JP7698591B2 - Action procedure information management device and action procedure information management system - Google Patents

Description

The present invention relates to an action procedure information management device and an action procedure information management system.

As part of managing the water quality of water supply and sewerage systems, processes such as injecting chemicals into raw water at water purification plants to make it potable water, and pumping air into sewage to settle sludge and turn it into treated water at sewage treatment plants are monitored and controlled. Recently, technology has been developed that allows unskilled drivers to refer to the driving history of experienced drivers by using a computer to search for past dates and times when the trends in specific indicators such as the turbidity and pH (hydrogen ion concentration) of raw water or sewage are similar to a reference date and time, and extracts event information such as alarms and operation information from those past dates and times as operating history.

Of these technologies, the management operation support device in Patent Document 1 extracts event information such as warnings and operation/driving history information from past dates and times when specific indicators are similar to a reference date and time, and obtains documents and action procedures (countermeasures) prepared to deal with the event information. Using the event information and documents, the device evaluates risk based on the amount of damage caused by the abnormality and the time required to deal with it, embeds the risk evaluation results in the action procedures, and provides feedback to the extraction of action procedures.

JP 2013-92954 A

When searching for action procedures based on past event information, a large amount of event information is matched, and some of the event information is unrelated to improving a specific index, so the search accuracy may be reduced.
The object of the present invention is to use event information from past dates and times when the trend of a specific indicator is similar to a reference date and time to search with high accuracy for documents to be viewed, their locations, and action procedures by an experienced driver, and to display them as the driving history of the experienced driver.

The action procedure information display device of the present invention is characterized by comprising a similar date and time extraction unit that extracts one or more past dates and times in which the trend of a specific indicator indicating water quality related to a plant is similar to a reference date and time, an event extraction unit that extracts multiple event information that is multiple alarms issued related to the specific indicator or the content of the operation/driving of the plant operator at each of the extracted one or more past dates and times, and an event screening unit that selects the extracted multiple event information based on the degree of involvement between the specific indicator and each of the multiple event information.
Other means will be described in the description of the embodiment of the invention.

According to the present invention, by using event information from past dates and times when the trend of a specific index is similar to a reference date and time, documents to be viewed, their locations, and action procedures taken by an expert driver can be searched for with high accuracy, and displayed as the driving history of the expert driver.

FIG. 1 is a diagram showing a configuration of an action procedure information display system. FIG. 2 is a diagram illustrating the hardware configuration of a water quality data collection server. 2 is a diagram showing the hardware configuration of a water quality monitoring control server. FIG. FIG. 2 is a diagram illustrating a hardware configuration of an action procedure information management server. FIG. 2 is a diagram illustrating a hardware configuration of a driver's information terminal. FIG. 2 is a diagram illustrating a hardware configuration of a surveillance camera. FIG. 2 is a diagram showing the hardware configuration of a water quality data storage unit. FIG. 2 is a functional block diagram of an action procedure information management server. FIG. 13 is a diagram illustrating a data structure. 13 is a process flow when action procedure information is generated. 13 is a process flow when action procedure information is displayed. FIG. 4 is a screen transition diagram of the driver's information terminal.

Below, an embodiment of the present invention (referred to as "the present embodiment") will be described with reference to the drawings. The present embodiment is an example of managing water quality in a water supply-related facility (not shown). A water supply management facility is a concept that includes water purification plants and sewage treatment plants. However, the present invention can also be applied to general plants (not shown) other than water supply management facilities.

Figure 1 is a diagram showing the configuration of an action procedure information display system in this embodiment. The action procedure information display system 100 comprises a water quality data collection server 1, a water quality data storage unit 2, an event information storage unit 3, a water quality monitoring and control server 4, a document information storage unit 5, a document viewing information storage unit 6, a driver's information terminal 7, an action procedure information storage unit 8, a surveillance camera 9, and an action procedure information management server (action procedure information management device) 10. These are connected by a network 11.

In plants other than water management facilities, the water quality data collection server 1 becomes a "sensor data collection server 1" that collects any sensor value in the plant, and the water quality monitoring and control server 4 becomes a "plant monitoring and control server 4" that monitors and controls any equipment that constitutes the plant. The action procedure information management device 10, the sensor data collection server 1, the plant monitoring and control server 4, and the driver's information terminal 7 constitute the action procedure information management system 100.

The water quality data collection server 1 uses sensors installed at specific locations such as water intake and receiving wells to collect data on specific water-related indicators such as turbidity and pH as water quality data, assigns a timestamp (date and time) to the data and transmits it to the water quality data accumulation unit 2. The water quality data collection server 1 comprises a water quality data measurement unit 1a and a water quality data transmission unit 1b. The water quality data measurement unit 1a uses sensors to collect data on specific indicators as water quality data. The water quality data transmission unit 1b assigns a timestamp to the water quality data and transmits it to the water quality data accumulation unit 2.

The water quality data storage unit 2 records the water quality data collected by the water quality data collection server 1, and transmits the water quality data to other servers in response to requests from the other servers. The event information storage unit 3 collects alarms and operation/driving history from the water quality monitoring and control server 4, records event information that consolidates these, and transmits the event information to the action procedure information management server 10 in response to requests from the action procedure information management server 10.

The water quality monitoring and control server 4 issues warnings when abnormalities occur in the water quality, controls or operates water quality control equipment such as coagulant injectors, and collects warnings and operation/driving history as event information and transmits it to others. The water quality monitoring and control server 4 includes a water quality data receiving unit 4a, an abnormality warning issuing unit 4b, an equipment control/operation unit 4c, an event information collecting unit 4d, and an event information transmitting unit 4e.

The water quality data receiving unit 4a receives water quality data from the water quality data collection server 1. The abnormality warning issuing unit 4b judges whether the water quality data is normal or abnormal, and issues a warning when an abnormality occurs. The equipment control/operation unit 4c operates and controls equipment that injects coagulants in water purification treatment and equipment that blows air in sewage treatment. The event information aggregation unit 4d digitizes information on the type of warning, coagulant injection rate, air blowing volume, and equipment operation volume when a warning is issued, when coagulant is injected, when air blowing treatment is performed, etc., and aggregates it as event information. The event information transmission unit 4e transmits the event information to the event information storage unit 3 or the action procedure information management server 10.

The document information storage unit 5 digitizes documents describing device control and stores them as electronic manuals.
The document viewing information storage unit 6 may store the viewing history of the electronic manual viewed by the driver of the equipment via the driver's information terminal 7, or may use the surveillance camera 9 to recognize the page of the document that the driver was viewing and create a viewing history.

The driver's information terminal 7 provides the driver with functions such as displaying an electronic manual, taking pictures with a camera, writing annotations, and accepting the selection of a similarity calculation formula type. The driver's information terminal 7 includes an electronic manual viewer 7a, a camera unit 7b, an electronic manual viewing history and annotation information transmission unit 7c, a similarity calculation formula type selection acceptance and transmission unit 7d, and an action procedure information reception and display unit 7e.

The electronic manual viewer 7a has a function of receiving and displaying data of the electronic manual from the document information storage unit 5 in response to a request from the driver, and also a function of adding annotations to specified locations in the electronic manual. The camera shooting unit 7b shoots videos or still images using a camera built into the driver's information terminal 7. The electronic manual viewing history/annotation information sending unit 7c stores the viewing history of the electronic manual and sends the viewing history and annotation information to the document viewing information storage unit 6. The similarity calculation formula type selection receiving/transmitting unit 7d receives the similarity calculation formula type (i.e., the type of specific index that is the target variable of the similarity calculation formula) and sends the result to the action procedure information management server 10. The action procedure information receiving/displaying unit 7e receives the action procedure information sent from the action procedure information management server 10, displays it on the display, and outputs it as sound to the speaker as necessary.

The action procedure information storage unit 8 stores the action procedure information generated by the action procedure information management server 10, and transmits it to the action procedure information management server 10 when searching for action procedure information.

The surveillance camera 9 captures images of the driver viewing a document (electronic manual), recognizes the manual name and page number from the images, and in some cases gestures as annotations, and transmits these to the document viewing information storage unit 6. The surveillance camera 9 includes a camera capture unit 9a, a manual name/page number/gesture recognition unit 9b, and an information transmission unit 9c. The camera capture unit 9a captures images of the driver viewing a document as video or still images. The manual name/page number/gesture recognition unit 9b recognizes the manual name and page number from the captured video or still images, and in some cases gestures as annotations, and stores these as data. The information transmission unit 9c transmits the recognized manual name, page number, and gestures to the document viewing information storage unit 6.

The action procedure information management server 10 generates, searches, or distributes action procedure information. The action procedure information management server 10 includes a water quality and event information receiving unit 10a, an electronic manual viewing history and annotation information receiving unit 10b, a similarity calculation formula type receiving unit 10c, an action procedure information generating unit 10d, an action procedure information searching unit 10e, and an action procedure information transmitting unit 10f. The water quality and event information receiving unit 10a receives water quality data and event information from the water quality data storage unit 2 and the event information storage unit 3. The electronic manual viewing history and annotation information receiving unit 10b receives electronic manual viewing history and annotation information from the document viewing information storage unit 6, as well as manual name, page number, and gesture recognition results recognized by the surveillance camera 9.

The similarity calculation formula type receiving unit 10c receives an input regarding the type of similarity calculation formula selected by the driver on the driver's information terminal 7 (i.e., selection of a specific index such as turbidity or pH). The action procedure information generating unit 10d aggregates the time-stamped water quality data and event information received by the water quality and event information receiving unit 10a, as well as the document browsing information and annotations received by the electronic manual browsing history and annotation information receiving unit 10b, and the similarity calculation formula type received by the similarity calculation formula type receiving unit 10c, as action procedure information. The action procedure information searching unit 10e searches the action procedure information from the action procedure information storage unit 8 using the water quality data and event information received by the water quality and event information receiving unit 10a as keys. The action procedure information transmitting unit 10f transmits the action procedure information aggregated by the action procedure information generating unit 10d to the action procedure information storage unit 8, and the action procedure information searched by the action procedure information searching unit 10e to the driver's information terminal 7.

Figure 2 is a diagram showing the hardware configuration of the water quality data collection server 1 in this embodiment. The water quality data collection server 1 includes a water quality data collection server main body 1c, a display output unit 1h, an operation input unit 1i, an A/D (Analog/Digital) converter 1j, and a water quality data sensor 1k. The water quality data collection server main body 1c measures and transmits water quality data according to the contents of a program. The water quality data collection server main body 1c includes a CPU (Central Processor Unit) 1d, a RAM (Random Access Memory) 1e, a recording device 1m, and an I/F (Interface) 1g. The CPU 1d executes calculation processing according to a program for collecting water quality data. The RAM 1e temporarily stores data in the program for collecting water quality data. The recording device 1m is a hard disk drive, a compact flash drive (registered trademark), or the like, that records the program for collecting water quality data, data, and the like.

The recording device 1m stores the water quality data measurement program 1a for implementing the water quality data measurement unit 1a, the water quality data transmission program 1b for implementing the water quality data transmission unit 1b, and also stores various data used when executing these programs. For ease of understanding, the same reference numerals are used for the XX unit and the XX program for implementing it (the same applies below).

The I/F 1g receives water quality data from the A/D converter 1j and receives data related to operation input from the operation input unit 1i, outputs the current status to the display output unit 1h, and transmits the water quality data to the network 11. The display output unit 1h is a display, printer, etc. that outputs information to the user about the current status of the water quality data collection server 1, such as normal operation and the occurrence of an abnormality. The operation input unit 1i is a keyboard, touch panel, mouse, etc. that converts operation input from the user into an electrical signal and transmits it to the water quality data collection server main body 1c. The water quality data sensor 1k is a water temperature sensor, turbidity sensor, etc. that measures specific indicators related to water quality from the water and converts them into an electrical signal. The A/D converter 1j converts the electrical signal from the sensor into a digital signal.

Figure 3 is a diagram showing the hardware configuration of the water quality monitoring and control server 4 in this embodiment. The water quality monitoring and control server 4 comprises a water quality monitoring and control server main body 4f, a display output unit 4n, an operation input unit 4k, and facility control equipment 4m. The water quality monitoring and control server main body 4f monitors and controls water quality according to the contents of a program. The water quality monitoring and control server main body 4f comprises a CPU 4g, a RAM 4h, a recording device 4p, and an I/F 4j. The CPU 4g executes calculation processing according to a program for monitoring and controlling water quality. The RAM 4h temporarily stores data in the program for monitoring and controlling water quality. The recording device 4p is a hard disk drive, compact flash drive, etc. that records the program, data, etc. for monitoring and controlling water quality.

The recording device 4p stores a water quality data receiving program 4a for implementing the water quality data receiving unit 4a, an abnormality alarm issuance program 4b for implementing the abnormality alarm issuance unit 4b, and an equipment operation/control program 4c for implementing the equipment operation/control unit 4c. The recording device 4p further stores an event information collection program 4d for implementing the event information collection unit 4d, an event information transmission program 4e for implementing the event information transmission unit 4e, and various data 4i used when executing these programs. As mentioned above, the XX unit and the XX program for implementing it are given the same reference numerals.

I/F4j receives data regarding operation input from operation input unit 4k, outputs the current state to display output unit 4n, and transmits control signals to facility control device 4m, as well as receiving water quality data from network 11 and transmitting event information to network 11.

The display output unit 4n is a display, printer, etc. that outputs information to the user about the current state of the water quality monitoring and control server 4, such as normal operation, occurrence of an abnormality, operation amount, control amount, etc. The operation input unit 4k is a keyboard, touch panel, mouse, etc. that converts operation input from the user into an electrical signal and transmits it to the water quality monitoring and control server main body 4f. The facility control equipment 4m is equipment such as a pump, coagulant injection equipment, and blower that controls water quality.

Figure 4 is a diagram showing the hardware configuration of the action procedure information management server 10 in this embodiment. The action procedure information management server 10 comprises an action procedure information management server main body 10g, a display output unit 10m, and an operation input unit 10n. The action procedure information management server main body 10g generates or distributes action procedure information according to the contents of a program. The action procedure information management server main body 10g comprises a CPU 10h, a RAM 10i, a recording device 10p, and an I/F 10k. The CPU 10h executes calculation processing according to a program for generating or distributing action procedure information. The RAM 10i temporarily stores data in the program for generating or distributing action procedure information. The recording device 10p is a hard disk drive, a compact flash drive, or the like, that records the program, data, etc. for generating or distributing action procedure information.

The recording device 10p stores a water quality and event information receiving program 10a for implementing the water quality and event information receiving unit 10a, an electronic manual viewing history and annotation information receiving program 10b for implementing the electronic manual viewing history and annotation information receiving unit 10b, and a similarity calculation formula type receiving program 10c for implementing the similarity calculation formula type receiving unit 10c. The recording device 10p also stores an action procedure information generating program 10d for implementing the action procedure information generating unit 10d, an action procedure information search program 10e for implementing the action procedure information search unit 10e, an action procedure information transmitting program 10f for implementing the action procedure information transmitting unit 10f, and various data 10j used when executing these programs.

The I/F 10k receives data related to operation input from the operation input unit 10n and outputs the current state to the display output unit 10m, and also receives water quality data, event information, electronic manual browsing history, annotation information, and similarity calculation formula types from the network 11 and transmits action procedure information to the network 11. The display output unit 10m is a display, printer, etc. that outputs information to the user about the current state of the action procedure information management server 10, such as action procedure information generation, normal operation in search processing, and abnormality warnings. The operation input unit 10n is a keyboard, touch panel, mouse, etc. that converts operation input from the user into an electrical signal and transmits it to the action procedure information management server main body 10g.

Figure 5 is a diagram showing the hardware configuration of the driver's information terminal 7 in this embodiment. The driver's information terminal 7 includes a CPU 7f, a RAM 7g, a recording device 7h, an I/F 7j, a display output unit 7k, an operation input unit 7m, and a camera 7n. The CPU 7f executes calculations according to the electronic manual viewer program 7a, the action procedure information receiving and display program 7e, and the like. The RAM 7g temporarily stores data in the electronic manual viewer program 7a, the action procedure information receiving and display program 7e, and the like. The recording device 7h is a hard disk drive, a compact flash drive, or the like that records the electronic manual viewer program 7a, the action procedure information receiving and display program 7e, various data 7i, and the like.

The recording device 7h stores an electronic manual viewer program 7a for implementing the electronic manual viewer 7a, a camera shooting program 7b for implementing the camera shooting unit 7b, and an electronic manual viewing history and annotation information transmission program 7c for implementing the electronic manual viewing history and annotation information transmission unit 7c. The recording device 7h further stores a similarity type selection reception and transmission program 7d for implementing the similarity type selection reception and transmission unit 7d, an action procedure information reception and display program 7e for implementing the action procedure information reception and display unit 7e, and various data 7i used when executing these programs.

The I/F 7j receives data related to operation input from the operation input unit 7m and outputs the current state to the display output unit 7k, and also receives electronic manuals and action procedure information and transmits electronic manual viewing history, annotation information, and similarity calculation formula types to the network 11. The display output unit 7k is a display, printer, etc. that outputs information to the user about the current state of the action procedure information management server 10, such as action procedure information generation, normal operation in search processing, and abnormality warnings. The operation input unit 7m is a keyboard, touch panel, mouse, etc. that converts operation input from the user into an electrical signal and transmits it to the action procedure information management server main body 10g. The camera 7n is used as an auxiliary to the surveillance camera 9 when recognizing the manual name, page number, and gestures from a video or still image of the driver viewing a document.

Figure 6 is a diagram showing the hardware configuration of the surveillance camera 9 in this embodiment. The surveillance camera 9 comprises a CPU 9d, RAM 9e, a recording device 9f, an I/F 9g, an operation input unit 9h, and a camera body 9i. The CPU 9d executes calculation processing according to a program for taking pictures with the camera. The RAM 9e temporarily stores data in the program for taking pictures with the camera. The recording device 9f is a hard disk drive, compact flash drive, etc. that records the program for taking pictures with the camera, data, etc.

The recording device 9f stores a camera shooting program 9a for implementing the camera shooting unit 9a, and a manual name, page number, gesture recognition, and information transmission program 9b for implementing the manual name, page number, and gesture recognition unit 9b and the information transmission unit 9c. The recording device 9f also stores various data 9j used when executing these programs.

The I/F 9g receives data related to operation input from the operation input unit 9h, outputs the current state to the network 11, and transmits the manual name, page number, and gesture recognition results to the network 11. The operation input unit 9h is a switch or the like that converts the operation input from the user into an electrical signal and transmits it to the action procedure information management server main body 10g. The camera main body 9i captures video or still images of the driver viewing a document according to the instructions of the camera shooting program 9a.

Figure 7 is a diagram showing the hardware configuration of the water quality data storage unit 2 (more precisely, a device that stores the water quality data storage unit 2 of Figure 1) in this embodiment. The water quality data storage unit 2 comprises a CPU 2a, a RAM 2b, a recording device 2c, and an I/F 2g. The CPU 2a executes calculation processing according to a program for storing water quality data. The RAM 2b temporarily stores data in the program for storing water quality data. The recording device 2c is a hard disk drive, compact flash drive, etc. that records the program for storing water quality data, water quality data, etc.

The recording device 2c stores a water quality data receiving and writing program 2d that receives water quality data from the network 11 and writes it to the recording device 2c, a water quality data reading and sending program 2e that reads water quality data from the recording device 2c and sends it to the network 11, and the water quality data 2f itself. The I/F 2g sends and receives the water quality data 2f to and from the network 11 according to the instructions of the program stored in the water quality data storage unit 2.

Figure 8 is a functional block diagram of the action procedure information management server 10 in this embodiment. The action procedure information search unit 10e includes a similar date and time extraction unit 10q, an event extraction unit 10r, an event screening unit 10s, and an action procedure information extraction unit 10t. The similar date and time extraction unit 10q receives water quality data from the water quality and event information receiving unit 10a. For the received water quality data, the similar date and time extraction unit 10q calculates the similarity between a specific index, such as turbidity, at a reference date and time such as the present and a specific index at each past time according to a similarity calculation formula selected by the similarity calculation formula type receiving unit 10c. The similar date and time extraction unit 10q extracts several past dates and times in descending order of the calculated similarity. As a similarity calculation formula, an example of a formula for calculating the deviation, which is the inverse of the similarity, is shown in formula (1).

[Equation 1]

"dtw" on the left side of equation (1) is the deviation, which is the inverse of the similarity. "turb(t)" on the right side is the value of a specific index at time t. The specific index is selected from turbidity, pH, etc. The specific index is also called a "target variable" in the sense that it is included as a variable in the similarity calculation formula. " _tb " on the right side is the reference date and time. " _ts " on the right side is the past date and time. "ΔT" on the right side is the sampling period. "i" on the right side is a coefficient (i = 0, 1, 2, ..., n) by which the sampling period is multiplied.

The similar date/time extraction unit 10q randomly acquires multiple past dates/times to be compared with a specific reference date/time (e.g., the present time). The similar date/time extraction unit 10q repeats the process of calculating the sum of the absolute values of the differences obtained by subtracting the value of a specific index at a time "iΔT" back from the past date/time from the value of a specific index at a time "iΔT" back from the reference date/time for each of the randomly acquired past dates/times. The similar date/time extraction unit 10q then extracts past time points with a small degree of deviation (i.e., a large degree of similarity) that meets a specified criterion. Ultimately, the similar date/time extraction unit 10q extracts past dates/times in which the trend of a specific water-related index in a water-related facility is similar to the reference date/time.

The event extraction unit 10r extracts event information for past dates and times extracted by the similar date and time extraction unit 10q from the water quality and event information receiving unit 10a. The event screening unit 10s selects and outputs event information that is highly related to the similarity calculation formula type (specific index, i.e., type of target variable) received by the similarity calculation formula type receiving unit 10c from the many extracted event information. The action procedure information extraction unit 10t searches for action procedure information using the event information selected by the event screening unit 10s as a key, and outputs the results to the action procedure information transmitting unit 10f.

The action procedure information generating unit 10d also aggregates the electronic manual viewing history and annotation information received from the electronic manual viewing history and annotation information receiving unit 10b, the similarity calculation formula type received from the similarity calculation formula type receiving unit 10c, and the event information selected by the event screening unit 10s for each time stamp, and generates the information as action procedure information.

Figure 9 is a diagram showing the data structure in this embodiment. The water quality data storage unit 2 stores water quality data 2f. The water quality data 2f aggregates specific indicators such as the turbidity of the intake well and the water temperature of the receiving well for each date and time.

The event information storage unit 3 stores multiple pieces of event information 3a, multiple pieces of warning information 3b, and multiple pieces of operation/driving information 3c. The event information 3a aggregates a group of warning IDs, operation amounts, and a group of operation/driving information IDs for each date and time. The warning information 3b is composed of a warning ID and warning content. The operation/driving information 3c is composed of a operation/driving information ID and operation/driving information content.

The document viewing information storage unit 6 stores multiple pieces of document viewing information 6a and multiple pieces of driver information 6b. The document viewing information 6a stores who viewed which part of the document (electronic manual) and when. That is, the document viewing information 6a stores the driver ID, document ID, page ID, date and time, annotator ID, and annotation content. The driver information 6b stores the driver ID, driver name, and driver attributes (driver's position, years of experience, etc.). Among the drivers of the equipment, the driver is the one who viewed the electronic manual, and among the drivers, the annotator is the one who added annotations (comments) to the electronic manual.

The document information storage unit 5 stores a plurality of electronic manuals 5a as documents. The electronic manuals 5a store a document ID, a document name, table of contents information, and a document file. The document file stores a plurality of page IDs and page contents. The page contents include documents in text format or image format.

The behavioral procedure information storage unit 8 stores multiple pieces of behavioral procedure information 8a that indicate the driver's past behavioral records (actual operations and driving, pages of electronic manuals that were actually viewed), an event information and involvement level table 8b, and a similar date and time information table 8c for the reference date and time.

The action procedure information 8a stores the action procedure information ID, the driver ID, the date and time when the action procedure was performed, a group of warning IDs issued at that date and time, a group of operation/driving information IDs, and multiple action procedure contents. The action procedure contents store the document ID, page ID, order, specific index, and the text of the action procedure. In other words, the action procedure information 8a as a whole is a combination of event information that occurred in the past and the page of the document viewed by the driver at that date and time. Furthermore, the event information indicates the warning issued at that date and time and the driver's operation/driving at that date and time. The event information/involvement level table 8b stores the involvement level, which is the degree of involvement between the event information and the specific index (the target variable of the similarity calculation formula), in a matrix format for each combination of event information and specific index.

Now, for example, suppose that the correlation between the event "Alarm 1" and the specific indicator "Intake well turbidity" is "1.0." Then, when the specific indicator "Intake well turbidity" is selected, the event screening unit 10s will select the event "Alarm 1" as a priority from among many events.

As another example, suppose that the correlation between the event "Alarm 2" and the specific indicator "Intake well turbidity" is "-1.0." Then, when the specific indicator "Intake well turbidity" is selected, the event screening unit 10s will preferentially exclude the event "Alarm 2" from among many events.

In this way, the action procedure information generating unit 10d of the action procedure information management server 10 normalizes the involvement level to the range of "-1.0 or more and less than 1.0." The action procedure information generating unit 10d determines the involvement level based on the past adoption frequency, etc. (especially the contents of the annotations associated with the electronic manual that was actually viewed in response to the event, etc.).

Furthermore, the action procedure information generating unit 10d may use as learning data a large number of combinations of event information corresponding to electronic manuals that the experienced driver viewed at a past date and to which the experienced driver added positive annotations, and specific indicators that the experienced driver selected at that past date and time, and machine-learn the following formula (2). "F" indicates that formula (2) is a type of function.

[Equation 2]
Involvement level = F (specific index, event information) ... (2)

The similar date and time information table 8c for the reference date and time stores the reference date and time, a period (a period for randomly obtaining past dates and times), and multiple similar dates and times (past dates and times that are highly similar to the reference date and time). The similar dates and times include a rank, a similarity (or a deviation), and the date and time itself.

As is clear from FIG. 9 as a whole, the water quality data 2f, the event information 3a, the document viewing information 6a, the action procedure information 8a, and the similar date and time table 8c for the reference date and time all contain "date and time." In other words, these are associated with each other via the same date and time (time stamp), and can be aggregated using the same date and time as a key.

In the event information/involvement table 8b in FIG. 9, one involvement value is defined for a combination of one specific indicator and one piece of event information. However, more generally, one involvement value may be defined for a combination of one or more specific indicators and one or more pieces of event information.

Figure 10 shows the process flow for generating action procedure information in this embodiment. This process flow is a procedure for accumulating action procedure information by an experienced driver (norms for non-experienced drivers).

In step S1, the water quality data collection server 1 collects water quality data.
In step S2, the water quality monitoring control server 4 collects event information.
In step S3, the driver's information terminal 7 requests the manual information from the document information storage unit 2. The driver's information terminal 7 here is operated by an experienced driver.

In step S4, the document information storage unit 2 transmits the electronic manual information to the driver's information terminal 7.
In step S5, the driver's information terminal 7 receives the similarity calculation formula (type of specific index) and annotations for the electronic manual.
In step S6, the surveillance camera 9 collects driver behavior information. Specifically, the surveillance camera 9 captures video or still images of the driver reading the electronic manual, and recognizes and collects the manual name, page number, and gestures indicating annotations in some cases. The surveillance camera 9 also captures video or still images of the driver operating various devices, and recognizes and collects the types of devices operated and whether or not they are operated. These collected results become the driver behavior information here.

In step S7, the action procedure information management server 10 requests the water quality monitoring control server 4 for event information.
In step S8, the water quality monitoring control server 4 transmits the event information to the action procedure information management server 10.
In step S9, the action procedure information management server 10 requests the monitoring camera 9 for driver action information.
In step S<b>10 , the monitoring camera 9 transmits the driver action information to the action procedure information management server 10 .

In step S11, the action procedure information management server 10 requests water quality data from the water quality data collection server 1.
In step S<b>12 , the water quality data collection server 1 transmits the water quality data to the action procedure information management server 10 .
In step S13, the action procedure information management server 10 requests the driver's information terminal 7 for driver action information.

In step S14, the driver's information terminal 7 collects driver's behavior information, specifically, electronic manual viewing history, similarity calculation formula type, device operation location, and the like.
In step S<b>15 , the driver's information terminal 7 transmits the driver action information to the action procedure information management server 10 .
In step S16, the action procedure information management server 10 screens (sorts) the event information according to the type of similarity calculation formula or annotation information.
In step S17, the action procedure information management server 10 generates action procedure information using the results of the filtering.

As is clear from the above, screening event information means sorting out event information with a high degree of involvement from event information with a low degree of involvement for each specific indicator selected by the driver, and storing the action procedure information corresponding to the event information with a high degree of involvement and the action procedure information corresponding to the event information with a low degree of involvement separately for each specific indicator.

Figure 11 shows the process flow for displaying action procedure information in this embodiment. This process flow is a procedure for displaying action procedure information of an experienced driver to an unskilled driver.

In step S21, the water quality data collection server 1 collects water quality data.
In step S22, the water quality monitoring control server 4 collects event information.
In step S23, the driver's information terminal 7 receives a similarity calculation formula (type of specific index).
In step S24, the action procedure information management server 10 requests the water quality monitoring control server 4 for event information.
In step S25, the water quality monitoring control server 4 transmits the event information to the action procedure information management server 10.

In step S26, the action procedure information management server 10 requests water quality data from the water quality data collection server 1.
In step S27, the water quality data collection server 1 transmits the water quality data to the action procedure information management server 10.
In step S28, the action procedure information management server 10 requests the driver's information terminal 7 for the type of similarity calculation formula (the type of specific index selected by the driver).
In step S29, the driver's information terminal 7 transmits the similarity calculation formula type to the action procedure information management server 10.

In step S30, the similar date and time extraction unit 10q of the action procedure information management server 10 receives water quality data from the water quality and event information receiving unit 10a. For the received water quality data, the similar date and time extraction unit 10q calculates the similarity between a specific index, such as turbidity, at a reference date and time, such as the current time, and a specific index at each past date and time, according to the similarity calculation formula selected (received) by the similarity calculation formula type receiving unit 10c. The similar date and time extraction unit 10q extracts a predetermined number of past dates and times in order of highest similarity.

In step S31, the event extraction unit 10r of the action procedure information management server 10 extracts event information for the past date and time extracted by the similar date and time extraction unit 10q from the water quality and event information receiving unit 10a. Furthermore, the event screening unit 10s of the action procedure information management server 10 selects, from the large amount of extracted event information, events whose degree of involvement with the specific index selected in step S23 is large enough to satisfy a predetermined criterion based on the event information and involvement degree table 8b (Figure 9).

In step S32, the action procedure information extraction unit 10t of the action procedure information management server 10 searches for action procedure information using the event information selected by the event screening unit 10s as a key. The key event information is selected here as information whose degree of involvement with the specific indicator received in step S23 is large enough to satisfy a predetermined criterion.

In step S33, the action procedure information transmission unit 10f displays the action procedure information of the search result on the driver's information terminal 7. The action procedure information displayed at this time corresponds to the event information that has a high degree of involvement with the specific indicator selected by the driver at the reference date and time (see FIG. 12).

As is clear from the above, if an unskilled driver selects, for example, "water intake well turbidity" as a specific indicator in step S23, only the action procedure information corresponding to the event information that is highly related to "water intake well turbidity" is displayed in step S33.

Figure 12 is a screen transition diagram of the driver information terminal 7 in this embodiment. The top row shows the top screen 21, the middle row shows the reason display screen 31, and the bottom row shows the event information display screen 41.

The driver's information terminal 7 accepts the driver's input of a target variable (either the turbidity of the intake well or the pH of the receiving well as a specific indicator) into the similarity calculation formula target variable field 22 on the top screen 21. The driver's information terminal 7 then displays a recommended action procedure 23 on the top screen 21. If the input target variable (specific indicator) is different, the content of the recommended action procedure 23 displayed thereafter will also be different.

The recommended action procedure 23 is a combination of a page of the electronic manual, a condition match degree, and a beginner recommendation statement. The page of the electronic manual is, for example, "Read pages 150 to 165 of document X", which is a page of the electronic manual that the driver viewed at a past date and time. The condition match degree is, for example, "60%". The condition match degree here is the proportion of positive annotations among the annotations added to the electronic manual. Furthermore, the condition match degree may be the similarity or divergence degree itself, or may be a predetermined correction of the similarity or divergence degree (such as a correction using the number of views in a recent predetermined period). The beginner recommendation statement is, for example, "Recommended for beginners!". The driver information terminal 7 displays the current state 24. The current state is, for example, "The current state is an increase in turbidity in the water intake well."

Now, let us say that the driver inputs "water intake well turbidity" in the similarity calculation formula target variable field 22, and then selects the record (row) in the bold frame in the recommended action procedure 23. Then, the driver's information terminal 7 displays the reason display screen 31.

The driver's information terminal 7 displays an index transition graph 32 on the reason display screen 31. The index transition graph 32 displays the transition of the intake well turbidity over a period of four sampling cycles (reference period) ending at the reference date and time (current date and time) (thick broken line). The index transition graph 32 also displays four intake well turbidities from a period of four sampling cycles in the past that have a high similarity to the intake well turbidity in the reference period (four broken lines other than the thick line). Furthermore, these four broken lines are further extended in the future direction by three sampling cycles in the past. Note that the "4" on the four broken lines corresponds to four past dates and times that have a high similarity (small deviation) that meets the specified criteria, and is not directly related to the "4" for four cycles.

The driver's information terminal 7 displays an operation amount transition graph 33 on the reason display screen 31. The operation amount transition graph 33 displays the amount of flocculant administered in the reference period (thick broken lines). The operation amount transition graph 33 also displays the amount of flocculant administered in a period of seven sampling periods before and after the sampling period, with each of the four past dates and times with the greatest similarity as its center point (four broken lines other than the thick lines).

The driver's information terminal 7 displays the past example list 34 on the reason display screen 31. In the type column of the past example list 34, "Reference" means the reference date and time (current date and time), and "Similar ○" means a past date and time. The date and time column displays the specific month, date, time, and minute of the reference date and time and the past date and time. The deviation column displays the deviation degree. The driver column displays the name or ID of the driver at the past date and time. The content column displays the page of the electronic manual that the driver viewed at the past date and time. The driver's information terminal 7 displays a title 35 in a prominent location on the reason display screen 31. The title 35 includes a specific character string that is written on a certain page of the electronic manual. In the example of FIG. 12, "Increase the coagulant injection rate by 5 mg/L" is a measure against the increase in turbidity of the water intake well.

When the driver selects a record with a bold frame from the past example list 34, the driver information terminal 7 displays the event information display screen 41. When the driver presses the "Back" button 36, the driver information terminal 7 displays the top screen 21. For the sake of explanation, let us assume that the driver has selected a record that includes the type "Similar 1".

The driver information terminal 7 displays event information 42 on the event information display screen 41. The event information 42 is a combination of the time series of the involvement level, warning, and operation/driving at the past date and time corresponding to "Similar 1". The warning is a warning that was actually issued at the past date and time. The operation/driving is an operation/driving that the driver "B" actually performed at the past date and time.

In the first line, the "turbidity rise" alarm corresponds to a contribution level of "1.0." In the fourth line, the "coagulant injection" operation also corresponds to a contribution level of "1.0." This shows that the change in intake well turbidity is sufficiently involved in the event information, the "turbidity rise" alarm, and the "coagulant injection" operation, so that an operator who encounters "increased intake well turbidity" at the reference date and time should pay attention to lines 1 and 4.

The degree of involvement "-1.0" on the second line indicates that the change in the turbidity of the intake well is not involved in the "pump abnormality" alarm at all. The degree of involvement "0.0" on the third line indicates that the change in the turbidity of the intake well is either involved or not involved in the "pump restart" operation. It is desirable for the driver's information terminal 7 to hide records with such a low degree of involvement (for the purpose of explanation, FIG. 12 shows an example in which they are displayed). If the event screening unit 10s did not exist, the number of records in the event information 42 would be enormous.

The driver information terminal 7 may predict the progress of a specific indicator after a reference date and time, and based on the prediction, may display a warning 43 that is likely to be issued in the future and an operation/driving action 44 that is likely to be performed in the future.

When the driver presses the "Back" button 45, the driver information terminal 7 displays the reason display screen 21.

(Effects of this embodiment)
(1) The action procedure information management device can limit event information of a past time point in which a specific index is similar to a reference time point to information related to the specific index.
(2) The action procedure information management device allows the driver to select a specific indicator.
(3) The action procedure information management device can aggregate event information and document browsing information having the same date and time to generate action procedure information.
(4) The behavior procedure information management device can determine the degree of association between a specific indicator and event information based on annotations in the electronic manual.

(5) The action procedure information management device can be applied to a water purification plant or a sewage treatment plant.
(6) The action procedure information management device can treat an alarm that is likely to be manifested to the outside or the contents of the driver's operation/driving as event information.
(7) The action procedure information management device can define the degree of association between a specific indicator that is of high interest to the driver and event information.

The present invention is not limited to the above-described embodiments, but includes various modified examples. For example, the above-described embodiments have been described in detail to make the present invention easier to understand, and are not necessarily limited to those having all of the configurations described. It is also possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. It is also possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

1. Water quality data collection server (sensor data collection server)
2 Water quality data storage unit 3 Event information storage unit 4 Water quality monitoring and control server (plant monitoring and control server)
5 Document information storage unit 6 Document viewing information storage unit 7 Driver information terminal 8 Action procedure information storage unit 10 Action procedure information management server (action procedure information management device)
10c Similarity calculation formula type receiving unit 10d Action procedure information generating unit 10q Similar date and time extracting unit 10r Event extracting unit 10s Event screening unit 11 Network

Claims (7)

a similar date and time extraction unit that extracts one or more past dates and times in which a change in a specific index indicating water quality related to the plant is similar to a reference date and time;
an event extraction unit that extracts a plurality of pieces of event information that are a plurality of warnings issued regarding the specific index at each of the extracted one or a plurality of past dates and times or a plurality of pieces of event information that are contents of operations or driving by an operator of the plant;
an event screening unit that selects the extracted plurality of pieces of event information based on a degree of association between the specific index and each of the plurality of pieces of event information;
An action procedure information management device comprising: a similarity calculation formula type receiving unit that receives the specific index selected by an operator of the plant;
2. The action procedure information management device according to claim 1, wherein: a behavior procedure information generating unit that collects the event information and document browsing information, and generates the collected event information and document browsing information, which have the same date and time, as behavior procedure information to be displayed to the driver;
3. The action procedure information management device according to claim 2, wherein: The action procedure information generation unit
determining the degree of involvement based on the contents of annotations added to an electronic manual actually viewed by the driver in response to the event information;
4. The action procedure information management device according to claim 3, wherein: The plant comprises:
Including a water purification plant or a sewage treatment plant;
5. The action procedure information management device according to claim 4, The degree of involvement is
A combination of the specific indicator and the event information is defined;
6. The action procedure information management device according to claim 5 , extracting one or more past dates and times at which a change in a specific indicator indicating water quality related to the plant is similar to a reference date and time;
extracting a plurality of pieces of event information, which are a plurality of warnings issued with respect to the specific index or a plurality of pieces of event information that are the contents of operations or driving by an operator of the plant at each of the extracted one or a plurality of past dates and times ;
an action procedure information management device that selects the extracted plurality of event information based on a degree of association between the specific indicator and each of the plurality of event information;
a sensor data collection server that collects sensor data related to the plant;
a plant monitoring and control server that monitors and controls the plant;
a driver's information terminal that receives the specific indicator from a driver of the plant and displays an action procedure corresponding to the event information to the driver;
An action procedure information management system comprising:

Images