern, and parameters (excess, less flag) (threshold) ( The number of occurrences) is stored in association with each other.

  An example of the primary analysis drowsiness detection pattern master is shown in FIG. 5B. The primary analysis drowsiness detection pattern ID, the sleepiness detection time zone, the determination pattern of the electroencephalogram band, and the parameters related to γ waves (excess Less than flag) (threshold value) (number of occurrences) and parameters related to the θ wave (excess / less flag) are stored in association with each other. Similarly, an example of the secondary analysis drowsiness detection pattern master is shown in FIG. 5C, and the secondary analysis drowsiness detection pattern ID corresponds to the determination pattern and parameter (threshold value) of the drowsiness detection time zone and the electroencephalogram band. Attached and memorized.

  In this way, finally, the detection pattern weighting data as shown in FIGS. 6A and 6B is collected. That is, the attribute classification, the number of attribute classification, the sleepiness detection, the matching rate, the weight, and the adoption availability are related. Of these, it is preferable to preferentially refer to a record with a high matching rate, that is, a record with a high detection rate, to control the timing of alerts related to driving. Actually, whether or not to adopt each record is determined based on the matching rate, and therefore, the control may be performed with reference to the adoption.

  Here, FIG. 7 shows a detailed configuration of the server apparatus and will be described.

  As shown in the figure, the server device 21 includes a control unit 211. The control unit 211 reads out and executes the program 216 in the storage unit 215, thereby providing a main control unit 211a, an authentication unit 211b, a pre-analysis unit 211c, a primary analysis unit 211d, a secondary analysis unit 211e, and a screen generation unit 211f. Function. In addition, the control unit 211 is connected to the communication unit 212, the user DB 213, and the master DB 214.

  In such a configuration, the control unit 211 functions as the pre-analysis unit 211c, the secondary analysis unit 211d, and the secondary analysis unit 211e in particular to perform appropriate operation management support based on the driver's brain waves. Make it possible.

  Next, a processing procedure performed by the system according to the embodiment of the present invention will be described in detail with reference to the flowchart of FIG. Here, the processing procedure from pairing of the electroencephalogram sensor unit and the portable terminal to preliminary measurement analysis will be mainly described.

  The portable terminal performs pairing with the electroencephalogram sensor unit (S1). Then, a signal is transmitted from the electroencephalogram sensor unit to the portable terminal by short-range wireless communication (S2), and a screen switching guidance display is performed on the portable terminal (S3). Then, a white screen is drawn (n seconds) (S4).

  Subsequently, when the signal is transmitted from the electroencephalogram sensor unit by short-range wireless communication (S5), the closed eye guidance display is performed (S6), and when the signal is transmitted again from the electroencephalogram sensor unit to the portable terminal (S7). Baseline measurement is started (S8).

  Next, the mobile terminal determines whether or not the eyes are continuously closed for m seconds (S9). When the eyes are closed, the baseline measurement is stopped (S10), and the eye opening guidance is performed (S11). The line is transmitted (S12).

  The server device stores the baseline in the DB (S13), and performs pre-measurement analysis (S14). Then, the analysis result is stored in the DB (S15), and the analysis result is transmitted to the administrator terminal by mail (S16). Further, the analysis result is returned to the portable terminal (S17). The analysis result to be returned includes data of a primary analysis drowsiness detection pattern master to be used later.

  The mobile terminal stores the analysis result (S18) and displays the analysis result (S19). When the administrator terminal sends an operation start approval to the server device (S20) and an approval notification is sent from the server device to the mobile terminal (S21), the mobile terminal starts operation (S22).

  In addition, the screen transition of FIG. 9 has shown the transition of the screen from the pairing in the flowchart of FIG. 8 to an operation start. At the time of pairing, a message to that effect is displayed on the screen, and screen guidance, white screen, eye-closed guidance, and eye-opening guidance are each suggested by switching screens as shown in the figure. In the result display, the start of operation can be selected.

  Next, a processing procedure performed by the system according to the embodiment of the present invention will be described in detail with reference to the flowchart of FIG. Here, the processing procedure from the primary analysis, the secondary analysis to the end of operation will be mainly described.

  When the mobile terminal starts driving (S30), the mobile terminal transmits a driving start notification to the server device (S31). The server device transmits an operation start mail to the administrator terminal (S32). On the other hand, when a signal is transmitted from the electroencephalogram sensor unit by short-range wireless communication (S33), the mobile terminal temporarily stores the electroencephalogram data, acquires the GPS data, and temporarily stores it (S34, S35). Then, these data are transmitted to the server device (S36).

  Next, the mobile terminal performs a primary analysis (S37).

  The server device stores the received data in the DB (S38) and performs a secondary analysis (S39). Then, after storing the analysis result in the DB (S40), the analysis result is sent by e-mail (S41). Moreover, an analysis result is returned to a portable terminal (S42). In this embodiment, the analysis result is sent by e-mail in step S41 only when, for example, a sign of sleepiness worthy of notification is confirmed, but it is of course not limited thereto. .

  In the portable terminal, the returned analysis result is stored (S43), and the analysis result is displayed (S44). About this display, you may use together with a character, a color, a sound, a vibration, etc.

  On the other hand, in the administrator terminal, the operation status confirmation is performed on the server device (S45). Or management alert setting is performed with respect to a server apparatus (S46). In the server device, management alert setting and saving to DB are performed (S47).

  In the portable terminal, the management alert setting is confirmed with respect to the server device (S48), and the confirmation is saved (S49). Management alert display (character, color, sound, vibration, etc.) is performed (S50), and prediction alert display is performed (S51). Then, it is determined whether or not the operation is to be ended (S52). When the operation is to be ended, the operation is ended (S52), and an operation end notification is sent to the server device (S53). The server device transmits an operation end mail to the administrator terminal (S54), and the administrator terminal confirms the operation result (S55).

  Note that the screen transition in FIG. 11 shows screen switching related to the suggestion of each state during driving. When it is in a safe state, “Good state” is displayed, and at the time of warning, “Concentration is low. Please take a break” is displayed. At one time, “Drowsiness was detected. It ’s dangerous. Take a break right away.” Is displayed (you can also notify by sound or vibration), and when you finish driving, “Thank you for your hard work. Please keep your regular lifestyle in mind. ”Is displayed. Of course, these are examples, and the present invention is not limited to these.

  12 and 13 show examples of management screens displayed on the administrator terminal. In the management screen shown in FIG. 12, it is possible to narrow down the operation list according to the operation state, and after narrowing down, the driver, the course, the operation status, etc. can be confirmed in the list. On the other hand, on the management screen shown in FIG. 13, the route and its quality (good, warning, dangerous) are further displayed on the map.

  With reference to the flowchart of FIG. 14, the detection pattern weighting index processing procedure will be described in detail. This is common to the preliminary analysis and the secondary analysis.

  As shown in FIG. 3, a record whose aggregation target flag of the attribute classification master is “1” is extracted, and the process is repeated for each record (S101). Then, measurement data is extracted based on the record condition of S101 (S102), and the process is repeated one record at a time from the group having a short sleepiness detection time zone (S103). Then, the process is repeated one record for each data of the sleepiness detection pattern candidate master in the same sleepiness detection time zone as shown in FIG. 5 (S104).

  Then, the process is repeated for each record for each measurement data extracted in step S102 (S106). Thus, based on the detection pattern candidate, the extracted measured electroencephalogram data is evaluated using deep learning, and it is determined whether or not the pattern matches (S107).

  If it is determined in S107 that they match, if they match the same drowsiness detection time zone, the number of detection is counted (S108). If it coincides with a different drowsiness detection time zone, the number of false detections is counted (S109). Thus, the repetition of S105 is finished (S110), and the detection rate and the false detection rate are registered (S111). Then, the repetition of S104 ends (S112), and the detection pattern weight is calculated from the detection rate weight master and the false detection rate weight master based on the detection rate and the false detection rate for each sleepiness detection time zone of the attribute classification. Then, it is decided whether or not to adopt (S113).

  Thus, the repetition of S103 is ended (S114), the repetition of S101 is ended (S115), and this process is ended.

  Next, with reference to the flowchart of FIG. 15, the processing procedure of the prior measurement analysis will be described.

  Only the detection pattern weighting data that matches the attribute to be analyzed is extracted from the detection pattern weighting table (S201), and the detection pattern record that matches the baseline electroencephalogram data is left out of the extracted detection pattern weighting records ( S202). Then, the sleepiness detection time zone of one record having the maximum weight among the remaining weighting data is returned as an analysis result (S203), and the pre-measurement analysis process is terminated.

  Next, the primary analysis process will be described with reference to the flowchart of FIG.

  When the primary analysis drowsiness detection pattern data is acquired (S301), it is confirmed whether or not the primary analysis sleepiness detection pattern data matches the real-time brain wave data (S302). If they match, the “sleepiness detection time zone” of the primary analysis sleepiness detection pattern data is returned as the analysis result (S303), and the primary analysis processing is terminated.

  Next, the secondary analysis process will be described with reference to the flowchart of FIG.

  Only the detection pattern weighting data matching the analysis target attribute is extracted from the secondary analysis detection pattern weighting table (S401), and the detection pattern record matching the real-time electroencephalogram data is left out of the extracted detection pattern weighting records. (S402). If they match, the “sleepiness detection time zone” of the secondary analysis sleepiness detection pattern data is returned as the analysis result (S403), and the secondary analysis processing is terminated.

  Although one embodiment of the present invention has been described above, the present invention is not limited to this, and it is needless to say that various improvements and changes can be made without departing from the spirit of the present invention. For example, in the above embodiment, a system that supports management related to the operation of a car has been illustrated. However, the present invention is not limited to this, and a wide range of work that requires safety and efficiency, such as work at a factory or class at a university. Of course, it is applicable to.

DESCRIPTION OF SYMBOLS 1 ... Operation management support system 10 ... Driver side system 11 ... EEG sensor unit 12 ... Mobile terminal 20 ... Cloud system 21 ... Web server apparatus 22 ... Database server apparatus 30 ... Administrator side system 31,32 ... Administrator terminal.

Claims (3)

  A driving management support system that receives and analyzes measurement data including baseline data that is brain wave data acquired before driving, brain wave data acquired during driving, and real-time acquired data that is GPS data,
  A storage unit storing a predetermined master table;
  Only the detection pattern weighting data that matches the attribute of the measurement data is extracted from the master table, and from the extracted detection pattern weighting data, the one that matches the baseline data is left, and the last remaining detection pattern weighting A pre-analysis unit that outputs a drowsiness detection time zone related to detection pattern weighting data with the largest weight among the data as an analysis result;
  A primary analysis drowsiness detection pattern is extracted from the master table, and whether or not the primary analysis drowsiness detection pattern data matches the real-time acquired data is determined. A primary analysis unit that outputs the sleepiness detection time zone as an analysis result;
  Only the detection pattern weighting data that matches the attribute of the measurement data is extracted from the master table, and from the extracted detection pattern weighting data, the one that matches the real-time acquisition data is left, and the last remaining detection pattern weighting A secondary analysis unit that outputs, as an analysis result, a sleepiness detection time zone related to detection pattern weighting data having the largest weight among the data,
  Supporting appropriate operation management based on a driver's brain wave based on the sleepiness detection time zone related to the output
  Operation management support system.   Equipped with a storage unit storing a predetermined master table, a pre-analysis unit, a primary analysis unit, and a secondary analysis unit, baseline data that is EEG data acquired before driving and EEG data acquired during driving And an operation management support method by an operation management support system that receives and analyzes measurement data including real-time acquisition data that is GPS data,
  The pre-analysis unit extracts only the detection pattern weighting data that matches the attribute of the measurement data from the master table, and leaves the extracted detection pattern weighting data that matches the baseline data, and finally Output the sleepiness detection time zone related to the detection pattern weighting data with the largest weight among the remaining detection pattern weighting data as an analysis result,
  The primary analysis unit extracts a primary analysis drowsiness detection pattern from the master table, checks whether the primary analysis sleepiness detection pattern data matches the real-time acquired data, and if it matches, the primary analysis Output the sleepiness detection time zone related to the analysis sleepiness detection pattern data as an analysis result,
  The secondary analysis unit extracts only the detection pattern weighting data that matches the attribute of the measurement data from the master table, and leaves the extracted detection pattern weighting data that matches the real-time acquisition data, Output the sleepiness detection time zone related to the detection pattern weighting data with the largest weight among the remaining detection pattern weighting data as an analysis result,
  Supporting appropriate operation management based on a driver's brain wave based on the sleepiness detection time zone related to the output
  Operation management support method.   Equipped with a storage unit that stores a predetermined master table and receives input of measurement data including baseline data that is brain wave data acquired before driving, brain wave data acquired during driving, and real-time acquired data that is GPS data A program executed on a computer that performs
  Computer
  Only the detection pattern weighting data that matches the attribute of the measurement data is extracted from the master table, and from the extracted detection pattern weighting data, the one that matches the baseline data is left, and the last remaining detection pattern weighting A pre-analysis unit that outputs a drowsiness detection time zone related to detection pattern weighting data with the largest weight among the data as an analysis result;
  A primary analysis drowsiness detection pattern is extracted from the master table, and whether or not the primary analysis drowsiness detection pattern data matches the real-time acquired data is determined. A primary analysis unit that outputs the sleepiness detection time zone as an analysis result;
  Only the detection pattern weighting data that matches the attribute of the measurement data is extracted from the master table, and from the extracted detection pattern weighting data, the one that matches the real-time acquisition data is left, and the last remaining detection pattern weighting Function as a secondary analysis unit that outputs the drowsiness detection time zone related to the detection pattern weighting data with the largest weight among the data as an analysis result,
  Supporting appropriate operation management based on a driver's brain wave based on the sleepiness detection time zone related to the output
  program.