Estimation device, estimation system, and trained model generation device

Abstract

Reduces the workload when jam occurs. [Solution] The estimation device according to the present disclosure includes: a jam detection unit capable of detecting a jam caused by the paper on the transport path when the paper is transported along the transport path by the operation of the paper transport unit; an acquisition unit that acquires information indicating the state of the paper transport unit and the state of the paper on the transport path when a jam is detected by the jam detection unit; and an estimation unit that receives an estimation result of the cause of the jam from the learning model by inputting information indicating one or more of the state of the paper transport unit and the state of the paper on the transport path to a learning model that has been trained using machine learning to estimate the cause of the jam caused by the paper.

Description

Technical Field

[0001] The present invention relates to an estimation device, an estimation system, and a learned model generation device.

Background Art

[0002] Many image forming apparatuses are provided with a paper conveyance unit for conveying the paper to be printed. In such an image forming apparatus, a so-called jam (paper jam), in which the paper taken into the paper conveyance unit is jammed and the operation is interrupted, may occur.

[0003] In Patent Document 1, in order to reduce the downtime due to jam occurrence, an AI function is used to recognize a sign of jam occurrence from data (paper size, environmental information, etc.) collected from an image forming apparatus, and when jam occurrence is predicted, maintenance is performed in advance. Thus, in Patent Document 1, a technique of preventing jam occurrence by performing maintenance when jam occurrence is predicted has been proposed.

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, Patent Document 1 is a technique for performing maintenance before jam occurs, and when jam occurs, the same operations as in the prior art, such as investigating the cause of the jam, are required.

[0005] An embodiment of the present invention aims to provide an estimation device, an estimation system, and a learned model generation device that reduce the work load by estimating the cause of a jam when the jam occurs.

Means for Solving the Problems

[0006] To solve the above-mentioned problems, the estimation device according to the present invention comprises: a jam detection unit capable of detecting jams caused by paper on a transport path when paper is transported along the transport path by the operation of a paper transport unit; an acquisition unit that acquires information indicating the state of the paper transport unit and the state of the paper on the transport path when a jam is detected by the jam detection unit; and an estimation unit that receives an estimation result of the cause of the jam from a learning model by inputting information indicating one or more of the state of the paper transport unit and the state of the paper on the transport path to a learning model that has been trained using machine learning to estimate the cause of the paper jam. [Effects of the Invention]

[0007] According to embodiments of the present invention, it is possible to reduce the workload when jam occurs. [Brief explanation of the drawing]

[0008] [Figure 1] This is a schematic diagram of an image forming apparatus according to one embodiment. [Figure 2] This figure shows an example of an image forming system according to the first embodiment. [Figure 3] This figure shows an example of the hardware configuration of an image forming apparatus according to the first embodiment. [Figure 4] This figure shows an example of a hardware configuration of an information processing device applicable to a machine learning server according to the first embodiment. [Figure 5] This figure shows an example of the software configuration for each device constituting the image forming system according to the first embodiment. [Figure 6] This figure shows the table structure of the jam-resistant holding table according to the first embodiment. [Figure 7] This figure shows an example of a procedure for estimating the cause of jamming in an image forming system according to the first embodiment. [Figure 8] This figure shows an example of a procedure for estimating the cause of jamming in an image forming system according to the first embodiment. [Figure 9]This figure shows an example of a procedure for estimating the cause of jamming in an image forming system according to the first embodiment. [Figure 10] This is an explanatory diagram illustrating the causes of jams based on the paper size set in the print settings and the paper size calculated from the transport sensor's detection results. [Figure 11] This figure shows an example of a screen displayed on the display unit by the display control unit according to the first embodiment. [Figure 12] This is an explanatory diagram illustrating the causes of jams based on paper damage information. [Figure 13] This figure illustrates the structure of a machine learning model generated by a machine learning server according to the first embodiment. [Figure 14] This is an explanatory diagram illustrating the concept of machine learning by the machine learning unit of the machine learning server according to the first embodiment. [Figure 15] This is a flowchart illustrating the procedure for processing jams in an image forming apparatus according to the first embodiment. [Figure 16] This figure shows an example of the software configuration of an image forming apparatus according to the first embodiment. [Modes for carrying out the invention]

[0009] Hereinafter, embodiments of the estimation device, estimation system, and trained model generation device according to the present invention will be described in detail with reference to the attached drawings.

[0010] (First embodiment) The image forming apparatus according to this embodiment may, for example, be a multifunction full-color digital copier (MFP (Multifunction Peripheral / Product / Printer)) that forms color images using an electrophotographic method.

[0011] FIG. 1 is a schematic configuration diagram of an image forming apparatus according to an embodiment. The image forming apparatus 100 shown in FIG. 1 is formed by a paper feeding unit 14 into which a paper 12 used for image formation is loaded, an image forming unit 16 that forms an image on the paper 12 by an electrophotographic process, and a paper conveyance unit 18 that is responsible for conveying the paper 12.

[0012] The paper feeding unit 14 is provided with a plurality of paper feeding trays 30 (30A, 30B, 30C, 30D) and a manual feed tray that can stack and accommodate a large number of papers 12 respectively. The paper feeding unit 14 has a plurality of delivery rollers 54 that send out the paper 12 from each of the plurality of paper feeding trays 30 and the manual feed tray. The plurality of delivery rollers 54 may include separation rollers that separate the sent-out papers 12 one by one.

[0013] The paper conveyance unit 18 operates various roller pairs to convey the paper 12 along the conveyance path. For example, the paper conveyance unit 18 has a plurality of conveyance roller pairs 56, 58 that convey the paper 12 with respect to the paper delivery conveyance path 32 from the paper feeding unit 14 to the image forming unit 16.

[0014] Furthermore, the paper conveyance unit 18 has a plurality of conveyance roller pairs 60, 62 with respect to the registration conveyance path 34 that conveys the paper 12 conveyed through the paper delivery conveyance path 32 to the transfer position.

[0015] The paper 12 taken out from the paper feeding tray 30 is fed to the transfer position through the paper delivery conveyance path 32 and the registration conveyance path 34.

[0016] The print engine (image forming unit) 16 includes a transfer module 22 and a fixing module 26.

[0017] The transfer module 22 transfers the toner image formed on a photoreceptor or the like to the paper 12 conveyed by the paper conveyance unit 18.

[0018] The fixing module 26 fixes the toner image transferred to the paper 12 by the transfer module 22 using a fixing (pressing) roller 28.

[0019] The paper transport unit 18 is equipped with an output transport path 36 that transports the paper 12 fed from the fuser module 26 toward the outside of the machine. After the image is formed on the paper 12, it passes through the output transport path 36 and is sent to an output tray (not shown) via a pair of output rollers 64 that discharge the paper outside the image forming apparatus 100.

[0020] The paper transport unit 18 is equipped with transport rollers 66 and 68 on transport paths 38 and 40 that transport the paper 12 fed from the fuser module 26 back to the transfer position for double-sided printing.

[0021] Meanwhile, the paper transport unit 18 is equipped with multiple transport sensors 70 between the rollers in each transport path to detect the passage of the paper 12. The multiple transport sensors 70 are capable of detecting whether or not the paper 12 is being transported at the appropriate timing.

[0022] The transport sensor 70 may, for example, be a reflective photosensor that optically detects the presence of the paper 12. While the transport sensor 70 is detecting the paper 12, it outputs a detection signal to the CPU 1201 of the image forming apparatus 100, which will be described later.

[0023] The transport sensors 70 are provided at arbitrary intervals along the transport path of the paper transport unit 18. Therefore, the CPU 1201 can recognize the position of the paper 12 along the transport path based on the detection signals from the transport sensors 70. If a jam occurs on the transport path, the CPU 1201 can recognize the location of the jam based on the detection signals from the transport sensors 70. Therefore, the transport sensors 70 function as a jam detection unit capable of detecting jams caused by paper 12 on the transport path.

[0024] The paper transport unit 18 has a motor (not shown) that rotates multiple rollers, and a clutch (not shown) that connects the rollers and the motor. The various rollers, motor, clutch, and other devices in the paper transport unit 18 may deteriorate or change in performance over time. Deterioration or changes in performance of these devices over time may cause paper jams in the paper 12. Similarly, the paper transport unit 18 may also cause jams if the paper information set by the user (paper size, paper thickness, etc.) differs from the actual characteristics of the paper 12.

[0025] Traditionally, if a jam occurred in an image forming machine during the image forming process, the process would stop, and the machine would enter a jammed state. In this case, the user would remove the jammed paper from the image forming machine. This would restore the image forming machine to a normal state, and it would be available for image forming and other processes. However, if the cause of the paper jam is not removed, the image forming machine will repeatedly jam and become unusable every time it performs an image forming process. Thus, as long as the cause of the paper jam is not removed, there is a problem of prolonged downtime for the image forming machine.

[0026] Therefore, in the image forming apparatus 100 according to this embodiment, when a jam occurs due to the paper 12, the cause of the jam is estimated. The user can then take action to eliminate the cause of the jam according to the estimated result, thereby reducing downtime.

[0027] Figure 2 shows an example of an image forming system according to this embodiment. The image forming system 1 shown in Figure 2 includes an image forming device 100 such as a printer, multifunction device, or fax machine, a machine learning server 102, a data server 105, and a general-purpose computer 103 that transmits print data to the image forming device 100, etc. Each of the devices shown in Figure 2 is connected by a network NW such as a LAN. The network NW may be wired or wireless. Furthermore, the network NW may include a public communication line. For example, the machine learning server 102 and the data server 105 may be configured to communicate with the image forming device 100 via a public communication line.

[0028] The image forming apparatus 100 is equipped with AI (Artificial Intelligence) functionality and functions as an estimation device that enables estimation of the cause of jamming using this AI functionality. The image forming apparatus 100 is equipped with a machine learning model 1204A (see Figure 3) to realize the AI ​​functionality.

[0029] The data server 105 collects information from the image forming apparatus 100, etc., including the detection results of various sensors 1211 installed in the image forming apparatus 100 when a jam occurs, and information indicating the current settings of the image forming apparatus 100. The data server 105 then provides the collected information to the machine learning server 102.

[0030] The machine learning server 102 generates training data based on information provided by the data server 105, and uses some or all of the generated training data to generate a machine learning model 1204A for realizing AI functionality.

[0031] Figure 3 shows an example of the hardware configuration of the image forming apparatus 100 according to this embodiment. As shown in Figure 3, the image forming apparatus 100 is equipped with hardware components such as a CPU 1201, RAM 1202, ROM 1203, HDD 1204, network IF 1210, GPU 1221, various sensors 1211, and a display unit 1205.

[0032] The CPU 1201 is a control unit that comprehensively controls the image forming apparatus 100. The RAM 1202 is a system work memory for the operation of the CPU 1201 and also serves as image memory for temporarily storing image data, etc. The ROM 1203 stores programs, etc., that the CPU 1201 executes.

[0033] The HDD (Hard Disk Drive) 1204 stores the system software, image data, software counter values, etc., of the image forming apparatus 100. Furthermore, the HDD 1204 stores the machine learning model 1204A. The machine learning model 1204A will be described later. Note that other storage devices such as solid-state drives (SSDs) may be provided instead of or in conjunction with the HDD.

[0034] The network IF1210 is connected to the network NW and communicates with the general-purpose computer 103, the machine learning server 102, the data server 105, and other computer terminals (not shown) on the network NW. The network IF1210 may also communicate data with an external facsimile device (not shown). Furthermore, the network IF1210 includes wireless communication functionality for wirelessly connecting with external terminals.

[0035] The various sensors 1211 are sensors provided in the image forming apparatus 100 to detect the state of the paper transport unit 18 of the image forming apparatus 100 and the state of the paper 12 on the transport path. For example, the various sensors 1211 include a transport sensor 70 as a sensor for detecting the state of the paper transport unit 18 of the image forming apparatus 100. Furthermore, the various sensors 1211 include a camera 1212 installed on the transport path as a sensor for detecting the state of the paper 12 on the transport path. Furthermore, the various sensors 1211 include a temperature and humidity meter 1213 as a sensor for detecting the environment of the paper transport unit 18.

[0036] Camera 1212 is configured to capture images of the paper 12 moving along the transport path. For example, when a jam is detected, it captures images of the paper 12 on the transport path and transmits the captured image information to the CPU 1201. Camera 1212 functions as a paper condition detection unit that can acquire the damage status (degree of damage) of the paper 12 moving along the transport path by capturing images of the paper 12 on the transport path.

[0037] The temperature and humidity meter 1213 detects the temperature and humidity around the paper transport unit 18 as information indicating the environment in which the paper transport unit 18 is installed, and transmits the detection results to the CPU 1201. This embodiment describes an example in which the temperature and humidity meter 1213 is used as an example of an environment detection unit for detecting the environment around the image forming apparatus 100, but it is not limited to the temperature and humidity meter 1213, and may include, for example, acceleration for detecting ambient vibrations.

[0038] The sensors included in the various sensors 1211 are not limited to the transport sensor 70, camera 1212, and temperature / humidity meter 1213, but may include other sensors.

[0039] The display unit 1205 is a display device such as a display that shows an operation screen or a settings screen.

[0040] The GPU 1221 is hardware capable of performing mathematical calculations at high speed. By having the GPU 1221 perform processing, the image forming apparatus 100 can process more data in parallel. Therefore, by equipping the image forming apparatus 100 with the GPU 1221, it is possible to perform calculations efficiently. In this embodiment, the image forming apparatus 100 may also use the GPU 1221 for processing in the estimation unit 1516, which will be described later. Furthermore, in this embodiment, the image forming apparatus 100 may be configured to perform the processing of the estimation unit 1516 using only the CPU 1201 or the GPU 1221.

[0041] Figure 4 shows an example of the hardware configuration of an information processing device applicable to the machine learning server 102 according to this embodiment.

[0042] As shown in Figure 4, the machine learning server 102 comprises a CPU 1301, RAM 1302, ROM 1303, HDD 1304, network IF 1310, I / O unit 1305, and GPU 1306, and a system bus 1307 that connects them to each other.

[0043] The CPU 1301 provides various functions by reading and executing programs such as the OS (Operating System) and application software from the HDD 1304. The RAM 1302 is the system work memory used by the CPU 1301 when executing programs. The ROM 1303 stores programs for starting the BIOS (Basic Input Output System) and OS, as well as configuration files.

[0044] The HDD (Hard Disk Drive) 1304 stores system software and other data. Alternatively, other storage devices such as SSDs may be provided instead of or in conjunction with the HDD.

[0045] The network IF1310 is connected to the network NW and communicates with external devices such as the general-purpose computer 103, the data server 105, and the image forming apparatus 100.

[0046] The I / O unit 1305 is an interface for inputting and outputting information to an operation unit (not shown) consisting of a liquid crystal display equipped with a multi-touch sensor, etc. The I / O unit 1305 outputs screen information according to the program, and a screen according to the screen information is displayed on the liquid crystal display of the operation unit with a predetermined resolution and number of colors. For example, a GUI (Graphical User Interface) screen is displayed on the liquid crystal display of the operation unit. The GUI screen contains various windows or data necessary for operation. The I / O unit 1305 then receives operation information from the operation unit based on input to the multi-touch sensor and passes it to the CPU 1301. Note that the machine learning server 102 does not necessarily have to be provided with an operation unit.

[0047] The GPU1306 is hardware capable of performing mathematical calculations at high speed. By having the machine learning server 102 process data on the GPU1306, it becomes possible to process more data in parallel. Therefore, the image forming apparatus 100 can perform calculations efficiently by equipping it with the GPU1306.

[0048] For example, if the machine learning server 102 performs machine learning, such as deep learning, multiple times, it is effective to process the data using the GPU 1306.

[0049] In this embodiment, when the machine learning unit 1532 (see Figure 4), described later, performs machine learning, it uses the GPU 1306 in addition to the CPU 1301. Specifically, the CPU 1301 and GPU 1306 work together to perform machine learning on a neural network using training data to generate a machine learning model. Note that this embodiment is not limited to a method in which the CPU 1301 and GPU 1306 work together to perform machine learning; machine learning may also be performed using only the CPU 1301 or the GPU 1306.

[0050] The data server 105 and the general-purpose computer 103 can be implemented with the same hardware configuration as the machine learning server 102, and therefore their explanation is omitted. Furthermore, the machine learning server 102 and the data server 105 may be implemented on the same computer. Additionally, the image forming apparatus 100 may be provided with the same functions as the machine learning server 102 and the data server 105.

[0051] Furthermore, the machine learning server 102 and the data server 105 may each be implemented by a single computer or by multiple computers. In addition, the machine learning server 102 and the data server 105 may each be implemented using cloud computing technology.

[0052] Figure 5 shows an example of the software configuration for each device constituting the image forming system 1 according to this embodiment. Figure 5 shows the software configuration realized by utilizing the hardware resources and programs of each device shown in Figures 2 to 4, which are included in the image forming system 1 according to this embodiment.

[0053] The image forming system 1 according to this embodiment functions as an estimation system that estimates the cause of a jam when one occurs in the image forming apparatus 100. The image forming system 1 according to this embodiment uses AI functionality to estimate the cause of the jam. In order to utilize AI functionality, it is necessary to perform a learning phase and an inference phase. Therefore, in this embodiment, the machine learning server 102 performs the learning phase, and the image forming apparatus 100 performs the inference phase.

[0054] The program required to implement the software configuration shown in Figure 5 is stored in a storage device (e.g., an HDD) provided for each device. The CPU of each device then loads the stored program into RAM and executes it.

[0055] For example, in the image forming apparatus 100, the program is stored in the HDD 1204. The CPU 1201 then loads the stored program into the RAM 1202 and executes it. In addition to the CPU 1201, the stored program may also be executed using the GPU 1221.

[0056] Furthermore, the image forming apparatus 100 implements the following functions through program execution: JOB control unit 1511, image reading unit 1512, counter unit 1513, acquisition unit 1514, state detection unit 1515, estimation unit 1516, determination unit 1517, and display control unit 1518.

[0057] Furthermore, the HDD 1204 of the image forming apparatus 100 stores a machine learning model 1204A, a data storage unit 1502, and a jam-response holding table 1503.

[0058] Machine learning model 1204A is a machine learning model that has been trained to estimate the cause of the jam caused by paper 12.

[0059] The machine learning model 1204A takes the state of the paper transport unit 18, the state of the paper 12 on the transport path, and environmental information as input when a jam occurs, and outputs an estimated result of the cause of the jam. The specific machine learning procedure using training data will be described later.

[0060] The status of the paper transport unit 18 includes, for example, the unit status of the paper transport unit 18, component information of the paper transport unit 18, and information based on the detection results of the transport sensor 70, including transport sensor information that can identify, for example, the location where a paper jam (paper jam) has occurred.

[0061] The unit status of the paper transport unit 18 refers to information indicating the status of the units constituting the paper transport unit 18, for example, as detected by various sensors 1211. For example, the unit status includes the degree of wear of the units, which can be identified based on image information captured by the camera 1212. The component information of the paper transport unit 18 refers to information about the components constituting the paper transport unit 18, such as the date and time of the last replacement of the component, the number of times the component has been used, and the temperature of the component. The component information of the paper transport unit 18 is stored, for example, in the HDD 1204 or ROM 1203.

[0062] The status of the paper 12 on the transport path includes paper size information (detected by the transport sensor 70) indicating the size of the paper 12 that caused the jam (paper jam). Furthermore, the status of the paper 12 on the transport path includes the paper size set for printing or other operations on the paper 12 on the transport path.

[0063] The condition of the paper 12 on the transport path also includes paper damage information, which indicates the degree of damage to the paper 12 that caused the jam (paper jam), based on image information captured by the camera 1212.

[0064] Environmental information includes, for example, information indicating the temperature and humidity around the paper transport unit 18, as detected by the thermometer / hygrometer 1213.

[0065] The data storage unit 1502 is a storage area for storing detection results from various sensors 1211 provided in the image forming apparatus 100, and data (image data, etc.) input and output to the image forming apparatus 100.

[0066] The jam-response holding table 1503 according to this embodiment is used to resolve the cause of a jam when one occurs. Figure 6 shows the table structure of the jam-response holding table 1503 according to this embodiment. As shown in Figure 6, the jam-response holding table 1503 according to this embodiment stores the cause of the jam and the method for resolving the cause in association. As shown in Figure 6, if the cause of a jam that occurs in the image forming apparatus 100 is "incorrect paper size setting", the user can resolve the cause of the jam by "correctly setting the paper size" as the resolving method.

[0067] The JOB control unit 1511 executes the basic functions of the image forming apparatus 100, such as copying, faxing, and printing, according to user input. Furthermore, the JOB control unit 1511 has the function of issuing instructions between multiple software components and controlling the transmission and reception of data when executing the basic functions.

[0068] The image reading unit 1512 has the function of reading the original document using a reading unit (not shown) of the image forming apparatus 100 when a copy or scan function is executed based on instructions from the JOB control unit 1511. Specifically, the image reading unit 1512 controls the optical reading of the contents represented on the original document using an in-line sensor included in the reading unit.

[0069] The counter unit 1513 records and manages various counter values ​​(for example, the total number of printed pages) in the image forming apparatus 100.

[0070] The acquisition unit 1514 acquires detection results from various sensors 1211, etc. Furthermore, the acquisition unit 1514 acquires the current settings of the image forming apparatus 100. The detection results from the various sensors 1211, etc. include, for example, the detection result of the paper 12 by the transport sensor 70, image information of the paper 12 captured by the camera 1212, and the temperature and humidity measurement results by the thermometer / hygrometer 1213. The current settings include, for example, the paper size set by the user.

[0071] The state detection unit 1515 detects (acquires) the state of each component included in the image forming apparatus 100 based on the detection results of various sensors 1211 etc. acquired by the acquisition unit 1514 and the current settings. For example, the state detection unit 1515 detects whether or not a jam has occurred in the paper 12 based on the detection result of the transport sensor 70, and acquires the location where the jam occurred.

[0072] Furthermore, the state detection unit 1515 acquires the state of the paper transport unit 18, the state of the paper 12 on the transport path, and environmental information based on the detection results of various sensors 1211 etc. acquired by the acquisition unit 1514 and the current settings. The information indicating the state of the paper transport unit 18 includes the unit state of the paper transport unit 18, component information of the paper transport unit 18, and transport sensor information. The information indicating the state of the paper 12 on the transport path includes the paper size in the print settings, paper size information identified from the detection results of the transport sensor 70, and paper damage information. The environmental information includes the temperature and humidity around the paper transport unit 18.

[0073] The state detection unit 1515 then stores information indicating the state of the paper transport unit 18, information indicating the state of the paper 12 on the transport path, and environmental information in the data storage unit 1502. The information stored by the state detection unit 1515 is not limited to information indicating the state of the paper transport unit 18, information indicating the state of the paper 12 on the transport path, and environmental information, but may also include, for example, detection results from various sensors 1211, or various settings and status information of the image forming apparatus 100.

[0074] When a jam occurs in the image forming apparatus 100, the estimation unit 1516 inputs information stored by the state detection unit 1515 (for example, the unit status of the paper transport unit 18, component information of the paper transport unit 18, transport sensor information, paper size in the print settings, paper size information identified from the detection results of the transport sensor 70, paper damage information, and environmental information) into the machine learning model 1204A, and receives an estimation result of the cause of the jam from the machine learning model 1204A. The estimation unit 1516 performs estimation processing using the machine learning model 1204A, for example, based on instructions from the JOB control unit 1511. This embodiment describes an example in which the unit status of the paper transport unit 18, component information of the paper transport unit 18, transport sensor information, paper size in the print settings, paper size information identified from the detection results of the transport sensor 70, paper damage information, and environmental information are input to the machine learning model 1204A. However, this embodiment does not limit the input to all of this information, and it is also possible to input at least one or more of this information to the machine learning model 1204A.

[0075] The estimation unit 1516 implements an AI function to estimate the cause of jamming by performing estimation and classification processing using the machine learning model 1204A.

[0076] The decision unit 1517 determines a method for resolving the cause of jam based on the estimation result of the estimation unit 1516, in accordance with the instructions of the JOB control unit 1511. For example, the decision unit 1517 refers to the jam correspondence storage table 1503 to identify a method for resolving the cause that is associated with the estimated cause of jam.

[0077] The display control unit 1518 performs control to display information on the display unit 1205 according to the instructions of the JOB control unit 1511. For example, when a jam occurs, the display control unit 1518 displays information on the display unit 1205 that shows the cause of the jam estimated by the estimation unit 1516 and the method for resolving the cause determined by the determination unit 1517. In this embodiment, it becomes possible to provide feedback to the user by displaying information, etc.

[0078] Similarly, in the data server 105, the program is stored in the HDD 1304. The CPU 1301 then loads the stored program into RAM 1302 and executes it.

[0079] The data server 105 then implements the data collection unit 1521 and the data provision unit 1522 through program execution. Furthermore, the HDD 1304 of the data server 105 houses the data storage unit 1523.

[0080] The data acquisition unit 1521 receives (collects) data from the image forming apparatus 100, including user environment-specific information related to the image forming apparatus 100. User environment-specific information related to the image forming apparatus 100 includes, for example, data stored in the data storage unit 1502 of the image forming apparatus 100. The data acquisition unit 1521 receives (collects) data from one or more image forming apparatuses 100.

[0081] The data storage unit 1523 is a storage area for storing information collected by the data acquisition unit 1521.

[0082] The data provision unit 1522 transmits (provides) the information stored in the data storage unit 1523 (collected by the data collection unit 1521) to the machine learning server 102.

[0083] On the machine learning server 102, programs are stored on the HDD 1304. The CPU 1301 then loads the stored programs into RAM 1302 and executes them. In addition to the CPU 1301, the stored programs may also be executed using the GPU 1306.

[0084] The machine learning server 102 then implements the training data generation unit 1531 and the machine learning unit 1532 through program execution. Furthermore, the HDD 1304 of the machine learning server 102 houses the data storage unit 1533.

[0085] The data storage unit 1533 serves as a storage area for storing information received from the data server 105. The data storage unit 1533 also serves as a storage area for storing training data generated by the training data generation unit 1531.

[0086] The training data generation unit 1531 generates training data to be used for machine learning using the information stored in the data storage unit 1533 (received from the data server 105).

[0087] The training data generation unit 1531 removes noise data from the information stored in the data storage unit 1533 (received from the data server 105) in order to obtain the desired learning effect. Any method can be used for data removal, regardless of whether it is a well-known method.

[0088] Furthermore, the training data generation unit 1531 optimizes the training data by adjusting it according to the format of the data to be input to the machine learning model. For example, as an example of preprocessing for effective machine learning, the training data generation unit 1531 may extract the unit status of the paper transport unit 18, component information of the paper transport unit 18, transport sensor information immediately after jam occurrence, paper size in the print settings, paper size information identified from the detection results of the transport sensor 70, paper damage information, and environmental information from the information received from the data server 105, and process the extracted information to include it in the training data. By performing this extraction, it becomes possible to learn the cause of jams more efficiently.

[0089] Furthermore, the training data generation unit 1531 generates training data by associating the extracted information described above with information indicating the cause of the jam. Therefore, the training data associates the state of the paper transport unit 18, the state of the paper 12 on the transport path, environmental information, and the cause of the jam. The cause of the jam is information entered by the service provider or analyst who dealt with the jam.

[0090] The machine learning unit 1532 generates machine learning models by performing machine learning based on training data. The machine learning models are generated by applying supervised learning, based on training data, to the base neural network. The specific generation method will be described later.

[0091] The machine learning model generated by the machine learning server 102 is stored in the HDD 1204 of the image forming apparatus 100. The machine learning model is a type of computational algorithm and is modularized as part of the control program of the image forming apparatus 100 and stored in the HDD 1204.

[0092] Next, the procedure for estimating the cause of jamming in the image forming system 1 according to this embodiment will be described.

[0093] Figure 7 shows an example of a procedure for estimating the cause of a jam in the image forming system 1 according to this embodiment. Figure 7 shows a service representative who visits the user and performs maintenance when a jam occurs, and an analyst who analyzes the cause of the jam according to a request from the service representative. The service representative may have a communication terminal 1701 for communication, and the analyst may have an information processing device 1702 for analyzing the cause of the jam and communicating with the service representative.

[0094] Traditionally, when a jam occurs, a service representative visits the user to take measures to resolve the cause. The service representative requests analysis from an analyst as needed. In this case, the service representative takes measures to resolve the cause according to the analysis results from the analyst.

[0095] In contrast, the image forming system 1 according to this embodiment uses the AI ​​function of the image forming apparatus 100 to estimate the cause of the jam without requiring an analysis request from an analyst.

[0096] First, a jam occurs in the paper transport unit 18 of the image forming apparatus 100. Then, the acquisition unit 1514 of the image forming apparatus 100 acquires a log showing the detection results from various sensors 1211 etc. at the time of the jam, as well as the current settings (Figure 7(1)). Then, the state detection unit 1515 extracts from the acquired information information such as the state of the paper transport unit 18 and the state of the paper 12 on the transport path, including the unit state of the paper transport unit 18, component information of the paper transport unit 18, transport sensor information immediately after the jam occurred, paper size in the print settings, paper size information identified from the detection results of the transport sensor 70, paper damage information, and environmental information.

[0097] The estimation unit 1516 then estimates the cause of the jam based on the extracted information (Figure 7(2)). Specifically, the estimation unit 1516 receives the estimated cause of the jam by inputting the information extracted by the state detection unit 1515 into the machine learning model 1204A. In the example shown in Figure 7, it is estimated that the jam occurred due to an operational problem.

[0098] Then, the determination unit 1517 refers to the jam response storage table 1503 and determines a method for resolving the cause of the jam. In Figure 7, the method for resolving the cause determined by the determination unit 1517 is assumed to be for cases where the cause of the jam can be resolved by user operation.

[0099] The display control unit 1518 then displays an improved operation plan on the display unit 1205, which associates the cause of the jam with a method for resolving it (Figure 7(3)). The user can resolve the cause of the jam by performing the operation according to the displayed improved plan. Therefore, it becomes unnecessary for the user to contact the service personnel. Furthermore, since it eliminates the need for work by the service personnel and analysis personnel, the workload of the service personnel and analysis personnel can be reduced.

[0100] Figure 8 shows an example of a jam cause estimation procedure in the image forming system 1 according to this embodiment. Figure 8 shows a service representative who visits the user and performs maintenance when a jam occurs, and an analyst who analyzes the cause of the jam according to a request from the service representative. The service representative has a communication terminal 1701 for communication, and the analyst has an information processing device 1702 for analyzing the cause of the jam and communicating with the service representative.

[0101] First, a jam occurs in the paper transport unit 18 of the image forming apparatus 100. Then, the acquisition unit 1514 of the image forming apparatus 100 acquires a log showing the detection results from various sensors 1211 etc. at the time of the jam, as well as the current settings (Figure 8 (1)). Then, the state detection unit 1515 extracts from the acquired information information such as the state of the paper transport unit 18 and the state of the paper 12 on the transport path, including the unit state of the paper transport unit 18, component information of the paper transport unit 18, transport sensor information immediately after the jam occurred, paper size in the print settings, paper size information identified from the detection results of the transport sensor 70, paper damage information, and environmental information.

[0102] The estimation unit 1516 then estimates the cause of the jam based on the extracted information (Figure 8 (2)). Specifically, the estimation unit 1516 receives the estimated cause of the jam by inputting the information extracted by the state detection unit 1515 into the machine learning model 1204A. In the example shown in Figure 8, it is estimated that the jam occurred due to a device malfunction. In this case, a service technician needs to visit the site, but analysis by an analyst is not required.

[0103] The determination unit 1517 then refers to the jam response retention table 1503 to determine a method for resolving the cause of the jam. In Figure 8, the method for resolving the cause determined by the determination unit 1517 is assumed to be when the cause of the jam can be resolved by replacing or adjusting the device. The resolving method includes information indicating the device to be replaced or adjusted, and the procedure for replacement or adjustment. The determination unit 1517 may also switch the contact person depending on the resolving method. In the example shown in Figure 8, the determination unit 1517 determines the contact person to be a service representative.

[0104] Therefore, the JOB control unit 1511 of the image forming apparatus 100 transmits to the service personnel's communication terminal 1701 information indicating that a jam has occurred in the image forming apparatus 100, along with information indicating the device to be replaced or adjusted, and the replacement or adjustment procedure (Figure 8 (3)).

[0105] Furthermore, the display control unit 1518 may, in accordance with the instructions of the JOB control unit 1511, display on the display unit 1205 that a jam has occurred and that a service representative has been contacted.

[0106] Based on the received information, the service technician's communication terminal 1701 displays that a jam has occurred in the image forming apparatus 100, along with information indicating the device to be replaced or adjusted, and the procedure for replacement or adjustment (Figure 8(4)).

[0107] Then, the service representative visits the user and takes measures such as replacing or adjusting the device that caused the jam in the image forming apparatus 100 (Figure 8 (5)). In the example shown in Figure 8, the service representative visits after recognizing the cause of the jam, so even if a device replacement is necessary, the replacement device can be prepared in advance. Therefore, the service representative does not need to make multiple visits, thus reducing their workload. Furthermore, since the work of the analysis staff is not required, the workload of the analysis staff is also reduced.

[0108] Figure 9 shows an example of a jam cause estimation procedure in the image forming system 1 according to this embodiment. Figure 9 shows a service representative who visits the user and performs maintenance when a jam occurs, and an analyst who analyzes the cause of the jam according to the request from the service representative. The service representative has a communication terminal 1701 for communication, and the analyst has an information processing device 1702 for analyzing the cause of the jam and communicating with the service representative.

[0109] First, a jam occurs in the paper transport unit 18 of the image forming apparatus 100. Then, the acquisition unit 1514 of the image forming apparatus 100 acquires a log showing the detection results from various sensors 1211 etc. at the time of the jam, as well as the current settings (Figure 9 (1)). Then, the state detection unit 1515 extracts from the acquired information information such as the state of the paper transport unit 18 and the state of the paper 12 on the transport path, including the unit state of the paper transport unit 18, component information of the paper transport unit 18, transport sensor information immediately after the jam occurred, paper size in the print settings, paper size information identified from the detection results of the transport sensor 70, paper damage information, and environmental information.

[0110] Then, the estimation unit 1516 estimates the cause of the jam based on the extracted information (Figure 9 (2)). Specifically, the estimation unit 1516 receives the estimated cause of the jam by inputting the information extracted by the state detection unit 1515 to the machine learning model 1204A. In the example shown in Figure 9, the machine learning model 1204A fails to estimate the cause of the jam. Therefore, the decision unit 1517 suppresses the decision on a method to resolve the cause of the jam. Furthermore, the decision unit 1517 determines that the contact person is a service representative.

[0111] Then, the JOB control unit 1511 of the image forming apparatus 100 transmits to the service representative's communication terminal 1701 that a jam has occurred in the image forming apparatus 100, and that it has failed to estimate the cause of the jam (Figure 9 (3)).

[0112] The service representative sends an analysis request to the analysis team via the communication terminal 1701 (Figure 9 (4)).

[0113] The information processing device 1702 for the analyst receives information and logs from the image forming apparatus 100 regarding the occurrence of a jam, according to the analyst's instructions (Figure 9(5)). The analyst then analyzes the received information and logs to identify the cause of the jam.

[0114] The analyst communicates (responds to) the service provider with the analysis results (Figure 9 (6)).

[0115] Then, the service representative visits the user and performs the necessary actions in the image forming apparatus 100 in accordance with the analysis results (Figure 9 (7)). This action resolves the cause of the jam.

[0116] The machine learning server 102 generates training data based on the jam occurrence information and logs received from the image forming apparatus 100, and the cause of the jam, which is the analysis result by the analyst. The machine learning model is then further trained using this training data. The image forming apparatus 100 stores the further trained machine learning model in the HDD 1204. Subsequently, by using the further trained machine learning model, the likelihood of being able to estimate the cause of the jam, which had previously failed to be estimated, is improved. Therefore, the image forming apparatus 100 according to this embodiment can achieve improved accuracy in estimating the cause of the jam.

[0117] Next, we will explain the specific estimation using the machine learning model 1204A. In this embodiment, the machine learning model 1204A estimates the cause of a jam by inputting the paper size in the print settings and the paper size based on the detection result of the transport sensor 70. Next, we will explain the estimation of the cause of a jam based on the paper size by the machine learning model 1204A.

[0118] Figure 10 is an explanatory diagram illustrating the cause of a jam based on the paper size in the print settings and the paper size calculated from the detection results of the transport sensor 70.

[0119] In the example shown in Figure 10(a), the paper length in the transport direction in the print settings for image forming is Lm1 [m], and the transport speed of the paper 12 on the transport path is V [sec / m]. Furthermore, the transport sensor 70A (70) is the upstream transport sensor 70, and the transport sensor 70B (70) is the downstream transport sensor 70A.

[0120] The paper transport unit 18 controls the rotation of the transport roller 58, so it is expected that the trailing end of the paper 12 will reach the transport sensor 70B after a time Lm1 / V [sec] has elapsed since the leading end of the paper 12 passed the transport sensor 70B.

[0121] Furthermore, the image forming apparatus 100 according to this embodiment has time margins Ta [sec] and Tb [sec] set for detecting jam.

[0122] In other words, after the leading edge of the paper 12 is detected by the transport sensor 70B, the state detection unit 1515 of the image forming apparatus 100 determines that a jam has occurred if the transport sensor 70B does not detect the trailing edge of the paper 12 within a time period of "Lm1 / V-Ta"sec" to "Lm1 / V+Tb"sec", taking into account a time margin, and the JOB control unit 1511 stops the printing process.

[0123] Furthermore, the state detection unit 1515 detects the paper length V × Tm [m] in the transport direction when it obtains the time Tm [sec] from when the transport sensor 70A, which is located upstream of the transport sensor 70B, detects the leading edge of the paper 12 until it detects the trailing edge of the paper 12.

[0124] In the example shown in Figure 10(b), the state detection unit 1515 considers that a jam has occurred because the time from when the leading edge of the paper 12 is detected by the transport sensor 70B until the trailing edge of the paper 12 is detected is shorter than the time "Lm1 / V-Ta[sec]".

[0125] In the example shown in Figure 10(b), the paper length Lm1 [m] is defined as the paper length in the transport direction according to the print settings. On the other hand, the state detection unit 1515 calculates the paper length Lm2 [m] based on the time from when the transport sensor 70A detects the leading edge of the paper 12 until when it detects the trailing edge of the paper 12.

[0126] In this case, the estimation unit 1516 inputs the paper length Lm1 in the transport direction according to the print settings and the paper size Lm2 calculated from the detection result of the transport sensor 70A to the machine learning model 1204A, and receives an estimation result that the cause of the jam is a mismatch between the paper size in the print settings and the actual paper size (the actual paper size is smaller).

[0127] In the example shown in Figure 10(c), the state detection unit 1515 considers a jam to have occurred because the time from when the leading edge of the paper 12 is detected by the transport sensor 70B until the trailing edge of the paper 12 is detected is longer than the time "Lm1 / V+Tb[sec]".

[0128] In the example shown in Figure 10(c), the paper length Lm1 [m] is defined as the paper length in the transport direction according to the print settings. On the other hand, the state detection unit 1515 calculates the paper length Lm3 [m] based on the time from when the transport sensor 70A detects the leading edge of the paper 12 until when it detects the trailing edge of the paper 12.

[0129] In this case, the estimation unit 1516 inputs the paper length Lm1 in the transport direction according to the print settings and the paper size Lm3 calculated from the detection result of the transport sensor 70A to the machine learning model 1204A, and receives an estimation result that the cause of the jam is a mismatch between the paper size in the print settings and the actual paper size (the actual paper size is longer).

[0130] In the example shown in Figure 10, the determination unit 1517 determines "correctly set the paper size" as the resolution method by referring to the jam-response retention table.

[0131] The display control unit 1518 then displays the cause of the jam and the method for resolving it on the display unit 1205.

[0132] Figure 11 shows an example of a screen displayed on the display unit 1205 by the display control unit 1518 according to this embodiment. In the example screen shown in Figure 11, "Incorrect paper size setting" is shown as the cause of the jam, and "Set the paper size correctly" is shown as the method to resolve the jam. By referring to this screen, the user can set the paper size correctly and resolve the cause of the jam.

[0133] In this embodiment, the jam-related storage table associates the cause of the jam with the method for resolving that cause. Therefore, after estimating the cause of the jam, the user can be informed of the method for resolving the jam. The user can take appropriate action according to the screen display shown in Figure 10.

[0134] Next, we will explain how the cause of jamming is estimated based on environmental information using the machine learning model 1204A. In image forming processes in low temperature or high humidity environments, condensation may occur in the image forming apparatus 100. If condensation occurs on a component on the transport path of the image forming apparatus 100, the paper 12 being transported may stick to that component, potentially reducing the transport speed of the paper 12.

[0135] Furthermore, the state detection unit 1515 of the image forming apparatus 100 determines that a jam has occurred if it fails to detect the arrival of the paper 12 as predicted based on the detection results of the transport sensor 70.

[0136] In this case, the estimation unit 1516 receives an estimation result that condensation is the cause of the jam by inputting environmental information, including temperature and humidity detected by the thermometer / hygrometer 1213, into the machine learning model 1204A.

[0137] In this case, the determination unit 1517 determines "raise the room temperature or lower the humidity" as the solution method by referring to the jam-response holding table. Therefore, the user can resolve the cause of the jam by operating the air conditioner in the environment (e.g., the room) in which the image forming apparatus 100 is located, thereby raising the temperature or lowering the humidity of the environment.

[0138] Next, we will explain how the machine learning model 1204A estimates the cause of a jam based on paper damage information.

[0139] Figure 12 is an explanatory diagram illustrating the cause of jamming based on paper damage information. Figure 12 shows the situation during image forming processing by the image forming apparatus 100, with the paper 12 passing through the fuser roller 28. When the fuser roller 28 shown in Figure 12 is hot, the paper 12 passing through the fuser roller 28 may curl. In the example shown in Figure 12, the curled paper 12 gets caught on a member 2101 installed downstream of the fuser roller 28 and is not transported along the discharge transport path 36.

[0140] In this case, the state detection unit 1515 of the image forming apparatus 100 cannot detect the arrival of the paper 12 as predicted based on the detection results of the transport sensor 70 downstream of the fuser roller 28, and therefore considers a jam to have occurred.

[0141] In this case, the estimation unit 1516 inputs paper damage information based on image information captured by the camera 1212 grounded on the discharge transport path 36 into the machine learning model 1204A, and receives an estimation result that the cause of the jam is a high fixing temperature.

[0142] The determination unit 1517 then refers to the jam-response holding table and determines that the solution is to "lower the fixing temperature to a level where the paper does not curl." The service technician can then resolve the cause of the jam by taking measures to lower the fixing temperature of the fixing roller 28.

[0143] Figure 13 illustrates the structure of a machine learning model generated by the machine learning server 102 according to this embodiment. As shown in Figure 13, the machine learning model is a machine learning model that causes the computer to function in order to output the cause of jamming based on information that causes jamming in the image forming apparatus 100. The example shown in Figure 13 illustrates a machine learning model using a neural network.

[0144] The machine learning model is configured as a neural network having an input layer 2301, an intermediate layer 2302, and an output layer 2303. In the example shown in Figure 13, the machine learning model inputs information such as transport sensor information, paper size information (including the paper size in the print settings and the paper size calculated from the detection results of the transport sensor 70), environmental information, and paper damage information as factors of jamming into the input layer 2301. The information input to the machine learning model may also include the unit status of the paper transport unit 18 and component information of the paper transport unit 18. Based on the input information, the machine learning model has been trained to determine the weighting coefficients of each component in the intermediate layer 2302 so that it outputs an estimated result of the cause of the jam from the output layer 2303.

[0145] The example shown in Figure 13 illustrates one possible cause of jam formation, but the factors are not limited to those shown in Figure 13. For example, detection results from various sensors 1211, etc., provided in the image forming apparatus 100 may be used as factors related to jam formation.

[0146] Figure 14 is an explanatory diagram illustrating the concept of machine learning performed by the machine learning unit 1532 of the machine learning server 102 according to this embodiment. As shown in Figure 14, the training data generation unit 1531 prepares (generates) a large amount of training data to be used for machine learning in advance (Figure 14 (1)).

[0147] The input data X included in the training data consists of the unit status of the paper transport unit 18, component information of the paper transport unit 18, transport sensor information, paper size information (including the paper size in the print settings and the paper size calculated from the detection results of the transport sensor 70), paper damage information, and environmental information, which are input to the input layer 2301 as causes of jamming. The input data X consists of data for which the correct answer value is known. The expected value (T) included in the training data is the cause of the jamming. The expected value T is the correct answer value corresponding to the input data.

[0148] Specific machine learning techniques include neural networks, nearest neighbor methods, Naive Bayes, decision trees, and support vector machines. Deep learning, which uses neural networks to generate features and connection weights for learning, is another example. Appropriate algorithms from the above can be used in this embodiment.

[0149] Then, the machine learning unit 1532 adjusts the weighting coefficients of the machine learning model so that when the input data X included in the training data is input to the machine learning model, the output data Y is as close as possible to the expected value T corresponding to the input data X.

[0150] The machine learning unit 1532 functions as both an error detection unit and an update unit. Specifically, the machine learning unit 1532 receives input data X into the input layer (Figure 14 (2)), and calculations are performed by the machine learning model (Figure 14 (3)). The machine learning unit 1532 then receives output data Y from the output layer as a result of the machine learning model (Figure 14 (4)). As an error detection unit, the machine learning unit 1532 uses a loss function to calculate a loss L that represents the magnitude of the discrepancy between the output data Y and the expected value T of the training data (Figure 14 (5)).

[0151] The machine learning unit 1532, acting as an update unit, updates the connection weight coefficients between nodes of the neural network, etc., based on the loss L, in order to reduce the loss L (to bring the loss L closer to 0) (Figure 14 (6)).

[0152] For machine learning unit 1532, the algorithm used to train the neural network is, for example, backpropagation, but other algorithms may also be used.

[0153] Figure 15 is a flowchart illustrating the procedure for processing jams in the image forming apparatus 100 according to the first embodiment.

[0154] First, the state detection unit 1515 of the image forming apparatus 100 detects a jam based on the detection results of various sensors 1211 etc. acquired by the acquisition unit 1514 and the current settings (S2501).

[0155] Next, the state detection unit 1515 collects data indicating the state of each component included in the image forming apparatus 100 at the time of jamming, based on the detection results of various sensors 1211 etc. acquired by the acquisition unit 1514 and the current settings, and stores this data in the data storage unit 1502 (S2502).

[0156] The estimation unit 1516 estimates the cause of the jam using the machine learning model 1204A (S2503). Specifically, the estimation unit 1516 inputs the data stored in the data storage unit 1502 into the machine learning model 1204A and receives the estimated cause of the jam from the machine learning model 1204A.

[0157] The determination unit 1517 refers to the jam correspondence storage table 1503 and determines a method for resolving the cause of the jam based on the estimation result of the estimation unit 1516 (S2504).

[0158] The display control unit 1518 displays information on the display unit 1205 indicating the cause of the jam estimated by the estimation unit 1516 and the method for resolving the cause determined by the determination unit 1517 (S2505).

[0159] In this embodiment, the user can recognize a method to resolve the cause of the jam by referring to the information displayed on the display unit 1205. By taking action to resolve the cause of the jam, the downtime of the image forming apparatus 100 can be reduced. Furthermore, since it is not necessary to call a service technician, the burden on the service technician can be reduced.

[0160] The image forming apparatus 100 according to this embodiment estimates the cause of a jam using AI (Artificial Intelligence) functionality. When the image forming apparatus 100 estimates the cause of a jam, it can improve the accuracy of its estimation by using various information such as the paper size in the print settings at the time of the jam, the paper size detected by the apparatus, environmental information, and paper damage information. Furthermore, it becomes possible to resolve the cause of the jam without waiting for a service representative to visit the user's site or for an analysis by an analysis specialist. Therefore, the image forming apparatus 100 according to this embodiment can reduce downtime.

[0161] In this embodiment, an example has been described in which the image forming apparatus 100 includes an acquisition unit 1514, a state detection unit 1515, an estimation unit 1516, a storage unit (HDD 1503) having a jam-response holding table 1503, a display control unit 1518, and a display unit 1205, all of which are provided in the image forming apparatus 100 having a paper transport unit 18. However, in this embodiment, the image forming apparatus 100 is not limited to having all of the above configurations. For example, the display control unit 1518 and the display unit 1205 may be implemented on a mobile terminal owned by the user, or the storage unit (HDD 1503) having the jam-response holding table 1503 and the estimation unit 1516 may be implemented on an information processing device that can communicate with the image forming apparatus 100. Furthermore, the acquisition unit 1514 and the state detection unit 1515 may be implemented on a diagnostic device that can communicate with the image forming apparatus 100. In this way, an estimation system may be implemented by a combination of multiple devices.

[0162] (Second embodiment) In the embodiments described above, an example was described in which the data server 105 and the machine learning server 102 are provided in the image forming apparatus 100. However, the embodiments described above are not limited to a configuration in which the data collection, learning phase, and inference phase are each performed by separate devices. For example, the image forming apparatus may have a data collection function and a machine learning model generation function.

[0163] Figure 16 shows an example of the software configuration of the image forming apparatus 100A according to this embodiment. Figure 16 shows the software configuration realized by utilizing the hardware resources and programs shown in Figure 3, which are included in the image forming apparatus 100A in this embodiment.

[0164] Figure 16 shows that in this embodiment, the image forming apparatus 100A further includes, in addition to the configuration of the image forming apparatus 100 shown in Figure 5, a data acquisition unit 1521 for collecting data to be used as training data, a training data generation unit 1531 for generating training data, and a machine learning unit 1532 for generating machine learning models.

[0165] The image forming apparatus 100A includes a data acquisition unit 1521, a training data generation unit 1531, and a machine learning model. It collects data when a jam occurs, generates training data, and generates and updates the machine learning model based on the training data. The specific processing of each component is the same as in the embodiment described above, and the same reference numerals as in the embodiment described above are assigned, so their explanation is omitted.

[0166] (Third embodiment) The embodiments described above describe a case in which the image forming apparatus 100 holds a machine learning model 1204A and the image forming apparatus 100 performs AI functions. However, the embodiments described above are not limited to cases in which the image forming apparatus 100 performs AI functions.

[0167] For example, the cloud server that can communicate with the image forming apparatus 100 may also be an estimation device that performs AI functions. For example, the CPU 1201 of the image forming apparatus 100 transmits to the cloud server information that can detect jams caused by paper 12 in the transport path, the unit status of the paper transport unit 18, component information of the paper transport unit 18, transport sensor information immediately after the jam occurred, paper size in the print settings, paper size information identified from the detection results of the transport sensor 70, paper damage information, and environmental information.

[0168] Then, when the cloud server receives information from the image forming apparatus 100, it assumes that a jam has occurred in the image forming apparatus 100. The cloud server then estimates the cause of the jam based on the received information. Furthermore, the cloud server determines a resolution method corresponding to the cause of the jam. The method for determining the resolution method is the same as in the embodiment described above.

[0169] The cloud server then transmits the estimated cause of the jam and the method for resolving the jam to the image forming apparatus 100. The image forming apparatus 100 then displays the estimated cause of the jam, the method for resolving the jam, and the display unit 1205.

[0170] This embodiment provides the same effects as the first embodiment described above, and since the image forming apparatus 100 does not need to perform the inference phase, the computational load can be reduced.

[0171] In the embodiments described above, when a jam occurs, the image forming apparatus or cloud server estimates the cause of the jam, thereby reducing the workload required to identify the cause of the jam. Furthermore, since users can resolve the cause of the jam based on the identified cause, the image forming apparatus can be used immediately, thus reducing the downtime of the image forming apparatus.

[0172] Although several embodiments for carrying out the present invention have been described above, the present invention is not limited in any way to these embodiments, and various modifications and substitutions can be made without departing from the spirit of the invention. Furthermore, the elements shown in the embodiments described above can be combined in any way as appropriate, as long as no technical inconsistencies arise. [Explanation of Symbols]

[0173] 100, 100A Image forming apparatus 70 Transport Sensors 1211 Various Sensors 1212 Camera 1213 Temperature / Hygrometer 1204 HDD 1204A Machine Learning Model 1502 Data Storage Unit 1503 Jam-compatible holding table 1511 JOB control section 1512 Image reading unit 1513 Counter section 1514 Acquisition Department 1515 State detection unit 1516 Estimation Department 1517 Decision Section 1518 Display Control Unit 105 Data Server 1521 Data Collection Unit 1522 Data Provision Department 1523 Data Storage Unit 102 Machine Learning Server 1531 Training Data Generation Unit 1532 Machine Learning Department 1533 Data Storage Unit 103 General-purpose computers [Prior art documents] [Patent Documents]

[0174] [Patent Document 1] Japanese Patent Publication No. 2019-160314

Claims

1. When paper is transported along the transport path by the operation of the paper transport unit, a jam detection unit capable of detecting a jam caused by the paper on the transport path is provided. When a jam is detected by the jam detection unit, an acquisition unit acquires information indicating the state of the paper transport unit and the state of the paper on the transport path. An estimation unit receives an estimation result of the cause of the jam from a learning model, which has been trained using machine learning to estimate the cause of the jam caused by the aforementioned paper, by inputting information indicating one or more of the state of the paper transport unit and the state of the paper on the transport path. An estimation device equipped with the following features.

2. The acquisition unit acquires the paper size set for printing on the paper, and the size of the paper identified by the detection unit based on the detection result. The estimation unit receives an estimated result of the cause of the jam from the learning model by inputting information indicating the paper size set for printing the paper and the paper size identified by the detection unit based on the detection result to the learning model. The estimation device according to claim 1.

3. The system further includes an environment detection unit capable of detecting information indicating the environment in which the paper transport unit is installed. The estimation unit further receives information indicating the environment in which the paper transport unit is installed, as detected by the environment detection unit, and receives the estimated cause of the jam from the learning model. The estimation device according to claim 1.

4. When the jam detection unit detects the jam, the system further includes a paper condition detection unit capable of detecting the degree of damage to the paper moving along the transport path. The estimation unit further receives information indicating the degree of damage to the paper obtained by the paper condition detection unit, and receives an estimated result of the cause of the jam from the learning model. The estimation device according to claim 1.

5. A storage unit that stores a table associating the cause of the jam with a method for resolving that cause, A display control unit that displays information on a display device indicating the cause of the jam estimated by the estimation unit and the method for resolving the cause, which is associated with the table, The estimation device according to any one of claims 1 to 4, further comprising:

6. A paper transport unit that operates to transport paper along a transport path, A jam detection unit capable of detecting jams caused by the paper on the transport path, When a jam is detected by the jam detection unit, an acquisition unit acquires information indicating the state of the paper transport unit and the state of the paper on the transport path. An estimation unit receives an estimation result of the cause of the jam from a learning model, which has been trained using machine learning to estimate the cause of the jam caused by the aforementioned paper, by inputting information indicating one or more of the state of the paper transport unit and the state of the paper on the transport path. A storage unit that stores a table associating the cause of the jam with a method for resolving that cause, The system includes a display control unit that displays information on a display device indicating a method for resolving the jam, which is associated with the cause of the jam estimated by the estimation unit and the table. At least one of the acquisition unit, estimation unit, storage unit, and display control unit is provided in the device having the paper transport unit. Estimation system.

7. A learning unit generates a learning model by performing machine learning using data that associates information indicating the state of the paper transport unit and the state of the paper on the transport path when a paper jam occurs in the transport path operated by the paper transport unit, with the cause of the jam, as training data. A pre-trained model generator equipped with the following features.

Images