JPH02226500A - Disaster prevention supervising device for precise environmental space - Google Patents

Abstract

PURPOSE:To improve the safety of an evacuation by forecasting a danger based on gas leakage on an initial stage before an alarm level, logically judging the level by means of a hydrogen gas, and displaying the evacuation direction on a communication panel. CONSTITUTION:A previous forecast inferring means 13 forecasts the gas leakage on the initial stage before the alarm level, and the measures against the gas leakage, etc., are instructed on the initial stage based on the forecast. A hydrogen gas logical judging part 13a logically judges the hydrogen gas level by means of three pieces of detected density information, estimates an approximate cause, and instructs measures. In addition, an intention determination supporting means 14 estimates the dangerous degree and the magnitude of damage from sensor information and the initialized information after the initial measures, and supports intention determination whether a processing should be continued in the precise environmental space or an operator should be evacuated. Further an evacuation route inferring means 15 controls a communication panel 33 according to an evacuation pattern, and displays the evacuation route. Thus the evacuation instruction can be suitably issued, and the safety and productivity can be improved.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a disaster prevention monitoring device applied to a precision environment space such as a clean room of a semiconductor factory, and more specifically, to a disaster prevention monitoring device applied to a precision environment space such as a clean room of a semiconductor factory. This invention relates to a disaster prevention monitoring device for a precision environmental space that is capable of predicting the extent of a disaster caused by a gas leak at an early stage and taking appropriate measures when a gas leak occurs.

[Conventional technology]

Recent precision processing technology is supported by many advanced support technologies, as seen in research and production facilities for electronics, typified by semiconductors, and biotechnology. One of the supporting technologies is the creation of a so-called clean room, which turns the space of a research or production facility into an ultra-precision working environment.

On the other hand, in semiconductor factories and various laboratories that have clean rooms, when producing semiconductor devices, they use highly toxic, flammable, self-combustible, or corrosive dangerous gases such as arsine, hydrogen, monosilane, and hydrogen chloride. Since many kinds of chemicals are used in large quantities, if even a small amount of these gases or chemicals leaks into the clean room, this can cause gas poisoning of facility workers, fire, explosion, and other disasters. Therefore, when a gas leak occurs, it is important to discover it early and take appropriate emergency measures, which will prevent disasters and improve the safety of the clean room.

Conventionally, disaster prevention monitoring systems for semiconductor factory clean rooms, etc. have installed various sensors depending on the material gas used at necessary locations in the clean room, and these sensors detect gas leaks that exceed a set value. When this occurred, an alarm was raised and local countermeasures were taken, such as shutting off the gas.

[Problem to be solved by the invention] By the way, in the conventional disaster prevention monitoring system as described above,
Since the response method is determined based on limited information from the sensor such as where and what kind of gas leaked, the response measures when an accident occurs will vary depending on the skill level of the operator. If there is a delay in responding, there is a risk that it will develop into a major secondary disaster. In addition, even if it is possible to respond to the gas leak by simply detecting a gas leak and issuing an alarm, how will the situation in other rooms related to this change and how should we deal with it? There is a problem that appropriate countermeasures cannot be taken at the initial stage of a gas leak, and furthermore, if the gas supply to normal production equipment is cut off due to sensor malfunction, production activities will be hindered.
This results in huge losses.

In addition, conventional disaster prevention monitoring systems focus on early detection of gas leaks and emergency measures such as shutting off the gas, but do not place emphasis on the evacuation of workers working in various facilities inside sealed clean rooms. Currently, most systems rely on the instructions of site managers or the judgment of individual workers, which can lead to confusion in evacuation in the event of a disaster.

The present invention has been made in order to solve the above-mentioned problems, and it is possible to predict the extent of a disaster at an early stage, to take appropriate disaster prevention measures, to make evacuation instructions accurate, and to improve safety. An object of the present invention is to provide a disaster prevention monitoring device for a precision environmental space, which is designed to improve production efficiency.

[Means to solve the problem]

The disaster prevention monitoring device for a precision environmental space according to the present invention has cause-and-effect relationships configured in a knowledge display format using specialized knowledge according to desired rules, initial setting information on gases used in the precision environment space, equipment, etc., and sensor information. and a memory means for storing the information as a knowledge base, and based on sensor information other than hydrogen gas obtained from the memory means, predicting the transition of change from the initial detection information exceeding the detection limit and instructing initial countermeasures. For hydrogen gas, inside the production equipment.

Pre-predictive inference means that makes logical judgments on gas concentration information under the floor and ceiling of precise environmental spaces, infers the cause, and instructs initial countermeasures, and stops production activities based on sensor information and initial setting information after initial countermeasures. A decision-making support means that supports decision-making on countermeasures for an evacuation decision plan, and countermeasure instructions determined based on the support information of the advance prediction reasoning means and the decision-making support means, and appropriate evacuation according to the location and size of the disaster. an evacuation route inference means for inferring and instructing a direction; and a central monitoring room arranged in the precise environmental space, controlled according to the evacuation pattern set by the evacuation route inference means and displaying the evacuation direction; It is equipped with a communication panel for communicating with the system, and a communication panel for communicating with the system.

[For production]

The advance prediction inference means predicts the danger from the gas leak in the initial stage before the alarm is issued, and accordingly instructs countermeasures against the gas leak in the early stage. For hydrogen gas, the system makes a logical judgment based on the detected concentration information from three locations, estimates the approximate cause, and instructs countermeasures.

In addition, the decision-making support means estimates the degree of risk and damage from the sensor information and initial setting information after initial countermeasures, and supports decision-making on whether to remain in the precise environment space and continue processing or to evacuate. .

Then, the evacuation route inference means infers and instructs an appropriate evacuation direction depending on the location and size of the disaster, and displays the evacuation route by controlling the communication panel in accordance with the determined evacuation pattern.

Therefore, according to the present invention, appropriate disaster prevention measures can be taken at an early stage, evacuation instructions can be given accurately, and safety and production efficiency can be improved. Furthermore, by making logical judgments using hydrogen gas, the reliability of the detected information is increased and countermeasures are easier to take.

〔Example〕

Embodiments of the present invention will be described below based on the drawings.

FIG. 1 is an overall configuration diagram showing an example in which a disaster prevention monitoring device according to the present invention is applied to a clean room of a semiconductor factory.

In the figure, the disaster prevention monitoring system IO is installed in a central monitoring room outside the clean room, and includes a central processing unit (hereinafter referred to as CPU) II that controls the entire system, and specialized knowledge such as IF-THEH. a memory 12 that stores, as a knowledge base, cause-and-effect relationships configured in a knowledge expression format, initial setting information representing characteristics of gases/chemicals used in semiconductor production and equipment status, and equipment/sensor information;
Based on the sensor information obtained from the memory 12 through the PU 11, the transition of change is predicted from the initial detection information that exceeds the detection limit and initial countermeasures are instructed, and for hydrogen gas, the gas concentration in the production equipment is , clean room 1
Pre-prediction inference means 13 includes a logic judgment unit 13a that makes logical judgments on the underfloor gas concentration and ceiling gas concentration information to estimate the cause, and production information for the operator based on the sensor information and initial setting information after initial measures. A decision-making support means 14 for quickly and appropriately supporting important decisions such as stopping activities or deciding to evacuate, and an evacuation route for inferring and instructing an appropriate evacuation direction depending on the location and size of a disaster. An inference means 15, a data transmission control device 16 that controls multiplex transmission of various data between the CPU I and the clean room system, and a peripheral device 18 connected via a channel device 17 that exchanges data with the CPU 1l. It consists of

The peripheral device 18 is a CR equipped with a multi-window display function.
T display device 18a, keyboard 18b for inputting data such as programs and initial setting information necessary for inference, printer 18c for printing display data, program list, etc.
, a disk device 18d for storing programs, display data, etc., and each device is connected to a channel device 17.

In FIG. 1, inside a clean room 1, there is a semiconductor device production equipment 20, an emergency exhaust fan 21 that urgently exhausts leaked gas etc. to the outside of the clean room 1, an exhaust damper 22 that is opened during emergency exhaust, and an exhaust damper 22 that cleans the inside of the clean room 1. An air conditioner 23 for air conditioning and a gas cutoff valve 24 for the production equipment 20 are installed. It is connected to the output device 26 and outputs command signals sent from the system 10 in the central monitoring room to each equipment, and status signals of each equipment are taken into the data transmission input/output device 26. It is adapted to be transmitted to system 10. Further, an emergency shutoff valve 27 that shuts off the supply source of the gas supply line is connected to the data transmission input/output device 26, and this emergency shutoff valve 27 has a similar configuration to the above-mentioned equipment.

The data transmission device 28 transmits sensor information to the system 10 side in real time, and is connected to the data transmission control device 16 of the system 10 via a signal wire 25, and on the input side of the data transmission device 2, Various gas detection sensors 29, temperature sensors 30, pressure sensors 31, earthquake sensors 32, etc. are connected.

The communication panel 33 provided in the clean room 1 is used to closely communicate information between the central monitoring room and the site and to display direction during evacuation, and is composed of an information transmission board 33a and an evacuation direction board 33b.

The information transmission panel 33a is capable of displaying the occurrence of an abnormal situation, its disaster mode, current situation, and predicted values on a plasma screen or the like, and is also capable of emitting sound using a voice synthesizer. It is equipped with an intercom function that enables direct voice communication between the field and the central control room.

Further, the evacuation direction indicator board 33b is installed near the ceiling or floor of a partition in a clean room, and on the wall, ceiling, or floor in a ship cabin, hallway, etc.

The communication panel 33 as described above is connected to the system 1 via the data transmission input/output device 34 and the signal line 35.
The system 10 is connected to the data transmission control device 16 of the system 10.
Based on the decision result of countermeasure instructions obtained by the side and based on this, <IIi!
Data indicating the m direction is sent to the communication panel 33. Further, the information transmission board 33b of the communication panel 33 is directly connected to a telephone set 36 installed in the central monitoring room so that operators can directly communicate between the site and the central monitoring room.

Next, the operation of the disaster prevention monitoring system according to this embodiment configured as described above will be explained with reference to FIGS. 2 to 6.

FIG. 2 is a flowchart showing the steps of prediction/judgment support and evacuation route inference, and is an example of a case where a gas leak occurs within the clean rule.

When the system IO program starts, first, in step 100, information on each sensor is transmitted to the CP through the data transmission input/output device 26 and the data transmission control device 16.
The leakage detection concentration level of each sensor is sequentially fetched into LIII, and the advance prediction inference means 13 determines whether the leakage detection concentration level of each sensor exceeds the detection limit (step 101). Then, in the next step 102, data from the time when the detection limit is exceeded is stored at predetermined time intervals (for example, every minute), and the transition of the concentration gradient is predicted and calculated from this data. In the next step 103, it is determined from the transition of the concentration gradient whether it is actual leakage or noise. If it is determined that there is a possibility of noise, the process proceeds to step 104, where an instruction to check the spot where a leak has been detected is displayed on the display device 18a in the central monitoring room, and a message to that effect is displayed on the communication panel 33 in the clean room. send. Upon receiving this information, clean room workers, etc., check the status of the sensor and its surroundings at the specified location.
I will

On the other hand, if it is determined that there is a possibility of leakage in the stem knob 103, the process proceeds to step 105, where it is determined whether it is hydrogen gas information, and if it is determined that the detected information is something other than hydrogen gas, the process proceeds to step 106. Further, when it is determined that it is hydrogen gas information, the process proceeds to step 107, where a logical judgment is made and the cause is estimated.

FIG. 3 is a schematic cross-sectional view of a clean room showing the arrangement of hydrogen gas sensors in the clean room 1 when logical judgments are made based on hydrogen gas information. A production device 20 is installed in the upper chamber 1a of the clean room 1 which is divided into upper and lower parts by a grating 40, and the ceiling of the upper chamber 1a and the royal room 1b are communicated with each other by a duct 41 and an air circulation fan 42, so that the air in the upper chamber 1a is is designed to circulate in the direction of the arrow. A first hydrogen gas sensor 29A is installed inside the production equipment 20, and a second hydrogen gas sensor 29B is installed under the grating (under the floor) near the production equipment 20.
Third hydrogen gas sensors 29C are installed on the ceiling of the upper chamber 1a.

Therefore, based on the gas concentration information inside the production equipment 20, which is detected by each of the hydrogen gas sensors 29A to 29C and obtained by predicting its change transition, the gas concentration information under the grating, and the gas concentration information on the ceiling, A logical judgment of the estimation is performed (step 107).

Here, examples of logical judgment using hydrogen gas will be listed.

碢± Case■ ・Hydrogen concentration inside the production equipment -800ppm ・Hydrogen concentration under the floor of the production equipment (under the grating -150ppm) ・Hydrogen concentration on the ceiling of the production equipment room -20ppm The logical judgment based on this is: ・Production There is a very high possibility that there is a leak inside the device.

・As a countermeasure, close the hydrogen supply valve in the production equipment and check the leak point using a leak checker (portable gas detector).

Case■ ・Hydrogen concentration inside the production equipment Oppm ・Hydrogen concentration under the floor of the production equipment -150ppm ・Hydrogen concentration at the ceiling of the production equipment -20ppm The logical judgment based on this is: ・There is a possibility that there is a leak outside the production equipment expensive.

・As a countermeasure, shut off the valve in front of the production equipment on the hydrogen gas piping that supplies the production equipment, and use a leak checker to check the leak location.

Case■ ・Hydrogen concentration inside the production equipment Oppm ・Hydrogen concentration under the floor of the production equipment -20ppm ・Hydrogen concentration at the ceiling of the production equipment -150ppm The logical judgment based on this is: ・Hydrogen gas is present somewhere outside the production equipment. There is a high possibility that there is a leak, but the fact that the concentration under the floor has not been detected is 2/15L of the ceiling, which raises the following question.

■ Was it leaking near the ceiling?

(Normally, piping should not be run near the ceiling, but under the floor.) ■ Air flow (usually downward flow) is weak and hydrogen gas is rising.

■ Hydrogen gas (with strong directionality) is being sprayed toward the ceiling.

■ An abnormality in the detector either under the floor or on the ceiling.

breakdown.

As a countermeasure, check the above items 1 to 2 with a leak checker.

The reasons for using hydrogen gas for logical judgments are: (a) To make logical judgments, it is best if a large number of detectors for the leaked gas are installed, but many types of gas are used in a clean room. Therefore, for gases other than hydrogen, one gas detector is normally installed for each production device in which the gas is used, and logical judgment is difficult in this situation.

(b) The sensitivity and reliability of detectors for special material gases such as silane gas and arsine gas are much higher than those for hydrogen gas, and are suitable for logical judgment.

(C) There are many points at which hydrogen is used (use points) in a clean room, and therefore many hydrogen detectors are installed. Further, the above-mentioned special material gas is often used as a mixture in hydrogen (for example, silane gas volume ratio 1%, hydrogen 99%), and detection can sometimes be performed with hydrogen instead of the special material gas.

(d) Hydrogen is the lightest of gases and is less adsorbed, so it has good diffusivity after leaking into the room, and there is little variation in detected concentration depending on the location of the detector.

It is from.

When the determination in step 105 is rYESJ and the process in step 107 is completed, the process proceeds to step 106, and from the transition of the concentration gradient, what type of gas (or chemical) leaks and in what manner is determined. IF-T when leaking
The pre-prediction inference means 13 makes an estimate based on the HEN rule, and performs support processing for initial countermeasures. At the earliest stage when it is determined that there is a possibility of leakage, the process proceeds to step 108, and the probable cause is determined in order to take early countermeasures using feedforward control. Information such as "What ppm will be in a few minutes" and corresponding countermeasure instructions are displayed on the display device 18a in the central monitoring room, and are also displayed on the communication panel 33 in the clean room, and the information is also output by voice. Based on this information, the operator in the clean room takes initial measures such as stopping the gas supply line, checking for leaks, and tightening pipe connections. Then, use the intercom to notify the central monitoring room.

If it is determined that the concentration gradient of the gas leak is steep and immediate evacuation is necessary (step 108A), step 11
Skip to step 4 and move on to the step of inferring an evacuation route.

In the next step 109, the processing after the initial measures is evaluated,
In the next step 111, it is determined whether the gas leakage concentration has decreased. Although initial measures have been taken, if it is determined that the gas leakage concentration has not decreased and is rising, we will consider whether to stop the production equipment 20 or decide to evacuate.
Further, the decision support means 14 executes support and processing for full-scale countermeasures. Steps IIIA to 113 correspond to this.

That is, the current degree of danger is calculated from the properties of gases and chemicals, sensor measurements, and the state of the production equipment, and the countermeasures to be taken are displayed on the screen of the display device 18a based on the score. For example, should production be stopped, and if so, when and how? And if not, how will you ensure safety?

It also gives instructions on whether to evacuate, whether some or all of the people should evacuate, and what to do before evacuating.

On the other hand, the magnitude of damage is estimated from the use of hazardous materials in the clean room and the situation of workers, and this is used as support information for determining evacuation measures. For example, the number of wafers that are lost or damaged, or the damage to production equipment.

FIGS. 4(a), 4(b), and 4(C) are explanatory diagrams showing an example of a case where the degree of risk is estimated and support information for determining countermeasures is displayed on the screen.

Furthermore, the risk evaluation rules by the decision-making support means 14 are set as follows, for example.

a. Gas cylinder (10 points) Capacity 411 or more -・-・----10
Point 10j2〃 Less than 47ffi----1・-・-
7 points 31// 101〃 ・---------・・・−・ 5
Less than 3 points -・------3 points b6 remaining amount
(10 points) Remaining capacity 70% or more ・---111・-I 0
Points 40% or more but less than 70% --- 7 points 40
Less than %...---5 points C, gas properties (out of 20 points) (1) Toxicity 20 points x A A: TLV (tolerable concentration) less than 1 ppm----A
= 11 or more 5 -...A = 0.5 5 or more A = 0.3 (2) Self-flammability 10 points Addition (3) Corrosion 3 points Addition (4) Flammability 5 points Addition (5) Explosion (6) Addition of 5 points for combustibility (6) Addition of 5 points for combustibility With reference to the support information obtained as described above, the measures to be actually taken are determined and displayed on the communication panel 33 in the clean room.

In addition, with the decision on the above measures, actions that would have a significant impact on production, such as shutting down the supply chain and cutting off exhaust and gas supply lines from the supply source, will be subject to approval by the person in charge. However, if the instruction is not given for a certain period of time (for example, 5 minutes) after the countermeasure display by the system 10, the above action is forcibly performed by remote control from the central monitoring room.

If it is determined that evacuation is necessary based on the support information provided by the advance prediction inference means 13 and the decision support means 14 (step 113A), the process proceeds to step 114, and the evacuation route inference means 15 infers the optimal evacuation method. Execute by.

That is, in step 11, the shortest evacuation route is inferred from sensor information and initial setting information such as the location of the location (room) where the trouble occurred in the clean room, and the location and number of doors.
4) This determines the evacuation route.Step 115) displays this on the display device 18a in the central monitoring room.

At the same time, the evacuation direction board 33b inside the clean room
data transmission control device 16 and data transmission input/output device 3
Output the direction instruction data through the evacuation direction indication board 3.
3b is displayed as an arrow to instruct the evacuation direction to the evacuating workers (step 116). Therefore, workers and the like can evacuate smoothly and safely by moving in the direction indicated by the arrow.

Figure 5 shows an example of the evacuation route monitor screen displayed on the display device 18a in the central monitoring room. The evacuation route for each room will be displayed pointing toward the clean room exit as shown by the arrow.

Further, FIG. 6 shows an example in which the state of the clean room at the time of gas leakage is displayed on the screen of the display device 18a using a multi-window display method.

In the figure, the hatched area is the leakage zone, and the color is displayed. Further, the portion corresponding to the sensor 29 that has detected the gas leak is flickered. Further, in the area where the gas leakage trend graph is displayed, a graph showing the change in gas leakage concentration is enlarged and displayed.

In addition, in the advance prediction inference means 13, I
As a method for predicting and inferring leakage using the P-THEN rule, the following method is used.

For example, in the case of a toxic gas, the way it leaks can be deduced in the same way as barley.

Table I As countermeasures against this, if the leaked gas is toxic, Case ■ If it is a leak [Countermeasures] - Close the emergency shutoff valve.

・Evacuate everyone.

・Switch to emergency exhaust mode.

・Wear an oxygen mask.

- Be careful of fire when using silane.

Case■　″′If there is a gradual leak at the office [Countermeasures] ・Close the emergency shutoff valve.

・Evacuate the people working on the leaked process.

・Switch to emergency exhaust mode.

・Wear an oxygen mask.

Case■　If low concentration or zadama soil causes sputum leakage [Countermeasures] ・Stop the gas supply to the production equipment in the vicinity of the detection.

・Evacuate people near the leak.

・Wear an oxygen mask.

・Check for gas leaks.

Case ■ 2y1 Wandering - If there is an omission [Countermeasures] ・Check for gas leaks.

- Carefully monitor whether the concentration increases again.

If immediate evacuation is required at the predictive inference stage as described above, the process immediately moves to the evacuation route inference step, where a predetermined shortest evacuation route is inferred.

In this embodiment as described above, danger is predicted from the initial gas leak, so workers can be informed of the situation before it reaches the alarm level, and this can be useful for earlier countermeasures. Can be done.

Furthermore, by logically determining hydrogen gas leakage based on information from multiple locations, it is possible to determine whether the leak is a false alarm due to noise, detection system abnormality, etc., and improve the reliability of the detected information. Furthermore, since it is possible to infer where the leak is occurring (for example, there is a high possibility that the leak is occurring in production equipment), it is easy to take countermeasures.

In addition, since the decision support means 14 supports important decisions such as stopping production activities or deciding to evacuate, decisions about whether to remain in the clean room and continue treatment or whether to evacuate can be made quickly and rationally. It is possible,
Accordingly, it is possible to prevent a decrease in the operating rate due to erroneous stoppage of production activities, reduce damage to production equipment, and improve production efficiency.

Furthermore, the appropriate evacuation direction is inferred and instructed according to the location and size of the disaster, and the evacuation direction is displayed on the communication panel 33 in the clean room, so that workers can evacuate smoothly and ensure evacuation safety. In addition to improving safety, it can also prevent man-made disasters and other disasters.

In the above example, the leakage of silane gas in semiconductor production was described, but the leakage is not limited to this, and the same applies to other gases such as arsine, ciborane, chlorine, hydrogen chloride, and ammonia. In addition, it is not limited to semiconductor production.

〔Effect of the invention〕

As explained above, according to the present invention, dangers are predicted from gas leaks at an early stage before the flu level is reached, so countermeasures against disasters can be taken at an early stage. In addition to being able to minimize and prevent this, the reliability of detection information is increased by making logical decisions based on hydrogen gas, and the need for sensors to detect special material gases such as silane can be reduced.

In addition, in the event of a gas leak after initial countermeasures have been taken, the system estimates the degree of risk and damage caused by sensor information and initial setting information and instructs countermeasures, such as stopping production activities or evacuation. This makes it possible for operators and the like to quickly and easily make decisions on important decisions, such as decisions, and to make accurate judgments.

Furthermore, the appropriate evacuation direction can be inferred and instructed according to the location and size of the disaster, and the evacuation direction can be displayed on the communication panel, making evacuation actions smoother for workers and improving evacuation safety. There is an effect.

[Brief explanation of the drawing]

FIG. 1 is an overall configuration diagram showing an example of a disaster prevention monitoring device according to the present invention. FIG. 2 is a flowchart for explaining the operation in this embodiment. FIG. 3 is an explanatory diagram showing the hydrogen gas sensor and arrangement state in this embodiment. FIGS. 4(a) to 4(C) are explanatory diagrams showing an example of displaying support information for determining countermeasures on a screen in this embodiment. FIG. 5 is an explanatory diagram showing an example of the evacuation route monitor screen in this embodiment. FIG. 6 is an explanatory diagram showing an example of a multi-window display of the state of the clean room at the time of gas leakage in this embodiment. [Explanation of symbols of main parts] 1...Clean room (Seiyu environmental space) IO...Disaster prevention monitoring system 11...Central processing unit 12...Memory 13...Advance prediction inference means 13a. ... Hydrogen gas logic side IVr section 14 ... Decision support means 15 ... Evacuation route inference means 16 ... Data transmission control device 18 ... Peripheral device 18a ... Display device 20 ・ 23 ・ 26° 28 ・ 29° 33 ・ 36 ・ ... Production equipment ... Air conditioner 34 ... Data transmission input/output device ... Data transmission device 30. 31 ... Sensor ... Communication panel ... Telephone device. Figure b Figure + a) Figure Figure (b) Figure (C)

Claims (1)

[Claims] (1) A memory means for storing, as a knowledge base, cause-and-effect relationships configured with specialized knowledge in a knowledge display format according to desired rules, initial setting information for gases used in precision environmental spaces, equipment, etc., and sensor information; Based on the sensor information other than hydrogen gas obtained from the memory means, the transition of change is predicted from the initial detection information exceeding the detection limit, and initial countermeasures are instructed. Pre-predictive inference means that makes logical judgments on gas concentration information under the floor and ceiling of precise environmental spaces, infers the cause, and instructs initial countermeasures, and stops production activities based on sensor information and initial setting information after initial countermeasures. A decision-making support means that supports decision-making on countermeasures for an evacuation decision plan, and countermeasure instructions determined based on the support information of the advance prediction reasoning means and the decision-making support means, and appropriate evacuation according to the location and size of the disaster. an evacuation route inference means for inferring and instructing a direction; and a central monitoring room arranged in the precise environmental space, controlled according to the evacuation pattern set by the evacuation route inference means and displaying the evacuation direction; A disaster prevention monitoring device for a precision environmental space, which is equipped with a communication panel for communicating with the .