Anomaly detection device, anomaly detection system, and anomaly detection method

Abstract

The present invention provides an abnormal condition detection device capable of appropriately detecting the occurrence or risk of fire in vehicles loaded on a car carrier. [Solution] The abnormal state detection device 100 includes a navigation information acquisition unit 111 that acquires location information, and a camera information acquisition unit 112 that acquires camera information that images the vehicle and the space around the vehicle. The abnormal state detection device 100 also includes a temperature information acquisition unit 113 that acquires temperature information and temperature change information per unit time. The abnormal state detection device 100 also includes an abnormal state determination unit 117 that determines whether a fire has occurred or is likely to occur based on assumed temperature information based on location information, temperature information, and temperature change information. If the abnormal state detection device 100 determines that a fire has occurred or is likely to occur, it includes a notification unit 118 that notifies at least one of the rotating light, alarm, user terminal, and administrator terminal that a fire has occurred or is likely to occur.

Description

Technical Field

[0001] The present invention relates to an abnormal state detection device, an abnormal state detection system, and an abnormal state detection method.

Background Art

[0002] Conventionally, technologies related to fire detection in automobile carriers have been proposed. Patent Document 1 discloses a fire detection notification system for an automobile carrier. The fire detection notification system disclosed in Patent Document 1 measures the temperature of the surface of an automobile mounted on a ship using a thermopile sensor, and detects the presence or absence of a fire from the measured temperature.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, in the fire detection notification system disclosed in Patent Document 1, since vehicles are usually loaded at as high a density as possible in an automobile carrier, the layout of the loaded vehicles often changes depending on the types and numbers of vehicles to be loaded. Therefore, with the conventional configuration, it is necessary to change the layout, orientation, etc. of the thermopile sensor unit each time according to the type and layout of the loaded vehicles, and it may be difficult to obtain high accuracy in detecting the occurrence of a fire depending on the layout, orientation, etc.

[0005] The present invention has been made in view of the problems of such conventional technologies. An object of the present invention is to provide an abnormal state detection device capable of appropriately detecting the occurrence of a fire or the risk of a fire occurring in a vehicle loaded on an automobile carrier.

Means for Solving the Problems

[0006] An abnormal condition detection device according to an aspect of the present invention is an abnormal condition detection device for detecting a fire occurring in a vehicle loaded on a car carrier, comprising: a navigation information acquisition unit that acquires position information of the car carrier; a camera information acquisition unit that acquires camera information which includes image data captured by a camera of the vehicle and the space around the vehicle; a temperature information acquisition unit that acquires temperature information and temperature change information per unit time of the vehicle and the space around the vehicle based on the camera information; a smoke generation detection unit that detects whether or not smoke is being generated in the vehicle and the space around the vehicle based on the camera information; and a smoke generation detection unit that detects whether or not smoke is being generated in the vehicle and the space around the vehicle based on the camera information The system includes a flame detection unit that detects whether or not flames are present, an abnormal condition determination unit that determines whether or not a fire has occurred or is likely to occur in the vehicle and the surrounding space based on assumed temperature information based on location information, temperature information, temperature change information, smoke generation status by a smoke detection unit, and flame generation status by a flame detection unit, and a notification unit that, if the abnormal condition determination unit determines that a fire has occurred or is likely to occur, notifies at least one of the rotating lights, alarms, user terminals, and administrator terminals installed on the car carrier that a fire has occurred or is likely to occur.

[0007] Another aspect of the present invention provides an abnormal condition detection system comprising the abnormal condition detection device described above, a camera for imaging vehicles loaded on a car carrier, and at least one of a rotating light, an alarm, a user terminal, and an administrator terminal.

[0008] Another aspect of the present invention relates to an abnormal condition detection method, which is performed by a computer and detects a fire occurring in a vehicle loaded on a car carrier, and includes acquiring location information of the car carrier, acquiring camera information which includes image data of the vehicle and the surrounding space captured by a camera, acquiring temperature information of the vehicle and the surrounding space and temperature change information per unit time based on the camera information, detecting smoke generation status indicating whether or not smoke is being generated in the vehicle and the surrounding space based on the camera information, detecting flame generation status indicating whether or not flames are being generated in the vehicle and the surrounding space based on the camera information, determining whether or not a fire is occurring in the vehicle and the surrounding space or whether there is a risk of a fire occurring, and if it is determined that a fire is occurring or there is a risk of a fire occurring, notifying at least one of the rotating lights, alarms, user terminals, and administrator terminals installed on the car carrier that a fire is occurring or there is a risk of a fire occurring. [Effects of the Invention]

[0009] According to the present invention, it is possible to provide an abnormal condition detection device that can appropriately detect the occurrence of a fire or the risk of a fire occurring in a vehicle loaded on a car carrier. [Brief explanation of the drawing]

[0010] [Figure 1] This figure shows the schematic configuration of the abnormal condition detection system according to this embodiment. [Figure 2] This figure shows the configuration of the abnormal condition detection device according to this embodiment. [Figure 3] This block diagram shows the functional configuration of the abnormal condition detection device according to this embodiment. [Figure 4A] This figure shows an example of image data applied to the abnormal condition detection system according to this embodiment. [Figure 4B] This figure shows an example of image data applied to the abnormal condition detection system according to this embodiment. [Figure 4C] This figure shows an example of judgment information applied to the abnormal state detection system according to this embodiment. [Figure 5] This flowchart shows an example of the processing of the abnormal condition detection device according to this embodiment. [Modes for carrying out the invention]

[0011] The abnormal condition detection system 10 and abnormal condition detection device 100 according to this embodiment will be described in detail below with reference to the drawings. Note that the dimensional ratios in the drawings are exaggerated for illustrative purposes and may differ from the actual ratios.

[0012] (Regarding the fire on the car carrier ship) In recent years, the adoption of electric vehicles (EVs) has surged due to increasing global environmental concerns. Furthermore, rapid and abnormal global warming is increasing the risk of fires occurring during the voyage of car carriers.

[0013] As a result, not only are there fires caused by wiring displacement due to waves, as disclosed in Patent Document 1, but there is also an increase in fires in electric vehicles equipped with lithium (Li) ion batteries due to abnormal temperature increases inside the vehicle's cabin.

[0014] For example, it is generally known that the maximum storage temperature for not only lithium-ion batteries but also regular lead-acid batteries is around 40-50°C. Furthermore, nowadays, not only EVs and hybrid vehicles but also gasoline-powered vehicles have an increasing number of electrical components, and it is common for cars to constantly consume around 5-10mA as backup power for the vehicle's ECU and electrical components.

[0015] Furthermore, in order to save fuel costs, the ventilation systems in the vehicle storage compartments of car carriers may be shut off while they are at sea. As a result, temperatures inside the storage compartments can exceed 40°C.

[0016] In addition, there are cases where one sails in the ocean in a high-temperature zone due to abnormal weather conditions these days, and the temperature in the storage room where the automobile is stored may exceed 45°C, which is the allowable ambient temperature of the lithium-ion battery. For this reason, in the case of a vehicle equipped with a lithium-ion battery, which has a relatively high risk of ignition at high temperatures compared to a lead-acid battery, it is common to store and transport it on a lower deck where the room temperature in the storage room is less likely to rise even inside an automobile carrier. However, in any case, such an increase in the room temperature in the storage room is considered to be one of the factors increasing the risk of a fire occurring in the automobile carrier.

[0017] Furthermore, once a fire breaks out in an automobile, it is likely to spread rapidly and become too late to take action. In that case, evacuation from the scene will be inevitable. In addition, there are many cases where the scene has been burned down, and there are also many cases where surveillance cameras and the like are not installed, so there is also a problem that it is impossible to trace the root cause of the problem and the circumstances of the accident.

[0018] Generally, ships are obliged to install fire alarm equipment such as smoke detectors. However, in existing smoke detectors, due to reasons such as smoke flowing and diffusing when the ventilation equipment in the installation room operates, the distance from the smoke generation source to the smoke detector being far, or the amount of smoke generated at the initial stage of a fire being small, it may take a long time from the occurrence of a fire until the detection operation is performed.

[0019] In addition, there is a high possibility that the scale of the fire will expand and cause primary and secondary damages. These fire detectors that detect smoke are generally so-called post-detection and post-reporting, which detect and operate after a fire has occurred. Therefore, combined with the detection delay due to the diffusion of smoke as described above, there is a problem that very serious damage is likely to be suffered.

[0020] Furthermore, as disclosed in Patent Document 1, in the method of measuring the surface temperature of the vehicle body to compensate for the disadvantages of conventional smoke detectors, in the case of a lithium-ion battery fire, it is often too late when the surface temperature of the vehicle body has already started to rise.

[0021] As of November 2023, Japan owns 298 car carriers, the most in the world, far ahead of Norway (142), South Korea (80), Israel (59), and China (40). Meanwhile, China has been rapidly increasing the number of new car carriers in recent years. In this context, it is extremely important for Japan's national interests to ensure that the car carriers currently owned by Japan are protected from risks such as fire, and to implement safety measures and safe operations more than ever before. Furthermore, as the world leader in the car carrier sector, it is necessary for Japan to recognize the aforementioned problems and to take the lead in developing abnormal condition detection devices to solve them.

[0022] The abnormal condition detection device 100 according to this embodiment detects the possibility of fire from various perspectives, including "temperature control and proper ventilation equipment operation," "monitoring of rapid temperature rises in the storage room," "smoke detection monitoring," and "flame detection monitoring," and records, monitors, and alerts on the risk of fire from multiple angles. In other words, the abnormal condition detection device 100 according to this embodiment detects, records, monitors, and alerts on the occurrence of a fire or an event that precedes it ("rapid temperature rise," "smoke emission") or the risk of such an event occurring.

[0023] As a result, the abnormal condition detection device 100 according to this embodiment can quickly and reliably detect various fire possibilities in car carriers, including not only fires caused by conventional wiring contact, but also fires caused by airflow in the storage compartment, ship movement, and temperature increases in the storage compartment due to navigation in tropical hot regions. In other words, the abnormal condition detection device 100 according to this embodiment can quickly and reliably detect various fire possibilities in car carriers, including Li-ion battery fires associated with the global and rapid spread of electric vehicles, thereby minimizing damage.

[0024] (Configuration of the abnormal condition detection system 10) Referring to Figure 1, the schematic configuration of the abnormal state detection system 10 according to this embodiment will be described. The abnormal state detection system 10 consists of a camera 40 that captures images of a vehicle 50 loaded on a car carrier and the space surrounding the vehicle 50, an administrator terminal 20, a user terminal 30 operated by a user 31, and an abnormal state detection device 100.

[0025] In this embodiment, the space surrounding the vehicle 50 includes the roofs, hoods, and other body surfaces of multiple vehicles 50 located within a radius of approximately 30m from the camera, which are within the camera's field of view. Furthermore, the space surrounding the vehicle 50 includes the surface of the body of vehicles that are partially obscured by shadows from adjacent vehicles and whose roofs or other body parts are not visible from the camera's field of view, as well as the space around the vehicle body surface, approximately 1 to 3m, from those body surfaces to the ceiling of the storage compartment.

[0026] Furthermore, the abnormal condition detection system 10 may include a rotating light 60 and / or an alarm 70 that notifies that a fire or an event that foreshadows a fire has occurred or is likely to occur. In addition, the abnormal condition detection system 10 may include a cloud server 200.

[0027] In the abnormal condition detection system 10, the abnormal condition detection device 100, the cloud server 200, the administrator terminal 20, and the user terminal 30 can communicate with each other via the network 80.

[0028] The administrator terminal 20 is comprised of a PC (personal computer) installed, for example, in the ship's command room or bridge, and displays fire detection information and / or image data transmitted from the abnormal condition detection device 100.

[0029] The user terminal 30 is operated by users 31, such as crew members or workers who patrol the ship, and consists of devices such as smartphones or tablets. The user terminal 30 also displays fire detection information and / or image data transmitted from the abnormal condition detection device 100.

[0030] Camera 40 consists of a thermal camera unit and a full HD camera unit, and acquires camera information, which is information including image data of the vehicle 50 and the space surrounding the vehicle 50, and sends it to the abnormal condition detection device 100. In this embodiment, camera 40 is a thermal camera with a wide field of view (60 to 90 degrees) and a full HD camera with a wide field of view (80 to 100 degrees). The wide field of view allows the camera 40 to monitor the entire body of the loaded vehicle, or a part of the body, as much as possible, in order to quickly detect changes in the vehicle body. Depending on the vehicle loading situation in the storage compartment, even if at least a part of the vehicle body is not within the field of view of camera 40, it is desirable to arrange the camera so that the immediate surrounding space in contact with that part of the vehicle body is within the field of view of camera 40.

[0031] In this embodiment, the surrounding space of the vehicle 50 includes, as described above, the body surfaces of the vehicle 50 that are within the camera's field of view, such as the tops and hoods of multiple vehicles 50 loaded within a radius of approximately 30m from the camera. Furthermore, the surrounding space of the vehicle 50 includes, for vehicles where only a portion of the vehicle body, such as the top, is visible from the camera's field of view due to being in the shadow of an adjacent vehicle, the surface of that vehicle body, and the space around the vehicle body surface, approximately 1 to 3m, from those vehicle body surfaces to the ceiling of the storage compartment.

[0032] Furthermore, since camera 40 is equipped with a wide-angle thermal camera and an ultra-wide-angle full HD camera, the vehicle 50 being imaged by camera 40 only needs to be a part of its body, and it is desirable to design the placement of camera 40 so that almost the entire vehicle 50 being mounted is within the imaging field of view of camera 40.

[0033] Furthermore, in order to achieve higher accuracy, it is desirable that the entire vehicle body, or a part of the vehicle body, captured by the camera 40, falls within both the thermal camera's and the full HD camera's field of view within the camera 40's imaging field of view.

[0034] Vehicle 50 refers to automobiles and the like loaded onto a car carrier, and multiple vehicles 50 of a wide variety of types are loaded onto the car carrier at an extremely high density.

[0035] The rotating light 60 and alarm 70 acquire information from the abnormal condition detection device 100 when a fire or an event that foreshadows a fire occurs, or when there is a risk of such an event occurring, and notify the surrounding area of ​​the risk by rotating the light or making a sound.

[0036] The cloud server 200 continuously acquires and stores information on "smoke" and "flame" events, including navigation position information and weather information entered daily by the operator, temperature change information inside the storage room collected and detected by the camera 40 and the abnormal condition detection device 100 (described later) via the network 80, and the judgment results from the abnormal condition judgment unit 117 (described later). This information is then compiled into a database, and the probability of occurrence in the near future is calculated. Note that the navigation position information and weather information may be information that is automatically collected and registered via the network 80, not just information entered by the operator.

[0037] For example, the cloud server continuously acquires and stores information such as "navigation position information and weather information such as outside temperature and weather," "temperature change information in the hangar rooms on each floor," and events such as "sudden abnormal temperature rises," "smoke emission," and "fire" in the vehicle hangar rooms of the car carrier, and builds a database.

[0038] As a result, the cloud server 200 calculates the correlation between "navigation position and weather information such as outside temperature and weather" and "room temperature inside the hangar on each floor," as well as predicting these correlations, and further calculates the probability of occurrences such as "rapid temperature rise," "smoke," and "flame" when these events occur. This makes it possible to notify the bridge in advance of "navigation position," "weather information," "which floor's vehicle hangar is most likely to experience a temperature rise," and "whether a rapid temperature rise is likely to occur as a result," based on "near-future navigation position and weather information."

[0039] Furthermore, vehicle type information such as "gasoline vehicle," "EV vehicle," and "PHEV vehicle" may be added to the cloud server 200. This will allow for consideration of improvements to operational methods, such as which vehicle types should be loaded on which tier, and what loading methods should be used on which routes and in which seasons. This will enable fire alarms to be not only based on "post-incident notification" or "post-incident alert" as in the past, but also on "pre-incident prediction" and "pre-incident notification."

[0040] For example, consider a scenario where a car carrier departs Japan, passes through the Strait of Hormuz 15 days later, and reaches the vicinity of Dubai in the UAE, with an expected outside temperature exceeding 40°C. In this case, if the temperature on the 10th floor of the ship is calculated to exceed 55°C based on past correlations between outside and inside temperatures, even if the operating company has not experienced any emergency incidents such as "smoke," "flame," or "rapid temperature rise" in the past, this system can clearly identify the risk in the correlation between outside and inside temperatures, which was previously handled based on empirical rules. This allows for more reliable risk recognition and action towards hazard avoidance at the operational level.

[0041] The calculation of the probability of occurrence on this cloud server 200 may be performed using a generation AI or the like, and the cloud server 200 may be located in a different location from the car carrier. Alternatively, the cloud server 200 may be located inside the car carrier.

[0042] (Configuration of abnormal condition detection device 100) Figure 2 is a block diagram showing the configuration of the abnormal state detection device 100. As shown in Figure 2, the abnormal state detection device 100 is composed of a general-purpose computer comprising a control unit 110, a storage unit 120, an input / output IF 130 (Interface), and a communication IF 140. Note that the configuration example of the abnormal state detection device 100 shown in Figure 2 is merely an example, and this embodiment is not limited thereto. Furthermore, when the administrator terminal 20 and user terminal 30 are implemented as personal computers, smartphones, tablets, etc., their configuration may be the same as the block diagram shown in Figure 2.

[0043] The control unit 110 controls the entire abnormal state detection device 100, for example, by operating the operating system. Furthermore, the control unit 110 operates based on a program (not shown) stored in the storage unit 120 and executes each of the functions provided. Note that the program is not limited to being stored in the storage unit 120, but may also be stored in a ROM or the like (not shown) within the abnormal state detection device 100.

[0044] The storage unit 120 includes an acquired information DB 121 (see Figure 3), a detection information DB 122 (see Figure 3), and a judgment information DB 123 (see Figure 3). The storage unit 120 may consist of one unit or multiple units. For example, a single storage unit 120 may be configured to store data in separate areas. Alternatively, data may be distributed and stored in multiple storage devices located in physically separate locations.

[0045] The input / output IF130 shown in Figure 2 is, for example, an interface for the user to exchange data with the abnormal condition detection device 100. Specifically, a pre-registered ship's floor plan and elevation view, along with the positions of the cameras 40, are plotted on the LCD screen connected to the abnormal condition detection device 100, and these are linked to the control unit 110. This makes it possible to clearly indicate to the crew, along with colors, signs, and sounds, the past and present indoor temperature information of the hangar captured by each camera 40, and which camera 40 is capturing a vehicle 50 where an emergency event such as "sudden temperature rise," "smoke," or "flame" has occurred, by displaying the camera image in a pop-up window.

[0046] Furthermore, weather information and navigation information, as well as future risk information calculated based on this information, are also displayed on the screen. In this embodiment, the input / output IF130 is connected to the rotating light 60 and the alarm 70 and also functions as an interface for outputting fire alarms.

[0047] The communication IF140 is an interface that enables the abnormal condition detection device 100 to communicate with the outside world via the network 80.

[0048] (Functional configuration of abnormal condition detection device 100) Next, the functional configuration of the abnormal state detection device 100 will be described. Figure 3 is a block diagram showing the functional configuration of the abnormal state detection device 100. As shown in Figure 3, the control unit 110 includes the following functions: a navigation information acquisition unit 111, a camera information acquisition unit 112, a temperature information acquisition unit 113, a rapid temperature rise detection unit 114, a smoke generation detection unit 115, a flame generation detection unit 116, an abnormal state determination unit 117, and a notification unit 118.

[0049] The navigation information acquisition unit 111 acquires the position information of the car carrier. Specifically, it acquires navigation plan information by inputting it from the operator or from the operation management system, and the navigation information acquisition unit 111 acquires navigation records by receiving radio waves from multiple GPS (Global Positioning System) satellites using a GPS receiver (not shown) installed on the car carrier to obtain the current position of the car carrier. The navigation information acquisition unit 111 also stores the acquired position information in the acquisition information DB 121.

[0050] The camera information acquisition unit 112 acquires camera information, which includes image data of the vehicle 50 and the surrounding space. The camera information acquisition unit 112 also stores the acquired camera information in the acquired information DB 121. The image data of the vehicle 50 and the surrounding space may be, for example, the data shown in Figure 4A. It is desirable that the image data of the vehicle 50 and the surrounding space consist of a high-resolution image of FullHD or higher on the left side of Figure 4A and thermal image data on the right side. By acquiring these two types of camera image information, camera information for "temperature information inside the storage room," "sudden temperature rise detection," "smoke detection," "flame detection," etc., can be acquired.

[0051] The temperature information acquisition unit 113 acquires temperature information of the vehicle and the surrounding space based on camera information. The temperature information acquisition unit 113 also acquires temperature information from each camera and calculates temperature change information, which shows the temperature change per unit time, from the acquired information. The temperature information acquisition unit 113 stores the acquired temperature information and temperature change information in the acquisition information DB 121.

[0052] The temperature surge detection unit 114 detects a rapid temperature rise based on the temperature information acquired and calculated by the aforementioned temperature information acquisition unit 113, and on a preset setting for each camera, "whether a rapid temperature rise is defined as a temperature rise when a predetermined temperature rise occurs per unit time." The predetermined temperature per unit time in the temperature surge detection unit 114 is, for example, a rise of 20°C or more per minute, in which case it is determined that the temperature has risen rapidly. However, the predetermined temperature per unit time in the temperature surge detection unit 114 is not limited to determining a temperature of 20°C or more per minute; it may also be determined by a temperature lower or higher than 20°C per shorter or longer period of time than one minute.

[0053] The smoke generation detection unit 115 detects whether or not smoke is being generated in the vehicle and the surrounding space based on camera information. Specifically, the smoke generation detection unit 115 detects whether or not smoke is being generated in the vehicle and the surrounding space based on data that has been pre-trained using images of smoke being generated, and camera information acquired via the full HD camera unit of the camera 40.

[0054] For example, the smoke generation detection unit 115 may use a pre-trained model that has been machine-learned based on image data such as the "color," "density," "shape," and "changes in this information over time" of the smoke to calculate the degree of smoke generation, indicating whether or not smoke is being generated, and generate this as a judgment status. In this case, for example, the degree of smoke generation may be expressed as a percentage with a maximum value of 100, but in cases where a time-sensitive decision is required, such as in this system, it is desirable to be able to adjust the similarity judgment level with the pre-trained model by setting parameters, etc., and to promptly issue an alarm notification when smoke generation (including suspected smoke generation) is detected. The smoke generation detection unit 115 also stores the results of detecting whether or not smoke is being generated in the vehicle and the surrounding space in the detection information DB 122.

[0055] The flame detection unit 116 detects whether flames are present in or around the vehicle based on camera information (temperature measurement information and images from the thermal camera) acquired via the thermal camera unit of the camera 40. Specifically, the flame detection unit 116 compares the extremely high temperature detection, which is likely to be a flame, from the temperature information measured by the thermal camera unit of the camera 40, with the camera image information acquired from the thermal camera unit at the high-temperature detection location, which was obtained using data that had been previously trained on machine learning to capture images of flames. Using both the temperature information and the image information, the unit detects whether or not flames are present in or around the vehicle.

[0056] For example, the flame detection unit 116 may compare the image information obtained from the thermal camera unit of the camera 40, which is an extremely high temperature of approximately 1000°C or higher, with the camera image information from the thermal camera unit obtained at the location where the high temperature was detected, using a pre-trained model that has been machine-learned based on image data such as the "color," "shape," and "changes in this information over time" of the flame. The unit may then determine whether or not a flame is present and generate a determination status. Alternatively, the flame detection unit 116 may use the high temperature information obtained from the thermal camera unit and the pre-trained model to calculate the degree of flame occurrence, and generate this as a determination status. In this case, for example, the degree of flame occurrence may be expressed as a percentage with a maximum value of 100. The flame detection unit 116 also stores the results of detecting whether or not a flame is present in the vehicle and the surrounding space in the detection information DB 122.

[0057] The abnormal state determination unit 117 determines whether a fire has occurred or is likely to occur in the vehicle or the surrounding space, based on location information, temperature information, temperature change information, the smoke generation status detected by the smoke generation detection unit 115, and / or the flame generation status detected by the flame generation detection unit 116. The abnormal state determination unit 117 also stores the determination result in the determination information DB 123.

[0058] The abnormal condition determination unit 117 determines, based on location information and temperature information, whether the temperature information is higher than the assumed temperature in the area currently being navigated. If the temperature is significantly higher than the assumed temperature, it determines that a fire has occurred or is likely to occur in the vehicle or the surrounding space.

[0059] The assumed temperature may be information stored in the memory unit 120 in advance, associated with location information. Alternatively, the assumed temperature may be obtained from a weather information provision application using weather information corresponding to the location information, and the temperature corresponding to the location information may be used as the assumed temperature. In other words, the assumed temperature is the temperature expected in the area where the car carrier is navigating, and corresponds to assumed temperature information.

[0060] Furthermore, the abnormal state determination unit 117 may determine that a fire has occurred or is likely to occur in the vehicle and the surrounding space if the temperature indicated by the temperature information is 100°C higher than the assumed temperature. This comparison temperature with the assumed temperature is not limited to the configuration of this embodiment, and a threshold value lower than 100°C or higher than 100°C may be set to determine that a fire has occurred or is likely to occur in the vehicle and the surrounding space.

[0061] Furthermore, the abnormal state determination unit 117 determines whether or not flames are present in the vehicle and its surroundings based on temperature change information. Specifically, the abnormal state determination unit 117 determines that flames are present in the vehicle and its surroundings if, for example, the temperature rises by 20°C or more per minute. However, the determination by the abnormal state determination unit 117 based on temperature change information is not limited to cases where the temperature rises by 20°C or more per minute. For example, the abnormal state determination unit 117 determines that flames are present or likely to occur in the vehicle and its surrounding space if the temperature rises by approximately 20°C over a period of time shorter than one minute or longer than one minute.

[0062] Furthermore, the abnormal condition determination unit 117 determines, based on the smoke generation status detected by the smoke generation detection unit 115, that smoke is being generated or is likely to be generated in the vehicle and the surrounding space. For example, if the smoke generation status detected by the smoke generation detection unit 115 is indicated by a smoke generation degree, the unit may determine that smoke is being generated or is likely to be generated in the vehicle and the surrounding space if the smoke generation degree is 70% or higher.

[0063] Furthermore, the abnormal condition determination unit 117 determines, based on the flame generation status detected by the flame generation detection unit 116, that flames are present or likely to be present in the vehicle and the surrounding space. For example, if the flame generation status detected by the flame generation detection unit 116 is indicated by the degree of smoke generation, the unit may determine that flames are present or likely to be present in the vehicle and the surrounding space if the degree of smoke generation is 70% or higher.

[0064] Figure 4C shows an example of the judgment information DB 123 in which the judgment results determined by the abnormal state determination unit 117 are stored. In the example shown in Figure 4C, for example, camera 1 is determined to be not experiencing a fire. On the other hand, in the example shown in Figure 4C, camera X is determined to be experiencing a fire based on the acquired camera information, using temperature information, temperature change information, smoke detection, and flame detection, and the judgment result is that a fire has occurred.

[0065] If the notification unit 118 determines that a fire has occurred or is likely to occur, it notifies at least one of the following devices installed on the car carrier: the rotating light 60, the alarm 70, the user terminal 30, and the administrator terminal 20, that a fire has occurred or is likely to occur.

[0066] For example, when the notification unit 118 notifies the rotating light 60, it may differentiate the content of the notification depending on whether it determines that a fire has occurred or that there is a risk of a fire occurring. For example, if the notification unit 118 determines that a fire has occurred, it may notify the rotating light 60 to rotate faster than if it determines that there is a risk of a fire occurring. This allows crew members who see the rotating light 60 to recognize the urgency of the fire or the risk of a fire occurring.

[0067] Similarly, when the notification unit 118 notifies the alarm 70, user terminal 30, and / or administrator terminal 20, it may differentiate the notification content depending on whether it determines that a fire has occurred or that there is a risk of a fire. For example, if the notification unit 118 determines that a fire has occurred, it may increase the volume of the alarm 70 and make the display on the user terminal 30 and / or administrator terminal 20 more urgent compared to when it determines that there is a risk of a fire. This allows crew members who hear the alarm 70, see the display on the user terminal 30, and see the display on the administrator terminal 20 to recognize the degree of urgency regarding the occurrence or risk of a fire.

[0068] (Outline of the processing flow of the abnormal condition detection system 10) Next, the process related to abnormal state detection in the abnormal state detection device 100 (abnormal state detection method) will be explained based on the flowchart in Figure 5. In the flowchart shown in Figure 5, the process will also terminate due to power off or processing termination interruption. In addition, in the explanation of the flowchart below, the same content as described in the above-mentioned explanation of the abnormal state detection system 10 and abnormal state detection device 100 will be omitted or simplified.

[0069] In step S501, the navigation information acquisition unit 111 acquires the position information of the car carrier. Specifically, the navigation information acquisition unit 111 acquires the current position of the car carrier using a system called AIS (Automatic Identification System), which is installed on the car carrier and automatically sends and receives information such as the position of the ship using GPS information. The navigation information acquisition unit 111 also stores the acquired position information in the acquisition information DB 121. After that, the process proceeds to step S502.

[0070] In step S502, the camera information acquisition unit 112 acquires camera information, which includes image data of the vehicle 50 and the surrounding space. The camera information acquisition unit 112 also stores the acquired camera information in the acquisition information DB 121. The image data of the vehicle 50 and the surrounding space may be, for example, the data shown in Figure 4A. Alternatively, the image data of the vehicle 50 and the surrounding space may be thermal image data shown on the right in Figure 4B. The process then proceeds to step S503.

[0071] In step S503, the temperature information acquisition unit 113 acquires temperature information of the vehicle and the surrounding space based on camera information. The temperature information acquisition unit 113 also calculates temperature change information, which shows the temperature change per unit time, from the acquired temperature information. The temperature information acquisition unit 113 stores the acquired temperature information and temperature change information in the acquired information DB 121. After that, the process proceeds to step S504.

[0072] In step S504, the temperature rise detection unit 114 determines, based on the temperature change information, whether or not flames are occurring in the vehicle and its surroundings. Specifically, the temperature rise detection unit 114 determines that flames are occurring in the vehicle 50 and its surroundings if, for example, the temperature rises by 20°C or more in one minute. However, the determination by the temperature rise detection unit 114 based on the temperature change information is not limited to cases where the temperature rises by 20°C or more in one minute. For example, the abnormal condition determination unit 117 determines that flames are occurring or are likely to occur in the vehicle and its surrounding space if the temperature rises by about 20°C in a time shorter than one minute or longer than one minute.

[0073] In step S504, if the temperature rapid rise detection unit 114 determines that the temperature is rising rapidly (step S504: YES), the process proceeds to step S507. On the other hand, in step S504, if the temperature rapid rise detection unit 114 determines that the temperature is not rising rapidly (step S504: NO), the process proceeds to step S505.

[0074] In step S505, the smoke generation detection unit 115 detects whether or not smoke is being generated in the vehicle 50 and the surrounding space based on the camera information. Specifically, the smoke generation detection unit 115 uses a trained model that has previously learned images of smoke being generated to detect whether or not smoke is being generated in the vehicle and the surrounding space based on the acquired camera information.

[0075] In step S505, if the smoke generation detection unit 115 detects that smoke is being generated in the vehicle 50 and the space surrounding the vehicle 50 (step S505: YES), the process proceeds to step S507. On the other hand, in step S505, if the smoke generation detection unit 115 detects that no smoke is being generated in the vehicle 50 and the space surrounding the vehicle 50 (step S505: NO), the process proceeds to step S506.

[0076] In step S506, the flame detection unit 116 detects whether flames are present in the vehicle 50 and its surroundings based on camera information. Specifically, the flame detection unit 116 uses a pre-trained model that has learned images of flames in advance to detect whether smoke is present in the vehicle 50 and the space surrounding the vehicle 50 based on the acquired camera information.

[0077] In step S506, if the flame detection unit 116 detects that flames are present in the vehicle 50 and the surrounding space (step S506: YES), the process proceeds to step S507. On the other hand, in step S506, if the flame detection unit 116 detects that no flames are present in the vehicle 50 and the surrounding space (step S506: NO), the process ends.

[0078] In step S507, the abnormal condition determination unit 117 determines whether a fire has occurred or is likely to occur in the vehicle 50 or the space surrounding the vehicle 50, based on the location information, temperature information, temperature change information, smoke generation status, and / or flame generation status. The abnormal condition determination unit 117 also stores the determination result in the determination information DB 123.

[0079] In step S508, if the notification unit 118 determines that a fire has occurred or is likely to occur, it notifies at least one of the following: the rotating light 60, the alarm 70, the user terminal 30, and the administrator terminal 20 that a fire has occurred or is likely to occur. After that, the process ends.

[0080] For example, when the notification unit 118 notifies the rotating light 60, it may differentiate the content of the notification depending on whether it determines that a fire has occurred or that there is a risk of a fire occurring. For example, if the notification unit 118 determines that a fire has occurred, it may notify the rotating light 60 to rotate faster than if it determines that there is a risk of a fire occurring. This allows crew members who see the rotating light 60 to recognize the urgency of the fire or the risk of a fire occurring.

[0081] Similarly, when the notification unit 118 notifies the alarm 70, user terminal 30, and / or administrator terminal 20, it may differentiate the notification content depending on whether it determines that a fire has occurred or that there is a risk of a fire. For example, if the notification unit 118 determines that a fire has occurred, it may increase the volume of the alarm 70 and make the display on the user terminal 30 and / or administrator terminal 20 more urgent compared to when it determines that there is a risk of a fire. This allows crew members who hear the alarm 70, see the display on the user terminal 30, and see the display on the administrator terminal 20 to recognize the degree of urgency regarding the occurrence or risk of a fire.

[0082] As described above, the abnormal state detection device 100 according to this embodiment is an abnormal state detection device 100 that detects a fire occurring in a vehicle 50 loaded on a car carrier. The abnormal state detection device 100 includes a navigation information acquisition unit 111 that acquires position information of the car carrier as it is navigating. The abnormal state detection device 100 also includes a camera information acquisition unit 112 that acquires camera information, which is information including image data captured by a camera of the vehicle and the space around the vehicle. The abnormal state detection device 100 also acquires temperature information of the vehicle 50 and the space around the vehicle 50, and temperature change information per unit time, based on the camera information, from a temperature information acquisition unit 113. The abnormal state detection device 100 also includes an abnormal state determination unit 117 that determines whether a fire has occurred or is likely to occur in the vehicle 50 and the space around the vehicle 50, based on the assumed temperature information, temperature information, and temperature change information based on the position information. Furthermore, the abnormal condition detection device 100 includes a notification unit 118 that notifies the system that a fire has occurred or is likely to occur when the abnormal condition determination unit 117 determines that a fire has occurred or is likely to occur. The notification unit 118 notifies at least one of the following devices installed on the car carrier: the rotating light 60, the alarm 70, the user terminal 30, and the administrator terminal 20, that a fire has occurred or is likely to occur.

[0083] With this configuration, the abnormal condition detection device 100 can detect the occurrence or risk of fire based on assumed temperature information based on the position information of the car carrier, and temperature information and temperature change conditions of the vehicle 50 and the space around the vehicle 50, thereby enabling detection according to the navigation conditions. As a result, the abnormal condition detection device 100 can appropriately detect the occurrence or risk of fire of a vehicle 50 loaded on the car carrier.

[0084] Furthermore, the assumed temperature information of the abnormal condition detection device 100 may correspond to the temperature expected in the area where the car carrier is navigating. This allows the abnormal condition detection device 100 to perform fire detection in accordance with the temperature conditions of the area where it is navigating, enabling more accurate fire detection.

[0085] Furthermore, the abnormal condition detection device 100 may further include a smoke generation detection unit 115 that detects whether or not smoke is being generated in the vehicle 50 and the space surrounding the vehicle 50 based on camera information. In addition, the abnormal condition determination unit 117 may determine whether or not a fire has occurred or is likely to occur in the vehicle 50 and the space surrounding the vehicle 50 based on location information, temperature information, temperature change information, and the smoke generation status detected by the smoke generation detection unit 115. This makes it possible for the abnormal condition detection device 100 to more appropriately detect the generation of smoke in the initial stages of a fire, and to more appropriately detect the occurrence of a fire or the risk of a fire occurring in the vehicle 50 loaded on the car carrier.

[0086] Furthermore, the smoke generation detection unit 115 may detect whether or not smoke is being generated in the vehicle 50 and the surrounding space based on data previously obtained by machine learning using images of smoke being generated, and camera information. This enables the abnormal condition detection device 100 to perform more accurate smoke generation detection processing.

[0087] Furthermore, the abnormal condition detection device 100 may further include a flame generation detection unit 116 that detects whether flames are occurring in the vehicle 50 and the space surrounding the vehicle 50 based on camera information. The abnormal condition determination unit 117 may determine whether a fire has occurred or is likely to occur in the vehicle 50 and the space surrounding the vehicle 50 based on location information, temperature information, temperature change information, smoke generation status detected by the smoke generation detection unit, and flame generation status detected by the flame generation detection unit. This enables the abnormal condition detection device 100 to more appropriately detect the occurrence of flames and to reliably detect the occurrence of a fire or the risk of a fire occurring in the vehicle 50 loaded on the car carrier.

[0088] Furthermore, the flame detection unit 116 may detect whether or not flames are present in the vehicle and the surrounding space based on data previously obtained by machine learning using images of flames and camera information. This enables the abnormal condition detection device 100 to perform more accurate flame detection processing.

[0089] Furthermore, the notification unit 118 may notify the rotating light 60, the alarm 70, the user terminal 30, and / or the administrator terminal 20 of different content depending on whether it is determined that a fire has occurred or that there is a risk of a fire occurring. With this configuration, crew members who see the rotating light 60, crew members who hear the sound of the alarm 70, crew members who see the display on the user terminal 30, and / or crew members who see the display on the administrator terminal 20 can recognize the degree of urgency regarding the occurrence or risk of a fire. As a result, crew members of a car carrier to which the abnormal condition detection device 100 according to this embodiment is applied can take appropriate firefighting or fire prevention actions according to the situation of the fire and the risk of a fire occurring.

[0090] (Other embodiments) Although these embodiments have been described above, the embodiments are not limited to these, and various modifications are possible within the scope of the gist of the embodiments. Furthermore, it is possible to combine some or all of the various embodiments to create new embodiments.

[0091] In the above-described embodiment, the control unit 110 of the abnormal condition detection device 100 is shown to have the functions shown in Figure 3, but the embodiment is not limited to this configuration. For example, some of the functions of the control unit 110 in Figure 3 may be provided on the camera 40. By providing some of the functions shown in Figure 3 on the camera 40, functions such as rapid temperature rise detection, smoke generation detection, and flame generation detection can be calculated by distributing the calculations among the cameras 40, making it possible to perform them more quickly. Furthermore, in cases like this one, where car carriers are extremely large and often do not have a wideband onboard network LAN, providing some of these functions on the camera terminal side reduces the load on the network and makes it easier to achieve more stable communication. In addition, it is possible to reduce the computational load of the abnormal condition detection device 100 and miniaturize the computer hardware in the limited space of the onboard server room, resulting in higher system stability and availability.

[0092] The computer program (abnormal state detection program) that causes a computer to execute the abnormal state detection method described above, and the computer-readable recording medium on which the program is stored, are included within the scope of this embodiment. Here, the type of computer-readable recording medium is arbitrary. Furthermore, the computer program described above is not limited to that stored on the recording medium described above, but may also be transmitted via telecommunication lines, wireless or wired communication lines, networks such as the Internet, etc. [Explanation of Symbols]

[0093] 10 Anomaly detection system 20 Administrator terminals 30 User terminals 40 Cameras 50 vehicles 60 Rotating Lights 70 Alarm 80 Networks 100 Abnormal State Detection Device 110 Control Unit 111 Navigation information acquisition section 112 Camera Information Acquisition Unit 113 Temperature information acquisition section 114 Temperature Rise Detection Unit 115 Smoke generation detection unit 116 Flame detection unit 117 Abnormal State Determination Unit 118 Notification Department 120 Storage section 121 Acquisition information DB 122 Detection Information DB 123 Judgment information DB 130 Input / Output Interfaces 140 Communication IF 200 cloud servers

Claims

1. An abnormal condition detection device for detecting fires occurring in vehicles loaded on a car carrier, A navigation information acquisition unit that acquires location information of the car carrier, A camera information acquisition unit acquires camera information, which is information including image data captured by a camera of the vehicle and the surrounding space of the vehicle. A temperature information acquisition unit acquires temperature information of the vehicle and the surrounding space, and temperature change information per unit time, based on the camera information. A smoke generation detection unit detects whether or not smoke is being generated in the vehicle and the surrounding space based on the camera information, A flame detection unit detects whether or not flames are present in the vehicle and the surrounding space based on the camera information, An abnormal state determination unit determines whether a fire has occurred or is likely to occur in the vehicle or the space surrounding the vehicle, based on the assumed temperature information based on the location information, the temperature information, the temperature change information, the smoke generation status detected by the smoke generation detection unit, and the flame generation status detected by the flame generation detection unit. An abnormal condition detection device comprising: a notification unit that, when the abnormal condition determination unit determines that a fire has occurred or there is a risk of a fire occurring, notifies at least one of the rotating lights, alarms, user terminals, and administrator terminals installed on the car carrier that a fire has occurred or there is a risk of a fire occurring.

2. The abnormal condition detection device according to claim 1, wherein the assumed temperature information corresponds to the temperature expected in the area where the car carrier is navigating.

3. The abnormal condition detection device according to claim 1, wherein the smoke generation detection unit detects whether or not smoke is being generated in the vehicle and the surrounding space of the vehicle based on data that has been previously trained using images of smoke generation and the camera information.

4. The abnormal condition detection device according to claim 1, wherein the flame generation detection unit detects whether or not flames are present in the vehicle and the surrounding space based on data that has been previously trained using images of flames and the camera information.

5. The abnormal condition detection device according to claim 1, wherein the notification unit notifies the rotating light, the alarm, the user terminal, and / or the administrator terminal of different content when it is determined that a fire has occurred and when it is determined that there is a risk of a fire occurring.

6. An abnormal condition detection device according to any one of claims 1 to 5, A camera for imaging the vehicles loaded onto the aforementioned car carrier, At least one of the rotating light, the alarm, the user terminal, and the administrator terminal, An abnormal condition detection system equipped with the following features.

7. An abnormal condition detection method, performed by a computer, for detecting a fire occurring in a vehicle loaded on a car carrier, The car carrier acquires its positional information as it navigates, Camera information is acquired, which includes image data captured by a camera of the vehicle and the surrounding space of the vehicle. Based on the camera information, temperature information of the vehicle and the surrounding space, and temperature change information per unit time are acquired. Based on the camera information, the system detects whether or not smoke is being generated in the vehicle and the surrounding space, Based on the camera information, the system detects whether or not flames are present in the vehicle and the surrounding space, Based on the assumed temperature information based on the location information, the temperature information, the temperature change information, the smoke generation status, and the flame generation status, it is determined whether a fire has occurred or is likely to occur in the vehicle and the surrounding space. An abnormal condition detection method that, when it is determined that a fire has occurred or there is a risk of a fire occurring, notifies at least one of the rotating lights, alarms, user terminals, and administrator terminals installed on the car carrier that a fire has occurred or there is a risk of a fire occurring.

Images