Quality control system and quality control method

Abstract

Provided is a quality control system for managing the quality of cargo on the basis of odors in a container in which the cargo is stored. The quality control system comprises: a plurality of sensors including at least an odor sensor for measuring odors in the container to generate odor data, the plurality of sensors measuring conditions in the container to generate measurement data; a calculation device for analyzing a state of the cargo on the basis of the odor data; and a monitoring device that determines whether to execute an alarm operation for giving a notification of the state of the cargo on the basis of an alarm level corresponding to the analysis result of the state of the cargo, analyzes an abnormality cause on the basis of the measurement data and a prescribed threshold corresponding to the measurement data when the alarm operation has been executed, and executes a prescribed action on the basis of the alarm level and the abnormality factor.

Description

Quality control system and quality control method

[0001] The present disclosure relates to quality control systems and methods.

[0002] Conventionally, quality control of goods during transportation has been required to prevent and address quality problems of goods during transportation. For example, Patent Document 1 discloses a transportation quality control system that grasps and manages stress experienced by goods during transportation. This transportation quality control system includes a stress sensor that detects stress experienced by goods packaged and transported in a real package, a sensor module having a transmitter that transmits data detected by the stress sensor via weak radio waves, and a dummy package in which some or all of the goods in the package are replaced with the sensor module. The transportation quality control system further includes a receiving module installed on a truck that receives data transmitted from the sensor module, and an external communication module connected to the receiving module and connected to a transportation management center via a communication line. The transportation quality control system then monitors the stress experienced by the goods at the transportation management center.

[0003] Japanese Patent Application Laid-Open No. 2008-94616

[0004] Conventionally, various measures have been considered for quality control of cargo during transportation, such as transporting cargo while controlling its temperature or transporting cargo in packaging that prevents it from collapsing. However, it is difficult to completely eliminate quality problems of cargo during transportation. When it is discovered that there is a risk of quality problems with cargo during transportation, it is possible to open the container in which the cargo is stored in order to check the specific condition of the cargo. However, if the container is opened during transportation, the opening may cause deterioration in the quality of the cargo. Furthermore, opening the container to check the condition of the cargo may be impractical when there are a large number of containers to be checked.

[0005] The present disclosure has been devised in view of the above-mentioned conventional situation, and aims to manage the quality of cargo based on the odor of the cargo.

[0006] The present disclosure provides a quality control system that manages the quality of cargo based on the odor inside a container storing the cargo, the quality control system comprising: a plurality of sensors that measure the condition inside the container and generate measurement data, including at least an odor sensor that measures the odor inside the container and generates odor data, a computing device that analyzes the condition of the cargo based on the odor data, and a monitoring device that determines whether or not to execute an alarm operation to notify the condition of the cargo based on an alarm level corresponding to the analysis result of the condition of the cargo, and if the alarm operation is executed, analyzes an abnormality factor based on the measurement data and a threshold value corresponding to the measurement data, and executes an action based on the alarm level and the abnormality factor.

[0007] The present disclosure also provides a quality control method for managing the quality of cargo based on the odor inside a container in which the cargo is stored, the quality control method including: measuring the odor inside the container to generate odor data; measuring the condition inside the container to generate measurement data; analyzing the condition of the cargo based on the odor data; determining whether to execute an alarm operation to notify the condition of the cargo based on an alarm level corresponding to the analysis result of the condition of the cargo; if the alarm operation is executed, analyzing abnormal factors based on the measurement data and a threshold value corresponding to the measurement data; and executing an action based on the alarm level and the abnormal factor.

[0008] Any combination of the above components, and conversion of the expression of the present disclosure into a method, device, system, storage medium, computer program, etc., are also valid aspects of the present disclosure.

[0009] According to the present disclosure, the quality of cargo can be controlled based on the odor of the cargo.

[0010] 1. Block diagram showing an example of the configuration of a quality control system according to the first embodiment. 2. Schematic diagram for explaining color changes associated with ripening of bananas. 3. Graphic diagram for explaining cleaning of odors adhering to an odor sensor according to the first embodiment. 4. Flowchart of processing by a computing device according to the first embodiment. 5. Schematic diagram for explaining an example of odor data according to the first embodiment. 6. Table diagram for explaining an example of an odor analysis message according to the first embodiment. 7. Table diagram for explaining an alarm level management table corresponding to bananas according to the first embodiment. 8. Table diagram for explaining a cargo management table according to the first embodiment. FIG. 1 is a table diagram illustrating an example of a freight transport destination management table according to the first embodiment; FIG. 2 is a table diagram illustrating an example of a transport destination tolerance level management table according to the first embodiment; FIG. 3 is a table diagram illustrating an example of a transport base management table according to the first embodiment; FIG. 4 is a flow chart of processing by the monitoring device according to the first embodiment; FIG. 5 is a flow chart of processing by the monitoring device according to the first embodiment;

[0011] Hereinafter, with reference to the drawings as appropriate, detailed descriptions of embodiments specifically disclosing the quality control system and quality control method according to the present disclosure will be provided. However, unnecessary detailed descriptions may be omitted. For example, detailed descriptions of well-known matters and redundant descriptions of substantially identical configurations may be omitted. This is to avoid unnecessary redundancy in the following description and to facilitate understanding by those skilled in the art. Note that the accompanying drawings and the following description are provided to enable those skilled in the art to fully understand the present disclosure and are not intended to limit the subject matter recited in the claims.

[0012] (First Embodiment) [System Configuration] Fig. 1 is a block diagram showing an example configuration of a quality control system 1 according to the first embodiment. The quality control system 1 is a system that controls the quality of cargo 6 based on the odor inside a container 10 in which the cargo 6 is stored. When the cargo 6 is transported by land, the container 10 is loaded onto, for example, a trailer 17. Note that the trailer 17 may also be read as a truck. When the cargo 6 is transported by sea, the container 10 is loaded onto, for example, a container ship 18. Note that in this specification, "cargo" includes goods, merchandise, materials, parts, or a combination thereof.

[0013] The quality control system 1 includes a sensor module 2 , a pump 3 , a power supply device 4 , a communication device 5 , a computing device 8 , and a monitoring device 9 .

[0014] The sensor module 2, the pump 3, the power supply device 4, and the communication device 5 are installed in a container 10. A power source can be supplied to the container 10. The container 10 may be, for example, a reefer container. Although not shown in FIG. 1 , an air conditioner is installed inside the container 10.

[0015] The sensor module 2 is composed of a plurality of sensors that measure the conditions inside the container 10 and generate measurement data. The sensor module 2 includes an odor sensor 11, a temperature sensor 12, a humidity sensor 13, an acceleration sensor 14, an illuminance sensor 15, and a Global Positioning System (hereinafter referred to as "GPS") sensor 16.

[0016] The odor sensor 11 measures the odor inside the container 10 and generates odor data as measurement data. More precisely, the odor sensor 11 measures the odor of the air sent in by the pump 3. The temperature sensor 12 measures the temperature inside the container 10 and generates temperature data as measurement data. The humidity sensor 13 measures the humidity inside the container 10 and generates humidity data as measurement data. The acceleration sensor 14 measures the acceleration of the container 10 and generates acceleration data as measurement data. The illuminance sensor 15 measures the illuminance inside the container 10 and generates illuminance data as measurement data. The GPS sensor 16 measures the position of the container 10 and generates position information of the container 10 as measurement data. The odor data will be described later with reference to FIG. 6. The temperature data, humidity data, acceleration data, illuminance data, and position information will be described later with reference to FIG. 13.

[0017] The pump 3 sends air to the odor sensor 11. The pump 3 alternately sends air from inside the container 10, in other words, air containing the odor of the cargo 6, and fresh air to the odor sensor 11 at specified time intervals. Here, fresh air refers to gas that does not contain an odor, and is a term used to distinguish it from air containing the odor of the cargo 6. An example of fresh air is air outside the container 10, in other words, outside air. In the example of FIG. 1 , the pump 3 takes in fresh air from outside the container 10, but, for example, a container storing odorless air may be installed in advance inside the container 10. The pump 3 may then send the air stored in the container as fresh air to the odor sensor 11.

[0018] The power supply device 4 supplies power to the sensors included in the sensor module 2, the pump 3, and the communication device 5. When the container 10 is loaded onto a trailer 17, for example, power may be supplied to the power supply device 4 from power supply equipment installed on the trailer 17 or a tractor towing the trailer 17. When the container 10 is loaded onto a container ship 18, for example, power may be supplied to the power supply device 4 from power supply equipment installed on the container ship 18.

[0019] The communication device 5 acquires measurement data from each sensor included in the sensor module 2 and transmits it to the computing device 8 via the network 7. The network 7 is not particularly limited, and the communication device 5 may communicate with the computing device 8, for example, via satellite communication, or may communicate with the computing device 8 using a wireless communication function provided in the trailer 17 or the container ship 18. It is desirable for the communication device 5 to transmit the acquired measurement data to the computing device 8 in real time. However, there are cases where communication with the computing device 8 is not possible, for example, when the cargo 6 is being transported by sea by the container ship 18. In such cases, the communication device 5 may hold the acquired measurement data until it is able to communicate with the computing device 8, and then transmit the held measurement data together to the computing device 8 when it is able to communicate with the computing device 8.

[0020] The computing device 8 is configured, for example, by a personal computer or a server computer. As will be described in detail later, it is preferable that the computing device 8 be capable of performing machine learning. The computing device 8 acquires measurement data from each sensor included in the sensor module 2. The computing device 8 analyzes the state of the cargo 6 based on the acquired odor data. Details will be described later with reference to FIGS. 4 , 5 , 6 , and 7 . The computing device 8 transmits the analysis results of the state of the cargo 6 to the monitoring device 9. The computing device 8 also transmits each measurement data, such as temperature data and humidity data, to the monitoring device 9. Although not shown in FIG. 1 , the functions of the computing device 8 are realized by cooperation between a processor and a memory included in the computing device 8. The computing device 8 may also have an interface for communicating with the communication device 5, the monitoring device 9, and the like. In the example of FIG. 1 , the computing device 8 acquires measurement data related to one container 10, but the computing device 8 may be capable of acquiring measurement data related to multiple containers.

[0021] The monitoring device 9 is configured, for example, by a personal computer or a server computer. The monitoring device 9 monitors the quality of the cargo 6. Specifically, the monitoring device 9 acquires the analysis results of the status of the cargo 6 by the computing device 8. As will be described in detail later, the monitoring device 9 determines whether to execute an alarm operation based on the analysis results of the status of the cargo 6. The alarm operation may, for example, notify a user, such as a person responsible for the transportation of the cargo 6 or a shipper of the cargo 6, of the status of the cargo 6. When the monitoring device 9 executes an alarm operation, it is considered that some abnormality has occurred in the quality of the cargo 6. When the monitoring device 9 executes an alarm operation, the monitoring device 9 performs a causal analysis of the abnormality in the quality of the cargo 6 based on measurement data such as temperature data and humidity data. The monitoring device 9 further executes an action such as canceling delivery of the cargo 6. Note that, although not shown in FIG. 1 , the functions of the monitoring device 9 are realized by cooperation between a processor and a memory included in the monitoring device 9. The monitoring device 9 may also have an interface for communicating with the computing device 8. Also, in the example of Figure 1, the monitoring device 9 monitors the quality of the cargo 6 stored in one container 10, but the monitoring device 9 may be capable of monitoring cargo stored in multiple containers.

[0022] In this way, the quality control system 1 analyzes the condition of the cargo 6 by sensing the odor of the cargo 6. Then, based on the analyzed condition of the cargo 6, the quality control system 1 determines whether or not an abnormality has occurred in the quality of the cargo 6, and if an abnormality has occurred, analyzes the cause of the abnormality and executes some kind of action. Therefore, it is assumed that the cargo 6 stored in the container 10 and transported is cargo that emits an odor. Examples of cargo 6 include fresh foods such as fruit, or liquids with an odor. For convenience of explanation, the cargo 6 will be assumed to be bananas.

[0023] The smell of a banana changes as it ripens. The color of its skin also changes as it ripens. Hereinafter, the color of the banana skin may be simply referred to as the "banana color." Figure 2 is a schematic diagram illustrating the color changes that occur as bananas ripen. As shown in Figure 2, bananas become more mature as the days pass. At their least ripe stage, a banana's color is all green. As the banana's ripeness increases, the color changes from all green to light green, light green to half green, half green to half yellow, half yellow to green tip, green tip to full yellow, full yellow to star, and star to dapple. The smell of a banana varies depending on the stage at which the banana is one of the eight colors listed above. In other words, the smell of the banana can determine which of the eight colors it is. Furthermore, the change in the smell of the banana can determine the change in the banana's color, i.e., the progress of the banana's ripening. The quality control system 1 senses the smell of the banana and analyzes its condition to detect changes in its ripeness if the banana is ripening. In this way, the quality control system 1 can monitor the quality of the bananas.

[0024] The odor of bananas is measured by an odor sensor 11. The odor sensor 11 is equipped with multiple types of molecular recognition materials that react to different odor components, or more precisely, specific odor molecules. The electrical properties of each molecular recognition material change when it adsorbs an odor molecule. Simply put, a current is generated when each molecular recognition material reacts with an odor component. The generated current is detected by a probe corresponding to each molecular recognition material. The odor sensor 11 is equipped with at least eight types of molecular recognition materials to measure the odor of bananas. The odor sensor 11 may be equipped with, for example, 16 types of molecular recognition materials to more accurately measure the odor of bananas.

[0025] If the molecular recognition material included in the odor sensor 11 continues to adsorb odor molecules, the odor molecules may adhere to the molecular recognition material, resulting in a decrease in the accuracy of odor measurement. Therefore, as described in the description of FIG. 1 , the pump 3 alternately sends fresh air and air containing the banana odor into the odor sensor 11 at specified time intervals. As a result, when the odor sensor 11 measures the banana odor, the odor molecules that have adhered to the molecular recognition material are removed from the molecular recognition material by the fresh air. In other words, the odor sensor 11 is cleaned of odors that have adhered to it. FIG. 3 is a graph illustrating the cleaning of odors that have adhered to the odor sensor according to embodiment 1.

[0026] 3, the vertical axis represents the current value, and the horizontal axis represents time. Characteristic 20, characteristic 21, and characteristic 22 respectively represent the change in current value when three different types of molecular recognition materials in the odor sensor 11 adsorb odor molecules. For convenience of explanation, only three characteristics, characteristic 20, characteristic 21, and characteristic 22, are shown. However, if the odor sensor 11 includes, for example, 16 types of molecular recognition materials, 16 patterns of characteristics representing the change in current value will also be generated.

[0027] As shown by characteristics 20, 21, and 22, periods T1 and T2 are alternately repeated during odor measurement by the odor sensor 11. The odor sensor 11 measures the odor of fresh air during period T1. At this time, the current value hardly changes. This is because fresh air is odorless. In other words, this is because fresh air does not contain odor molecules that react with the molecular recognition material provided in the odor sensor 11. During period T2, the odor sensor 11 measures the air inside the container 10, i.e., the banana odor. At this time, the current value increases. By repeating measurements during periods T1 and T2, the odor sensor 11 can maintain the accuracy of odor measurement. The lengths of periods T1 and T2 may be preset by, for example, the user.

[0028] [Odor Analysis] Next, the process of analyzing the state of bananas based on odor data by the computing device 8 will be described. The computing device 8 analyzes the state of the cargo 6 using a trained model based on training data for each type of cargo 6. The types of cargo 6 include, for example, bananas, apples, fish, etc. In the first embodiment, the computing device 8 analyzes the state of the bananas using the trained model.

[0029] The computing device 8 may have generated a trained model in advance. The trained model may be generated by a computer other than the computing device 8. The trained model analyzes, based on banana odor data, which of the eight states of bananas shown in FIG. 2 the odor of the banana being analyzed corresponds to. In other words, the trained model analyzes which of the eight states of bananas, all green, light green, half green, half yellow, full yellow, star, and dapple, the banana being analyzed, that is, the banana that is the source of the odor data, is in. Examples of analysis results will be described later with reference to FIG. 6.

[0030] The trained model analyzes odor data based on odor feature data. Feature data represents the characteristics of the odor data. Feature data differs depending on the state of the banana. For example, feature data extracted from the odor data of bananas that are all green will differ from feature data extracted from the odor data of bananas that are dappled.

[0031] Before describing the series of processes for odor analysis by the computing device 8, a brief description of the generation of a trained model will be given. For convenience of explanation, it is assumed that the generation of a trained model is performed by the computing device 8. The computing device 8 acquires odor data, for example, of a banana whose color, in other words, ripeness is all green. The computing device 8 extracts feature data from the odor data. For example, if the odor sensor 11 includes one each of 16 types of molecular recognition materials, the computing device 8 acquires 16 types of characteristics indicating changes in current value over time, as shown in FIG. 3 . In this case, the computing device 8 extracts, for each of the 16 types of characteristics, a specific value of the current value from a certain time to another time as feature data. The specific value is, for example, the maximum current value. In this way, the computing device 8 acquires feature data when the banana is all green. The computing device 8 acquires feature data not only when the banana is all green, but also when the banana is in each state.

[0032] The computing device 8 uses odor data stored together with labels representing states, such as all green, as training data. The computing device 8 trains a model using the training data, feature data, and an arbitrary algorithm. The model learns to classify each state based on the feature values ​​of the odor data. This generates a trained model. The algorithm used to generate the trained model is not particularly limited, and may be, for example, a logistic regression, a decision tree, a random forest, or the like.

[0033] The computing device 8 uses the generated trained model to execute each process in the flowchart shown in Fig. 4. The trained model may be stored in a memory of the computing device 8, or may be stored in an external device or the like that can communicate with the computing device 8. Fig. 4 is a flowchart of the process of the computing device 8 according to the first embodiment.

[0034] The calculation device 8 acquires the banana odor data generated by the odor sensor 11 (step St30). An example of the odor data will be described with reference to FIG.

[0035] FIG. 5 is a schematic diagram illustrating an example of odor data 40 according to the first embodiment. The odor data 40 includes a device ID 41, target item data 42, date data 43, and an odor data body 44. The device ID 41 is an ID for identifying the odor sensor 11. The target item data 42 is data indicating the target item that is the source of the odor data, i.e., the cargo 6, and is pre-registered in the odor sensor 11. In the example of FIG. 5, the target item data 42 indicates that the target item is a banana. The date data 43 is data indicating the year, month, day, hour, minute, and second when generation of the odor data 40 began, in other words, when odor measurement began. In the example of FIG. 5, the date data 43 indicates that odor measurement began at 9:00:00 on September 23, 2023. The odor data body 44 is digitized data of the measured odor. For example, if the odor sensor 11 is equipped with one of each of 16 types of molecular recognition materials, the odor data body 44 will include current values ​​detected by the probes corresponding to each molecular recognition material. Numerical data on the odor is recorded as the odor data body 44, for example, at specified time intervals. Specifically, the odor sensor 11 may record digitized data on the odor, such as current values, as the odor data body 44 every 1 / 100th of a second. Thus, for example, if odor measurement is started at 9:00:00 on September 23, 2023, the measurement results will be recorded in the odor data body 44 every 1 / 100th of a second starting from 9:00:00 on September 23, 2023.

[0036] Returning to the description of the flowchart in FIG. 4 , the computing device 8 extracts and acquires odor data bodies from the odor data bodies included in the odor data for a specified period starting from a reference time (step St31). The reference time is the year, month, day, hour, minute, and second indicated by the date data included in the odor data. Extracting odor data bodies for a specified period means, for example, extracting one hour's worth of odor data bodies from 9:00:00 on September 23, 2023 to 10:00:00 on September 23, 2023. This is because the amount of data contained in the odor data bodies can be enormous. The specified period may be preset by the user, for example.

[0037] If the computing device 8 is unable to extract the odor data body (step St32; NO), it terminates the processing flow. The case where the computing device 8 is unable to extract the odor data body is when all of the odor data body contained in the odor data has already been extracted by repeating the processing of step St31. Therefore, in this case, the computing device 8 terminates the processing flow.

[0038] If the computing device 8 has successfully extracted the odor data body (step St32; YES), it inputs the odor data acquired in step St30 into the trained model and has it analyze the state of the banana (step St33). More precisely, the computing device 8 extracts feature data from the odor data and has the trained model analyze the state of the banana based on the feature data.

[0039] The calculation device 8 acquires the analysis result based on the trained model (step St34). An example of the analysis result will be described with reference to FIG. 6.

[0040] FIG. 6 is a table diagram illustrating an example of an odor analysis result by the computing device 8 according to the first embodiment. The trained model outputs a match rate for each state of the banana. The match rate indicates the probability that the banana is in a certain state. In the example of FIG. 6, the state of the banana that is the source of the analyzed odor data is all green with a 30% probability. Furthermore, the states of the banana are light green with a 63% probability, half green with a 5% probability, half yellow with a 2% probability, green tip with a 0% probability, full yellow with a 0% probability, star with a 0% probability, and dapple with a 0% probability.

[0041] Returning to the description of the flowchart in Fig. 4, the calculation device 8 creates an odor analysis message (step St35) based on the odor data acquired in step St30 and the analysis results acquired in step St34. An example of the odor analysis message will be described later with reference to Fig. 7.

[0042] The calculation device 8 updates the reference time (step St36), returns to step St31, and repeats the process. For example, if the reference time, i.e., the year, month, day, hour, minute, and second when odor measurement started, is 9:00:00 on September 23, 2023, and the specified period for extracting the odor data main body is 30 minutes, the calculation device 8 updates the reference time to 9:30:00 on September 23, 2023. As will be described later with reference to FIG. 7 , by repeating the process of step St35, analysis result elements are added to the odor analysis message.

[0043] FIG. 7 is a schematic diagram illustrating an example of an odor analysis message 50 according to the first embodiment. The odor analysis message 50 includes date data 51 and, as analysis result elements, at least an analysis result element 52 and an analysis result element 53. The date data 51 is similar to the date data included in the odor data and indicates the year, month, day, hour, minute, and second when odor measurement began. In the example of FIG. 7 , the date data 51 indicates that odor measurement began at 9:00:00 on September 23, 2023. The analysis result elements include the device ID included in the odor data, target item data, and analysis results for a specified period. The analysis result element 52 indicates the match rate for each state of the banana from 9:00:00 on September 23, 2023 to 9:30:00 on September 23, 2023. The analysis result element 53 indicates the match rate for each state of the banana for the period from 9:30:00 on September 23, 2023 to 10:00:00 on September 23, 2023. In the example of FIG. 7 , the odor data body for 30 minutes from the reference time is extracted in step St31 of the flowchart shown in FIG. 4. One analysis result element is added to the odor analysis message each time the processing of step St35 of the flowchart shown in FIG. 4 is performed once.

[0044] [Alarm Operation] Next, we will explain the alarm operation that the monitoring device 9 executes based on the odor analysis message etc. after receiving the odor analysis message from the computing device 8. The monitoring device 9 executes the alarm operation when necessary by referring to the alarm level management table shown in Fig. 8 and the cargo management table shown in Fig. 9 in addition to the odor analysis message.

[0045] FIG. 8 is a table diagram illustrating an alarm level management table for bananas according to the first embodiment. The alarm level management table may be stored, for example, in the memory of the monitoring device 9. The alarm level management table indicates alarm levels corresponding to each state of the cargo 6. In the example of the alarm level management table for bananas shown in FIG. 8, alarm levels corresponding to each state of the banana are shown. The banana's ripeness level indicates the ripeness of the banana. When the banana's color is all green, it is 1, and increases by 1 as the color changes from all green to dapple. When the banana's color is dapple, the ripeness level is 8. The alarm level indicates the strength of the warning regarding the quality of the cargo 6; the higher the level, the stronger the warning. In the example of bananas, when the banana's color is all green and the ripeness level is 1, the alarm level is 0. As the banana ripens, changing from light green to half green and the ripeness level reaching 3, the alarm level becomes 2. As the banana ripens further and the ripeness level reaches 4 or higher, the alarm level becomes 3. Although the alarm levels are represented by "1," "2," and "3," the present invention is not limited to this and the alarm levels may be represented by a character string, for example.

[0046] 9 is a table diagram for explaining the cargo management table according to the first embodiment. The cargo management table may be stored in the memory of the monitoring device 9, for example. The cargo management table includes, as items, a container ID for identifying the container 10, a loading capacity of the container 10 to which the container ID is assigned (for example, 33.1 m), ３ ), the type of container 10, the name of the company that owns the container 10, the item of cargo 6, the load amount of cargo 6 (for example, 1,000 pieces), a device ID for identifying the odor sensor 11 installed in the container 10, and the alarm level of the cargo 6. The cargo management table may further include items other than those listed above.

[0047] Next, a description will be given of the processing of the monitoring device 9. Figures 10 and 11 are flowcharts of the processing of the monitoring device 9 according to embodiment 1. "A," "B," and "C" shown in the flowcharts of Figures 10 and 11 indicate the connection between the flowchart shown in Figure 10 and the flowchart shown in Figure 11.

[0048] The monitoring device 9 receives the odor analysis message from the computing device 8 (step St60).

[0049] The monitoring device 9 identifies the target item from the target item data in the odor analysis message received in step St60 (step St61). In the description of this flowchart, the target item is bananas.

[0050] The monitoring device 9 reads the alarm level management table corresponding to the target item identified in step St61 (step St62). The monitoring device 9 reads the alarm level management table corresponding to bananas shown in FIG.

[0051] The monitoring device 9 focuses on the first analysis result element among the analysis result elements included in the odor analysis message received in step St60 (step St63). In this specification, focusing on certain data or information means extracting and acquiring the data or information so that it can be processed in some way. The first analysis result element is the analysis result element that was added earliest to the odor analysis message. In the example of FIG. 7 , the first analysis result element among the analysis result elements included in the odor analysis message 50 is analysis result element 52, and the analysis result element next to analysis result element 52 is analysis result element 53.

[0052] When the monitoring device 9 can focus on an analysis result element (step St64; YES), the monitoring device 9 acquires a match rate for each state of the banana from the focused analysis result element (step St65).

[0053] If the monitoring device 9 is unable to focus on any analysis result element (step St64; NO), the processing flow ends. The case where the monitoring device 9 is unable to focus on any analysis result element occurs when the monitoring device 9 has finished focusing on all analysis result elements included in the odor analysis message received in step St60 by repeating the processing of step St73 described below.

[0054] Based on the match rate for each state of the banana obtained in step St65, the monitoring device 9 determines whether the match rate for at least one of half yellow, green tip, full yellow, star, and dapple is 1% or more (step St66).

[0055] First, let us consider the case where the match rate for at least one of Half Yellow, Green Tip, Full Yellow, Star, and Dapple is 1% or higher (step St66; YES). In this case, the monitoring device 9 determines that the alarm level for the analyzed banana is 3 (step St67).

[0056] The monitoring device 9 refers to the cargo management table and identifies the alarm level of the bananas that were the subject of analysis (step St68). The monitoring device 9 identifies the alarm level of the bananas listed on the cargo management table, for example, based on the device ID included in the odor analysis message.

[0057] The monitoring device 9 determines whether the alarm level recognized in step St67, step St75 described later, or step St77 described later is higher than the alarm level identified in step St68 (step St69).

[0058] If the monitoring device 9 determines that the alarm level recognized in step St67, step St75 described later, or step St77 described later is not higher than the alarm level identified in step St68 (step St69; NO), it proceeds to step St73 described later.

[0059] If the monitoring device 9 determines that the alarm level recognized in step St67, step St75 (described later), or step St77 (described later) is higher than the alarm level identified in step St68 (step St69; YES), the monitoring device 9 creates an alarm message (step St70). The alarm message will be described later with reference to FIG. 12.

[0060] The monitoring device 9 executes an alarm operation (step St71). The alarm operation may involve, for example, simultaneously notifying the transport operator or responsible party by email or the like that the alarm level has increased, i.e., that there has been a change in the condition of the bananas, or displaying an alarm message or the like on the display screen of a display device (not shown) provided in the monitoring device 9. Alternatively, the monitoring device 9 may notify another system (not shown) different from the quality control system 1 as the alarm operation. This allows the relevant parties or some other system to schedule unloading work or the like.

[0061] The monitoring device 9 rewrites the alarm level of the bananas that were the subject of analysis, which is listed on the cargo management table, to the alarm level recognized in step St67, step St75 described below, or step St77 described below (step St72).

[0062] The monitoring device 9 focuses on the analysis result element next to the analysis result element of interest among the analysis result elements included in the odor analysis message received in step St60 (step St73).The monitoring device 9 then returns to step St64 and repeats the process.

[0063] Next, the case where the match rate for each of Half Yellow, Green Tip, Full Yellow, Star, and Dapple is less than 1% (step St66; NO) will be described. In this case, the monitoring device 9 determines whether the match rate for Half Green is 10% or more (step St74).

[0064] If the match rate with respect to the half green is 10% or more (step St74; YES), the monitoring device 9 determines that the alarm level of the banana being analyzed is 2 (step St75). Then, the monitoring device 9 proceeds to step St68.

[0065] If the match rate for half green is less than 10% (step St74; NO), the monitoring device 9 determines whether the match rate for light green is 50% or more (step St76).

[0066] If the match rate with light green is 50% or more (step St76; YES), the monitoring device 9 determines that the alarm level of the banana being analyzed is 1 (step St77). Then, the monitoring device 9 proceeds to step St68.

[0067] If the match rate with light green is less than 50% (step St76; NO), the monitoring device 9 determines that the analyzed banana is normal with respect to the analysis result element of interest (step St78).Then, the monitoring device 9 proceeds to step St73.

[0068] The determinations in steps St66, St74, and St76 are an example of the case where the cargo 6 is bananas. The alarm levels recognized by the monitoring device 9 are three levels, "1," "2," and "3," but are not limited to these. The states of bananas and the threshold match rates for each state shown in FIGS. 10 and 11 for determining the alarm level are merely examples and are not intended to be limiting. For example, in step St74, the monitoring device 9 recognizes the alarm level as 2 when the match rate for half green is 10% or higher, but is not limited to this. For example, the monitoring device 9 may recognize the alarm level as 2 when the match rate for at least one of half green and light green is 20% or higher.

[0069] FIG. 12 is a schematic diagram illustrating an example of an alarm message 80 according to the first embodiment. The alarm message 80 includes a container ID 81, target item data 82, date data 83, alarm level data 84, and match rate data 85. The container ID 81 is the container ID of the container 10 in which the bananas to be analyzed are stored. The target item data 82 indicates that the target item is bananas. The date data 83 indicates the year, month, day, hour, minute, and second that is advanced by the next hour from the year, month, day, hour, minute, and second indicated by the date data included in the odor analysis message. This will be explained in detail below.

[0070] In step St31 of the flowchart shown in FIG. 4 , the computing device 8 extracted the odor data main body for a specified period from the odor data. Date data 83 indicates the year, month, day, hour, minute, and second indicated in the date data included in the odor analysis telegram, with the specified period advanced by the number of times the process has returned to step St64 of the flowchart shown in FIG. 10 . A specific example will be described using the odor analysis telegram 50 shown in FIG. 7 . In the example of FIG. 7 , the specified period for extracting the odor data main body from the odor data is 30 minutes. When the monitoring device 9 focuses on the analysis result element 53 included in the odor analysis telegram 50, the process returns to step St64 of the flowchart shown in FIG. 10 once. When the monitoring device 9 creates an alarm telegram while focusing on the analysis result element 53 included in the odor analysis telegram 50, the date data included in the alarm telegram is as follows: That is, the date data included in the alarm message indicates 9:30:00 on September 23, 2023, which is advanced by one 30 minutes from 9:00:00 on September 23, 2023, as indicated by the date data 51 included in the odor analysis message 50. The analysis result element 53 indicates the match rate for each state of the banana from 9:30:00 on September 23, 2023 to 10:00:00 on September 23, 2023. That is, the date data included in the alarm message indicates the approximate time when the alarm level increased, i.e., when a change occurred in the banana's state.

[0071] The alarm level data 84 indicates the alarm level recognized by the monitoring device 9. The match rate data 85 indicates the match rate for each state of the banana indicated by the analysis result element of interest.

[0072] In this way, the monitoring device 9 obtains the analysis results of the banana's condition, and determines whether or not to execute an alarm operation to notify the user of the banana's condition based on the alarm level corresponding to the banana's condition.

[0073] [Cause Analysis] Next, an analysis of the cause of the increase in the alarm level, i.e., the cause of the abnormality, performed when the monitoring device 9 executes an alarm operation will be described. The monitoring device 9 analyzes the cause of the abnormality based on the measurement data shown in Fig. 13 and the sensing table shown in Fig. 14.

[0074] 13A and 13B are table diagrams illustrating examples of measurement data according to the first embodiment. FIG. 13A is a table diagram illustrating temperature data and humidity data. FIG. 13B is a table diagram illustrating acceleration data. FIG. 13C is a table diagram illustrating illuminance data. FIG. 13D is a table diagram illustrating location information. Each measurement data is acquired by the calculation device 8 from each sensor. Then, the calculation device 8 transmits each measurement data to the monitoring device 9.

[0075] In the first embodiment, the acceleration sensor 14 measures acceleration along three orthogonal axes, namely, the X-axis, the Y-axis, and the Z-axis. The unit of acceleration is m / s. ２ In the first embodiment, the illuminance sensor 15 is configured by an element using cadmium sulfide (hereinafter referred to as "CdS"). When the illuminance is low, the resistance value of the illuminance sensor 15 becomes high. On the other hand, when the illuminance is high, the resistance value of the illuminance sensor 15 becomes low. Furthermore, the location information includes longitude, latitude, and altitude.

[0076] For example, from the table shown in FIG. 13(c), it can be seen that the resistance value of the illuminance sensor was low at 0:18 on August 19, 2023. In other words, it can be seen that the container 10 was unpacked at this time. This makes it possible to determine, for example, from the location information of the container 10, which transport base the container 10 was near at this time.

[0077] FIG. 14 is a table diagram illustrating an example of a sensing table according to the first embodiment. The sensing table may be stored in, for example, the memory of the monitoring device 9. The sensing table shows the priority and threshold for each of four data types: temperature data, humidity data, illuminance data, and acceleration data. Note that the units of each data type are omitted from the sensing table. The monitoring device 9 analyzes the presence or absence of anomalies by focusing on the data types in descending order of priority. Note that a higher priority refers to a smaller priority number. That is, temperature, with a priority of 1, has the highest priority, and acceleration, with a priority of 4, has the lowest priority. The range from the lower limit to the upper limit of the threshold for each data type shown in the sensing table is the normal value range for that data type. If an upper limit for the threshold is not specified, as in the case of illuminance data, the range above the lower limit is the normal value range. If a lower limit for the threshold is not specified, as in the case of acceleration data, the range below the upper limit is the normal value range. The monitoring device 9 determines that the temperature data is abnormal if the temperature indicated by the temperature data is below 10°C or above 15°C. Furthermore, if the humidity indicated by the humidity data is below 85% or above 90%, the monitoring device 9 determines that the humidity data is abnormal. If the resistance value indicated by the illuminance data is below 100 kΩ, the monitoring device 9 determines that the illuminance data is abnormal. If any of the accelerations on the X-axis, Y-axis, or Z-axis indicated by the acceleration data is above 20 m / s ２ If the acceleration data exceeds the threshold value, the acceleration data is determined to be abnormal. Note that these threshold values ​​are examples for the purpose of explanation. Also, for example, an upper limit value for the threshold value of the illuminance data may be specified, and a lower limit value for the threshold value of the acceleration data may be specified.

[0078] Next, the cause analysis process by the monitoring device 9 will be described. Fig. 15 is a flowchart of the process by the monitoring device 9 according to the first embodiment. At the start of the flowchart shown in Fig. 15, the monitoring device 9 is executing an alarm operation. In other words, the execution of the alarm operation triggers the start of the flowchart shown in Fig. 15.

[0079] The monitoring device 9 reads the sensing table (step St90). The monitoring device 9 reads the sensing table shown in FIG.

[0080] The monitoring device 9 focuses on the data type with the highest priority among the data types listed in the sensing table read in step St90 (step St91). The monitoring device 9 focuses on the temperature data.

[0081] The monitoring device 9 reads measurement data for the data type of interest for a specific period prior to the date and time the alarm was activated (step St92). More precisely, the monitoring device 9 reads measurement data for a specific period prior to the year, month, day, hour, minute, and second indicated by the date data included in the alarm message. The specific period may be, for example, one day or one hour. The alarm message 80 shown in FIG. 12 will be used as an example. For example, the monitoring device 9 reads one day's worth of temperature data prior to 9:30:00 AM on September 23, 2023, as indicated by the date data 83 included in the alarm message 80. In other words, the monitoring device 9 reads one day's worth of temperature data from 9:30:00 AM on September 22, 2023 to 9:30:00 AM on September 23, 2023.

[0082] The monitoring device 9 reads the threshold value for the data type of interest from the sensing table read in step St90 (step St93).

[0083] The monitoring device 9 focuses on the earliest data among the measurement data for the specific period read in step St92 (step St94). The earliest data will be described using the temperature data in the table diagram of temperature data and humidity data shown in FIG. 13(a) as an example. In the example of FIG. 13(a), the earliest temperature data is the data for 0:00 on August 19, 2023. Furthermore, the data for the time next to the data for 0:00 on August 19, 2023 is the data for 0:02 on August 19, 2023.

[0084] The monitoring device 9 determines whether the data under consideration exceeds the upper limit or falls below the lower limit of the threshold value read in step St93 (step St95).

[0085] If the monitoring device 9 determines that the data under consideration does not exceed the upper limit of the threshold value and does not fall below the lower limit of the threshold value (step St95; NO), the monitoring device 9 proceeds to the processing of step St97 described below.

[0086] If the monitoring device 9 determines that the data under consideration exceeds the upper limit of the threshold or falls below the lower limit (step St95; YES), it extracts the data under consideration as abnormal data (step St96).

[0087] The monitoring device 9 determines whether or not attention has been paid to the data for all times during the specific period acquired in step St92 (step St97). In other words, the monitoring device 9 determines whether or not determination of whether or not there is an abnormality based on the threshold has been completed for the data for all times during the specific period acquired in step St92.

[0088] If the monitoring device 9 determines that attention has not been paid to all the data for all the time periods during the specific period acquired in step St92 (step St97; NO), the monitoring device 9 focuses on the data for the time period following the data for which attention has been paid (step St98). Then, the monitoring device 9 returns to step St95 and repeats the process.

[0089] If the monitoring device 9 determines that it has completed looking at the data for all times during the specific period acquired in step St92 (step St97; YES), it determines whether it has looked at all data types included in the sensing table (step St99).

[0090] If the monitoring device 9 determines that it has not focused on all data types included in the sensing table (step St99; NO), it focuses on a data type that is one level lower in priority than the data type that it is currently focused on (step St100). As shown in the sensing table in Fig. 14, for example, the data type that is one level lower in priority than temperature data is humidity data. Then, the monitoring device 9 returns to step St92 and repeats the process.

[0091] If the monitoring device 9 determines that it has considered all data types included in the sensing table (step St99; YES), it creates a cause message and notifies the user (step St101). Then, the monitoring device 9 ends this processing flow. The method of notifying the user is not particularly limited, and may be, for example, sending an email or displaying on a display device. The cause message will be described later with reference to FIG. 16.

[0092] Fig. 16 is a schematic diagram for explaining an example of a cause message according to embodiment 1. In addition to the cause message 110, Fig. 16 also shows a schematic diagram 120 for explaining the time sequence of the occurrence of an abnormality cause and the alarm operation.

[0093] The cause message 110 includes a container ID 111, target item data 112, and alarm level data 113. These data are the data included in the alarm message added to the cause message 110 as is. The cause message 110 also includes the abnormality data extracted in step St96 of the flowchart shown in FIG. 15. In the example of FIG. 16, the cause message 110 includes abnormality data 114, abnormality data 115, and abnormality data 116. The abnormality data 114 is abnormality data regarding temperature. The abnormality data 115 is abnormality data regarding acceleration. The abnormality data 116 is abnormality data regarding illuminance. Note that if no abnormality data is extracted, the cause message does not include any abnormality data.

[0094] The schematic diagram 120 includes a graph 121 representing humidity changes, a graph 122 representing temperature changes, a graph 123 representing acceleration changes, and a graph 124 representing illuminance changes. The schematic diagram 120 also includes dates to explain the time series of changes in the measurement data represented by each graph.

[0095] The color of the bananas being transported was all green until time P1 on August 19, at which point they turned light green. This information was obtained by measuring and analyzing the odor. Furthermore, as the bananas became more ripe, the monitoring device 9 recognized the alarm level as 1 and created an alarm message. The monitoring device 9 then acquired various measurement data for a specific period prior to time P1, performed a causal analysis, and extracted abnormal data 114, 115, and 116. From the schematic diagram 120, it can be seen that graphs 122, 123, and 124 each fluctuate significantly around the times indicated by the abnormal data 114, 115, and 116, respectively. Based on the causal message 110, the monitoring device 9 can make the following judgments: That is, the monitoring device 9 can determine that some kind of impact was applied to the container 10 at around 8:00 PM on August 18th, that the container 10 was opened during transportation at around midnight on the 19th to check safety, that this caused the temperature inside the container 10 to rise, and that this temperature rise caused the bananas to ripen. Note that the monitoring device 9's determination that some kind of impact was applied to the container 10 is based on abnormal acceleration, that the container 10 was opened is based on abnormal illuminance, and that the temperature rose is based on abnormal temperature.

[0096] [Execution of Action] After the cause analysis, the monitoring device 9 executes an action for a change in the state of the bananas, i.e., a possible abnormality in the quality of the bananas, based on the alarm level and the abnormality factor. The monitoring device 9 executes a countermeasure action for quality control based on the cargo destination management table shown in Fig. 17, the destination-specific tolerance level management table shown in Fig. 18, and the transportation base management table shown in Fig. 19.

[0097] 17 is a table diagram illustrating an example of a cargo destination management table according to the first embodiment. The cargo destination management table may be stored, for example, in the memory of the monitoring device 9. The cargo destination management table includes, as items, a container ID, a destination of the container 10 to which the container ID is assigned, a type of the destination, an address of the destination, and a predicted arrival date and time at the destination. Hereinafter, the predicted arrival date and time may be referred to as a scheduled arrival date and time.

[0098] FIG. 18 is a table diagram illustrating an example of a destination-specific tolerance level management table according to the first embodiment. The destination-specific tolerance level management table may be stored, for example, in the memory of the monitoring device 9. The destination-specific tolerance level management table indicates the allowable ripeness level of bananas determined according to the type of destination. In the example of FIG. 18 , when the type of destination is a supermarket, bananas up to a ripeness level of 2 are allowed. That is, supermarkets only accept bananas up to a ripeness level of 2, i.e., bananas that are all green or light green in color. Furthermore, when the type of destination is a confectionery manufacturer, bananas up to a ripeness level of 5 are allowed. For ease of explanation, the destination-specific tolerance level management table illustrated in FIG. 18 only illustrates examples of a supermarket and a confectionery manufacturer. Note that the destination-specific tolerance level management table illustrated in FIG. 18 is a table for when the cargo 6 is bananas, and tables corresponding to the types of cargo 6 may be prepared.

[0099] FIG. 19 is a table diagram illustrating an example of a transportation base management table according to the first embodiment. The transportation base management table may be stored, for example, in the memory of the monitoring device 9. The transportation base management table includes, as items, a company ID for identifying the company in charge of transportation, in other words, the transportation base, the company name, the latitude and longitude of the company, the company address, contact information for the company's person in charge, and the company's evaluation. The company evaluation may be determined, for example, by numbers such as "8," "9," and "10," as shown in the example of FIG. 19. In the first embodiment, the larger the number, the higher the evaluation. Furthermore, the maximum number used for evaluation is 10.

[0100] 20 and 21, a process for executing a corrective action for quality control by the monitoring device 9 will be described. FIGS. 20 and 21 are flowcharts of the process of the monitoring device according to the first embodiment. "D" and "E" shown in the flowcharts of FIGS. 20 and 21 indicate the connection between the flowchart shown in FIG. 20 and the flowchart shown in FIG. 21. At the start of the flowchart shown in FIG. 20, a cause message has been created by the monitoring device 9. In other words, the creation of the cause message triggers the start of this processing flow.

[0101] The monitoring device 9 reads the created cause message (step St130).

[0102] The monitoring device 9 determines whether or not the cause message read in step St130 includes abnormal data (step St131).

[0103] If the cause message does not contain any abnormal data (step St131; NO), the monitoring device 9 sends an instruction to cancel the transport to the user or the transport system (step St142) and an instruction to reorder to the user or the transport system (step St143). The monitoring device 9 then terminates this processing flow. This is because the alarm level rose even though no abnormal data was contained, and the monitoring device 9 determines that the bananas being transported are defective. The transport system may be, for example, a system that manages the ordering and receiving of cargo 6. When sending the reorder instruction, the monitoring device 9 obtains the loaded items and load amount of the cargo 6 to be reordered from the cargo management table based on, for example, the container ID, and obtains the destination of the cargo 6 whose transportation was canceled from the cargo destination management table. The monitoring device 9 then transmits an instruction to the ordering and receiving system to retransport the same items and the same quantity of the cargo 6 whose transportation was canceled to the same destination.

[0104] If the error message contains abnormal data (step St131; YES), the monitoring device 9 determines whether the error cause is temperature or humidity (step St132). That is, the monitoring device 9 determines whether the error data contains temperature data, humidity data, or both.

[0105] If the monitoring device 9 determines that the abnormality factor is temperature or humidity (step St132; YES), it controls the air conditioner in the container 10 (step St133). Then, the monitoring device 9 proceeds to step St134, which will be described later.

[0106] If the monitoring device 9 determines that the abnormality factor is not temperature or humidity (step St132; NO), in other words, if it determines that the abnormality factor is acceleration or illuminance, it proceeds to step St135 described below.

[0107] The monitoring device 9 determines whether the abnormality factors include acceleration or illuminance (step St134).

[0108] If the monitoring device 9 determines that the abnormality factors do not include acceleration or illuminance (step St134; NO), the monitoring device 9 proceeds to step St137, which will be described later.

[0109] When the monitoring device 9 determines that the cause of the abnormality includes acceleration or illuminance (step St134; YES), the monitoring device 9 acquires the position information of the time of the abnormal data (step St135).

[0110] The monitoring device 9 determines whether or not a transportation base station close to the location indicated by the latitude and longitude of the location information acquired in step St135 is included in the transportation base station management table (step St136).

[0111] If the monitoring device 9 determines that a transportation base station close to the location indicated by the latitude and longitude of the location information acquired in step St135 is not included in the transportation base station management table (step St136; NO), it proceeds to step St139 described below.

[0112] If the monitoring device 9 determines that a transportation base close to the location indicated by the latitude and longitude of the location information acquired in step St135 is included in the transportation base management table (step St136; YES), the monitoring device 9 notifies the person in charge of the transportation base of the date and time of the abnormality occurrence and its cause (step St137). Alternatively, if the monitoring device 9 determines in step St134 that acceleration or illuminance is not included in the abnormality causes, the monitoring device 9 notifies the person in charge of the transportation base who set up the air conditioner of the container 10 of the date and time of the abnormality occurrence and its cause (step St137).

[0113] The monitoring device 9 lowers the evaluation of the transportation base to which the date and time of the abnormality occurrence and its cause were notified in step St137, and reflects this in the transportation base management table (step St138).

[0114] The monitoring device 9 determines whether the alarm level indicated by the cause message read in step St130 is level 1, level 2, or level 3 (step St139).

[0115] If the alarm level is 1 (step St139; level 1), the monitoring device 9 ends this processing flow. This is because, if the alarm level is 1, the monitoring device 9 determines that there is no problem with the quality of the bananas.

[0116] If the alarm level is 2 (step St139; level 2), the monitoring device 9 executes a process of switching the destination of the bananas (step St140). Details will be described later with reference to FIG.

[0117] If the switching of the destination of the bananas has been successful or if switching is not necessary (step St141; YES), the monitoring device 9 ends this processing flow.

[0118] If the monitoring device 9 fails to change the destination of the bananas (step St141; NO), it sends an instruction to cancel the transportation to the user or the transportation system (step St142), and sends an instruction to reorder to the user or the transportation system (step St143).Then, the monitoring device 9 ends this processing flow.

[0119] If the alarm level is 3 (step St139; level 3), the monitoring device 9 sends an instruction to cancel the transport to the user or the transport system (step St142), and sends an instruction to reorder to the user or the transport system (step St143).Then, the monitoring device 9 ends this processing flow.

[0120] The process of switching the destination of bananas by the monitoring device 9 will be described with reference to Fig. 22. Fig. 22 is a flowchart of the process of the monitoring device according to the first embodiment.

[0121] The monitoring device 9 acquires the destination type of the bananas from the cargo destination management table based on the container ID of the bananas to be swapped (step St150). A specific example will be described with reference to FIG. 23 . FIG. 23 is a schematic diagram for explaining an example of swapping destinations according to the first embodiment. In the example of FIG. 23 , bananas with a container ID of "2847469" are the bananas to be swapped, i.e., bananas with an alarm level of 2. The destination type of the bananas to be swapped is a supermarket. In the example of FIG. 23 , the destination of the bananas with a container ID of "2847469" is swapped with the destination of the bananas with a container ID of "837498233."

[0122] Based on the banana transport destination type acquired in step St150, the monitoring device 9 acquires the banana ripeness level acceptable for that transport destination type from the banana transport destination-specific tolerance level management table (step St151). The transport destination type of the bananas to be replaced is a supermarket. As shown in the example of FIG. 18, the banana ripeness level acceptable for the supermarket is 2. Since the alarm level of the banana to be replaced is 2, the current ripeness level of the banana is 3, and its color is half green. More precisely, the current ripeness level is the ripeness level at the date and time of alarm detection, i.e., the date and time indicated by the date data in the alarm message.

[0123] The monitoring device 9 acquires the predicted arrival date and time of the bananas to be replaced from the cargo destination management table based on the container ID (step St152). In the example of Fig. 23, the predicted arrival date and time of the bananas to be replaced is 15:00 on August 20, 2023.

[0124] The monitoring device 9 uses known techniques to predict the color of the banana to be replaced at the predicted arrival date and time and the number of days it will take for the banana's current ripeness level to rise to each higher ripeness level (step St153). In the example of FIG. 23 , the monitoring device 9 predicts the color of the banana to be replaced at 15:00 on August 20, 2023. Furthermore, the monitoring device 9 predicts the number of days until the current color of the banana to be replaced, i.e., half green, changes to half yellow, green tip, full yellow, star, and dapple. The monitoring device 9 may, for example, use the banana's odor data, the odor data acquisition date and time, the harvest date, the temperature data, the humidity data, and the color of the banana to generate a trained model that predicts the color of the banana on a specific day after harvest under a specific temperature and humidity. By inputting odor data into the trained model, the monitoring device 9 can predict the color of the banana that is the source of the odor data and the number of days until the banana changes to a specified color.

[0125] For ease of explanation, assume that the maturity level of the banana to be replaced at 15:00 on August 20, 2023 is predicted to be 3 and its color to be half green. Also assume that the current color of the banana to be replaced, half green, is predicted to change to green tip in four days. Because the alarm detection date is August 19, 2023, the color of the banana to be replaced will be green tip on August 23, 2023.

[0126] The monitoring device 9 determines whether the maturity level permitted by the destination of the bananas to be replaced is equal to or higher than the maturity level at the date and time when the bananas arrive at the destination (step St154).

[0127] If the monitoring device 9 determines that the maturity level acceptable to the destination of the bananas to be replaced is equal to or higher than the maturity level at the date and time the bananas arrive at the destination (step St154; YES), it determines that replacement is not necessary and terminates this processing flow.

[0128] If the monitoring device 9 determines that the ripeness level acceptable by the destination of the bananas to be replaced is lower than the ripeness level at the date and time the bananas arrive at the destination (step St154; NO), the processing proceeds to step St155. In the example of FIG. 23 , the ripeness level acceptable by the destination of the bananas to be replaced, i.e., the supermarket, is 2. That is, the color acceptable by the supermarket is light green. As predicted in step St153, the color of the bananas to be replaced at the supermarket is half green, and the ripeness level is 3. Therefore, the monitoring device 9 determines that the ripeness level acceptable by the destination of the bananas to be replaced is lower than the ripeness level at the date and time the bananas arrive at the destination, and proceeds to step St155.

[0129] The monitoring device 9 refers to the cargo management table and determines whether there is a container storing the same type and amount of cargo as the cargo 6 to be exchanged (step St155).

[0130] If the monitoring device 9 refers to the cargo management table and determines that there is no container containing the same item and load amount of cargo as the cargo 6 to be replaced (step St155; NO), it determines that the replacement has failed and terminates this processing flow.

[0131] When the monitoring device 9 determines, with reference to the cargo management table, that there is a container storing cargo of the same item and the same load as the cargo 6 to be replaced (step St155; YES), it selects the destination of the container as a candidate destination for the cargo 6, and proceeds to step St156. For the sake of convenience, it is assumed that the monitoring device 9 determines, with reference to the cargo management table, that a container with a container ID of "837498233" is loaded with the same amount of bananas as the bananas to be replaced. In this case, the monitoring device 9 selects "CakeMaker," which is the destination of the container with the container ID of "837498233," as a candidate destination.

[0132] The monitoring device 9 refers to the cargo destination management table and determines whether the candidate destination satisfies the replacement condition (step St156). That is, the monitoring device 9 determines whether "CakeMaker", the destination of the container with the container ID "837498233", satisfies the replacement condition. The replacement condition is, for example, the following four conditions (A1) to (A4).

[0133] (A1) The maturity level acceptable to the candidate destination is equal to or higher than the maturity level of the bananas to be replaced at the predicted arrival date and time when the bananas to be replaced are transported to the candidate destination.

[0134] (A2) The maturity level of the bananas scheduled to be transported to the candidate destination at the predicted arrival date and time when the bananas scheduled to be transported to the destination of the bananas to be replaced is equal to or lower than the maturity level permitted by the destination of the bananas to be replaced.

[0135] (A3) The predicted arrival date and time of the bananas to be replaced when transported to a candidate destination is before the date and time obtained by adding the number of days until the color of the bananas to be replaced changes to the banana color acceptable to the candidate destination to the alarm detection date and time.

[0136] (A4) Among the candidate destinations that satisfy the three conditions (A1) to (A3), the candidate destination is the destination with the earliest predicted arrival date and time if the bananas scheduled for transport are transported to the destination of the bananas to be swapped.

[0137] Condition (A1) is a condition for avoiding the situation where, for example, even if the bananas to be replaced are transported to "CakeMaker", the bananas have ripened to the point where "CakeMaker" cannot accept them when they arrive at "CakeMaker".

[0138] Condition (A2) is a condition to avoid, for example, the situation where, even if bananas intended for transport to "CakeMaker" are transported to "SuperX," the bananas have ripened to the point where "SuperX" cannot accept them when they arrive at "SuperX."

[0139] The condition (A3) is a condition for avoiding the following cases, for example.

[0140] In step St153, it was predicted that it would take four days for the current color of the banana to be replaced, half green, to change to green chip, the color acceptable by "CakeMaker." That is, since the alarm detection date was August 19, 2023, the banana to be replaced would change to green chip by August 23, 2023. Even if the banana to be replaced is transported to "CakeMaker," if the banana arrives at "CakeMaker" after August 23, the banana may have ripened further, causing "CakeMaker" to be unable to accept the banana. To avoid such a scenario, the predicted arrival date and time of the banana to be replaced when transported to a candidate destination must be before the alarm detection date and time plus the number of days required for the color of the banana to be replaced to change to the color acceptable by the candidate destination. For simplicity, the time of day is omitted from the date and time.

[0141] The condition (A4) is a condition for narrowing down the candidates for the transport destination to one when the monitoring device 9 selected multiple candidates for the transport destination in step St155.

[0142] In addition, the monitoring device 9 may use the known technology used in step St153, for example, a trained model generated in advance, to determine whether the candidate transport destination satisfies the replacement condition.

[0143] If the monitoring device 9 determines that the candidate transport destination does not satisfy the replacement condition (step St156; NO), it determines that the replacement has failed and ends this processing flow.

[0144] If the monitoring device 9 determines that the destination candidate satisfies the switching condition (step St156; YES), it modifies the cargo destination management table (step St157). Then, the monitoring device 9 determines that the switching is successful and terminates this processing flow. In the example of FIG. 23 , the monitoring device 9 modifies the cargo destination management table so that the container ID "2847469" and the container ID "837498233" are swapped. Also, in the example of FIG. 23 , the predicted arrival dates and times are not modified. That is, each cargo is transported according to its original predicted arrival date and time. In this way, even if the destination of the cargo 6 is switched, it is preferable that the predicted arrival date and time of each cargo does not change. Furthermore, by modifying the cargo destination management table by the monitoring device 9, an instruction to switch the destination of the cargo 6 may be sent to a user, a transportation system, or the like. In this way, when an alarm operation is executed, the monitoring device 9 analyzes the cause of the abnormality of the cargo 6 and executes an action based on the alarm level and the abnormality cause. The monitoring device 9 executes the action, for example, changing the destination of the cargo 6, reordering the cargo 6, or notifying the cause of the abnormality. Here, changing the destination of the cargo 6 includes switching the destination of the cargo 6 with the destination of another cargo, canceling the transportation of the cargo 6, returning the cargo 6 to the shipper, etc.

[0145] (Modification of First Embodiment) In the first embodiment described above, an example has been described in which the cargo 6 stored in the container 10 is bananas. However, this is not limiting, and the cargo 6 may also be a motorcycle. In this case, the alarm level management table and the destination-specific tolerance level management table may be, for example, as shown in Figures 24 and 25, respectively.

[0146] 24 is a table diagram illustrating an alarm level management table corresponding to oil according to a modification of the first embodiment. When the cargo 6 is a motorcycle, the alarm level is determined based on whether or not there is an oil smell from the motorcycle. When there is no oil smell, i.e., when there is no oil leakage, the odor level indicating the degree of odor is 0, and the alarm level is also 0. On the other hand, when there is an oil smell, i.e., when oil is leaking, the odor level is 1, and the alarm level is also 1.

[0147] 25 is a table diagram illustrating an example of a destination-specific tolerance level management table according to a variation of the first embodiment. When the cargo 6 is a motorcycle, if the destination type is a personal importer, only an odor level of 0 is permitted. On the other hand, if the destination type is a dealer, manufacturer, or other type other than a personal importer, an odor level of 1 is also permitted. This is because dealers and manufacturers can repair oil leaks themselves.

[0148] Fig. 26 is a schematic diagram for explaining the time series of alarm operation and abnormality occurrence according to a modification of embodiment 1. Schematic diagram 160 shown in Fig. 26 includes graph 161 representing humidity change, graph 162 representing temperature change, graph 163 representing acceleration change, and graph 164 representing illuminance change. Schematic diagram 160 also includes dates for explaining the time series of changes in measurement data represented by each graph. For ease of explanation, only dates are shown, with the times omitted.

[0149] Oil from a motorcycle being transported leaks at date and time P2, and the odor of the oil is measured by odor sensor 11. Based on the odor data, calculation device 8 analyzes and determines that there is an oil odor. Furthermore, monitoring device 9 recognizes the alarm level as 1 and creates an alarm message. Then, monitoring device 9 acquires various measurement data from a specific period prior to date and time P2 and performs a causal analysis. Through this causal analysis, monitoring device 9 determines that some kind of impact was applied to container 10 on August 8, and that this impact caused the motorcycle oil to leak.

[0150] (Summary of First Embodiment) The above description of First Embodiment discloses at least the following techniques. Note that, in parentheses, examples of corresponding components in First Embodiment are shown, but the present invention is not limited to these.

[0151] (Technology 1) A quality control system (for example, quality control system 1) is a quality control system that controls the quality of cargo (for example, cargo 6) based on the odor inside a container (for example, container 10) in which the cargo is stored, and includes: a plurality of sensors that measure the condition inside the container and generate measurement data, including at least an odor sensor (for example, odor sensor 11) that measures the odor inside the container and generates odor data (for example, odor data 40); a calculation device (for example, calculation device 8) that analyzes the condition of the cargo based on the odor data; and a monitoring device (for example, monitoring device 9) that determines whether to execute an alarm operation to notify the user of the cargo condition based on an alarm level corresponding to the analysis result of the cargo condition, and if the alarm operation is executed, analyzes an abnormality factor based on the measurement data and a threshold value corresponding to the measurement data, and executes an action based on the alarm level and the abnormality factor.

[0152] This allows the quality control system to analyze the condition of the cargo based on the cargo's odor. Furthermore, the quality control system can execute an alarm operation based on the analyzed cargo condition. This allows users, such as transportation personnel, to recognize that an abnormality may have occurred in the quality of the cargo during transport. Furthermore, the quality control system can analyze the cause of an abnormality that may have occurred in the cargo based on measurement data related to the condition inside the container. Furthermore, the quality control system can execute quality control actions based on the analyzed abnormality cause and the cargo alarm level. In this way, by analyzing the cargo condition based on odor, it is possible to manage the quality of the cargo without unpacking the container. Furthermore, because there is no need to unpack the container, the quality control system can manage the quality of the cargo even when there are a large number of containers.

[0153] (Technology 2) In the quality control system described in Technology 1, the quality control system further includes a pump (e.g., pump 3) that sends air to the odor sensor, and the pump sends the air inside the container and fresh air to the odor sensor alternately at specified time intervals, and the odor sensor alternately measures the odor of the cargo and the odor of the fresh air at specified time intervals.

[0154] This allows odor components from cargo that have adhered to the odor sensor to be removed periodically, thereby preventing the accuracy of odor measurement by the odor sensor from decreasing.

[0155] (Technology 3) In the quality control system described in Technology 1 or 2, the monitoring device predicts the condition of the first cargo at the scheduled arrival date and time based on the scheduled arrival date and time of the first cargo at the first destination, and determines whether or not to transport the first cargo to the first destination based on the prediction result and the acceptable range of cargo condition specified for each cargo destination.

[0156] This allows the monitoring device to predict the condition of the first cargo at the scheduled arrival date and time at the first destination. Furthermore, the monitoring device can determine that the first cargo cannot be transported to the first destination based on the allowable range of cargo condition specified for each destination.

[0157] (Technology 4) In the quality control system described in Technology 3, when the monitoring device determines that the first cargo cannot be transported to the first destination, the monitoring device changes the destination of the first cargo or places a reorder.

[0158] As a result, if the monitoring device determines that the first cargo cannot be transported to the first destination based on the acceptable range of cargo condition specified for each destination, it can, for example, change the destination of the first cargo, cancel the transport, return the cargo, or reorder the cargo.

[0159] (Technology 5) In the quality control system described in Technology 4, when the destination of a first cargo is changed, the monitoring device selects a candidate destination as a second destination, predicts the expected arrival date and time when the first cargo is transported to the second destination, predicts the condition of the first cargo at the expected arrival date and time, and determines that the first cargo can be transported to the second destination based on the prediction result and an acceptable range, and predicts the expected arrival date and time when a second cargo scheduled to be transported to the second destination is transported to the first destination, predicts the condition of the second cargo at the expected arrival date and time, and if it determines that the second cargo can be transported to the first destination based on the prediction result and an acceptable range, swaps the first destination of the first cargo with the second destination of the second cargo.

[0160] This allows the monitoring device to change the destination of the first cargo from the first destination to the second destination, and switch the destination of the second cargo from the second destination to the first destination.

[0161] (Technology 6) In the quality control system described in Technology 4, the monitoring device selects multiple candidate destinations, predicts the expected arrival date and time for each of the multiple candidates when the cargo to be transported is transported to the first destination, and replaces the candidate with the earliest predicted expected arrival date and time with the first destination.

[0162] This allows the monitoring device to narrow down multiple candidate destinations for switching the destination of the first cargo to one.

[0163] (Technology 7) In the quality control system described in Technology 5, the items and load of the second cargo are the same as the items and load of the first cargo, respectively.

[0164] This allows the monitoring device to switch the destinations of containers containing the same cargo item and the same load.

[0165] (Technology 8) In the quality control system according to any one of technologies 1 to 7, the action includes changing the destination of the cargo.

[0166] This allows the quality control system to change the destination of the cargo for quality control of the cargo, i.e., to change the destination of the cargo, cancel the shipment, or return the shipment, for example.

[0167] (Technology 9) In the quality control system according to any one of techniques 1 to 8, the action includes reordering the cargo.

[0168] This allows the quality control system to reorder the cargo for quality control of the cargo.

[0169] (Technology 10) In the quality control system according to any one of techniques 1 to 9, the action includes notification of the cause of the abnormality.

[0170] This allows the quality control system to notify transportation-related parties and the like of the causes of any abnormalities that may have occurred in the cargo, in order to control the quality of the cargo.

[0171] (Technology 11) In the quality control system according to any one of Technologies 1 to 10, the plurality of sensors include a humidity sensor (e.g., humidity sensor 13), a temperature sensor (temperature sensor 12), an acceleration sensor (acceleration sensor 14), an illuminance sensor (illuminance sensor 15), and a GPS sensor (GPS sensor 16), and the measurement data includes humidity data, temperature data, and illuminance data inside the container, as well as acceleration data and position information of the container.

[0172] This allows the quality control system to analyze the causes of any abnormalities that may be occurring in the cargo based on the temperature, humidity, and illuminance data inside the container, as well as the acceleration data and position information of the container.

[0173] (Technology 12) In the quality control system according to any one of techniques 1 to 11, the monitoring device acquires measurement data for a specific period based on the time at which the alarm action is executed.

[0174] This allows the monitoring device to acquire measurement data for a period during which an abnormality is thought to have occurred.

[0175] (Technology 13) A quality control method for managing the quality of cargo based on the odor inside a container in which the cargo is stored includes measuring the odor inside the container to generate odor data, measuring the condition inside the container to generate measurement data, analyzing the condition of the cargo based on the odor data, determining whether to execute an alarm operation to notify the cargo condition based on an alarm level corresponding to the analysis result of the cargo condition, and if the alarm operation is executed, analyzing the cause of an abnormality based on the measurement data and a threshold value corresponding to the measurement data, and executing an action based on the alarm level and the cause of the abnormality.

[0176] This allows the quality control method to achieve the same effects as Technique 1.

[0177] The functions of the above-described embodiments can also be realized by supplying programs and applications for realizing the functions of the above-described embodiments to a system or device using a network or storage medium, etc., and having one or more processors in the computer of that system or device read and execute the programs.

[0178] Furthermore, the functions of the above-described embodiments may be realized by a circuit that realizes one or more functions (for example, an Application Specific Integrated Circuit (hereinafter referred to as "ASIC") or an FPGA).

[0179] Although the embodiments of the present disclosure have been described above with reference to the drawings, it goes without saying that the present disclosure is not limited to such examples. It is clear that a person skilled in the art can conceive of various modifications, alterations, substitutions, additions, deletions, and equivalents within the scope of the claims, and it is understood that these also naturally fall within the technical scope of the present disclosure. Furthermore, the components of the above-described embodiments may be combined in any manner as long as they do not deviate from the spirit of the invention.

[0180] The present disclosure is useful for a quality control system and a quality control method.

[0181] 1 Quality control system 2 Sensor module 3 Pump 4 Power supply unit 5 Communication device 6 Cargo 7 Network 8 Calculation device 9 Monitoring device 10 Container 11 Odor sensor 12 Temperature sensor 13 Humidity sensor 14 Acceleration sensor 15 Illuminance sensor 16 GPS sensor 17 Tractor 18 Container ship 40 Odor data 50 Odor analysis message 80 Alarm message 110 Cause message

Claims

1. A quality control system for managing the quality of goods based on the odor inside a container storing the goods, comprising at least an odor sensor that measures the odor inside the container to generate odor data, a plurality of sensors that measure the state inside the container to generate measurement data, a computing device that analyzes the state of the goods based on the odor data, and a monitoring device that determines whether to execute an alarm operation for notifying the state of the goods based on an alarm level corresponding to the analysis result of the state of the goods, and when the alarm operation is executed, analyzes an abnormal factor based on the measurement data and a threshold value corresponding to the measurement data, and executes an action based on the alarm level and the abnormal factor. Quality control system.

2. The quality control system according to claim 1, further comprising a pump for sending air into the odor sensor, wherein the pump alternately sends the air inside the container and fresh air to the odor sensor at a prescribed time interval, and the odor sensor alternately measures the odor of the goods and the odor of the fresh air at the prescribed time interval.

3. The monitoring device according to claim 1 predicts the state of the first goods at the scheduled arrival date and time at the first destination of the first goods based on the scheduled arrival date and time, and based on the prediction result and the allowable range of the state of the goods defined for each destination of the goods, determines whether the first goods can be transported to the first destination. Quality control system described.

4. When the monitoring device determines that the first goods cannot be transported to the first destination, the monitoring device changes the destination of the first goods or places a reorder. Quality control system described in claim 3.

5. When the monitoring device changes the destination of the first cargo, it selects candidates for the destination as the second destination, predicts the scheduled arrival date and time when the first cargo is transported to the second destination, predicts the state of the first cargo at the predicted scheduled arrival date and time, determines that the first cargo can be transported to the second destination based on the prediction result and the allowable range, and predicts the scheduled arrival date and time when the second cargo scheduled to be transported to the second destination is transported to the first destination, predicts the state of the second cargo at the predicted scheduled arrival date and time, and when it is determined that the second cargo can be transported to the first destination based on the prediction result and the allowable range, it exchanges the first destination of the first cargo and the second destination of the second cargo. The quality management system according to claim 4.

6. The monitoring device selects a plurality of candidates for the destination, predicts the respective scheduled arrival dates and times when the cargo scheduled to be transported to each of the plurality of candidates is transported to the first destination, and exchanges the candidate with the earliest predicted scheduled arrival date among the plurality of candidates with the first destination. The quality management system according to claim 4.

7. The item and the loading quantity of the second cargo are respectively the same as those of the first cargo. The quality management system according to claim 5.

8. The action includes changing the destination of the cargo. The quality management system according to claim 1.

9. The action includes reordering the cargo. The quality management system according to claim 1.

10. The action includes notification of the abnormal factor. The quality management system according to claim 1.

11. The plurality of sensors include a humidity sensor, a temperature sensor, an acceleration sensor, an illuminance sensor, and a GPS sensor. The measurement data includes humidity data, temperature data, and illuminance data inside the container, and acceleration data and position information of the container. The quality management system according to claim 1.

12. The monitoring device acquires the measurement data during a specific period based on the time when the alarm operation is executed. The quality management system according to claim 1.

13. A quality control method for controlling the quality of goods based on the odor inside a container storing the goods, the method comprising: measuring the odor inside the container to generate odor data; measuring the state inside the container to generate measurement data; analyzing the state of the goods based on the odor data; determining whether to execute an alarm operation for notifying the state of the goods based on an alarm level corresponding to the analysis result of the state of the goods; and when the alarm operation is executed, analyzing an abnormal factor based on the measurement data and a threshold value corresponding to the measurement data, and executing an action based on the alarm level and the abnormal factor. Quality control method.

Images