Baumkuchen baking system, baumkuchen baking support system, program, and baumkuchen manufacturing method

Abstract

A baumkuchen baking system 10 comprises: a baumkuchen baking machine 1 including an oven 2, a batter container 4, a rotating roller 3, and a camera 7; and a control unit 8. A server 20 can access a storage unit 30 which stores a trained model obtained by learning baking control or baked state determination or based on an image of the outer peripheral surface of baumkuchen batter. The control unit 8 includes: a user interface unit 83; and an automatic control unit that, on the basis of an image, captured by the camera 7, of the outer peripheral surface of the baumkuchen batter that is being baked, assesses a baked state or baking control 82 using a trained model provided from the server 20, and that uses the assessment results to automatically control the baking of each batter layer of the baumkuchen.

Description

[Technical Field]

[0001] The present invention relates to a computer system that utilizes a Baumkuchen baking machine. [Background technology]

[0002] The Baumkuchen baking machine rotates a rotating shaft coated with Baumkuchen batter in an oven. This causes the outer periphery of the batter to be baked almost uniformly all around. Once the outer periphery is baked, the Baumkuchen baking machine applies more batter on top of it and rotates it in the oven. In this way, by repeating the application of batter and the baking of the outer periphery of the batter by rotation in the oven, a Baumkuchen with a layered structure reminiscent of tree rings is baked.

[0003] The degree of doneness of Baumkuchen is important as it has a major impact on its quality. For this reason, skilled craftsmen operate the Baumkuchen baking machine to bake the Baumkuchen at the appropriate degree of doneness.

[0004] For example, Japanese Patent Application Laid-Open Publication No. 2021-010333 discloses a Baumkuchen baking machine. This Baumkuchen baking machine includes a rotating drum installed in a baking furnace and a drive mechanism that controls the revolution of six supporting rods hung on the rotating drum, causing them to revolve in sequence from a first intermittent revolution stop position to a sixth intermittent revolution stop position, and the rotation of the supporting rods. Furthermore, the Baumkuchen baking machine is provided with a first partition shutter and a second partition shutter that move forward and backward in synchronization with the intermittent revolution of the drive mechanism to block the revolution movement trajectory. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent Publication No. 2021-010333 Summary of the Invention [Problem to be solved by the invention]

[0006] The conventional Baumkuchen baking machine described above can bake multiple Baumkuchen simultaneously. However, it is difficult for small stores, such as Western-style confectionery shops that are rooted in their local communities, to install such a large Baumkuchen baking machine. Furthermore, even if a Baumkuchen baking machine is installed, it is difficult to serve Baumkuchen without a craftsman who can operate the Baumkuchen baking while monitoring the degree of doneness. However, the number of Baumkuchen craftsmen is limited. For example, in Germany, the birthplace of Baumkuchen, there are few small stores that manufacture and sell Baumkuchen. For the average German citizen, Baumkuchen is not a familiar confection, but a special luxury confection.

[0007] Therefore, the present application discloses a Baumkuchen baking system, a Baumkuchen baking support system, a program, and a Baumkuchen manufacturing method that can produce high-quality Baumkuchen without a Baumkuchen baking machine with a complex mechanism or operation by a skilled craftsman. [Means for solving the problem]

[0008] A Baumkuchen baking system according to an embodiment of the present invention includes a communication unit that communicates data with a server, a Baumkuchen baking machine having an oven, a dough container, a rotating roll that is movable between a baking position in the oven and the dough container, and a camera that photographs a portion of the outer periphery of the Baumkuchen dough layered on the rotating roll, and a control unit that controls the Baumkuchen baking machine. The server can access a memory unit that stores a trained model obtained by learning to determine the degree of doneness or to control baking based on images of the outer periphery of the Baumkuchen dough layered on the rotating roll and being baked. The control unit includes an automatic control unit that uses the trained model provided by the server to determine the degree of doneness or to control baking based on images of the outer periphery of the Baumkuchen dough being baked at the baking position in the oven photographed by the camera, and automatically controls the baking of each layer of dough of the Baumkuchen using the determination result. [Effects of the Invention]

[0009] According to the present disclosure, high-quality Baumkuchen can be produced without a Baumkuchen baking machine with a complicated mechanism or operation by a skilled craftsman. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a diagram showing an example of the overall configuration of a system according to this embodiment. [Figure 2] FIG. 2 is a diagram illustrating an example of the configuration of the control unit of the Baumkuchen baking system. [Figure 3] FIG. 3 is a flowchart showing an example of a process for baking Baumkuchen under automatic control. [Figure 4] FIG. 4 is a flowchart showing an example of a process for baking Baumkuchen by remote control. [Figure 5] FIG. 5 is a flowchart showing an example of the control process of S205 in FIG. [Figure 6] FIG. 6 is a diagram showing an example of a screen displayed in the process of S205 in FIG. [Figure 7] FIG. 7 illustrates an example of the operation of the server illustrated in FIG. [Figure 8] FIG. 8 is a front view of the Baumkuchen baking machine according to this embodiment. [Figure 9] FIG. 9 is a side view of the Baumkuchen baking machine shown in FIG. [Figure 10] FIG. 10 is a diagram showing the rotating roll 3 in the dough coating position. [Figure 11] FIG. 11 is a flowchart showing an example of the control process of the Baumkuchen baking machine 1 by the control unit 8. [Figure 12] FIG. 12 is a diagram showing an example of an image captured by the camera 7. As shown in FIG. [Figure 13] FIG. 13 is a diagram showing an example of the configuration of a neural network used in the determination process. [Figure 14]FIG. 14 is a diagram showing an example of a group of images acquired by the control unit 8 from the start of baking to the end of baking. [Figure 15] FIG. 15 is a flowchart showing an example of processing for collecting training data for learning processing based on the baking operation of the Baumkuchen baking machine 1. [Figure 16] FIG. 16 is a diagram showing a modified example of the entire system of this embodiment. [Figure 17] FIG. 17 is a diagram showing an example of the content indicated by dough recipe data. DETAILED DESCRIPTION OF THE INVENTION

[0011] The inventor sensed and analyzed the actions of a craftsman operating a Baumkuchen baking machine to bake Baumkuchen. He attempted to digitize the craftsman's skills and reproduce the craftsman's skills through automatic control using the data. After trial and error, the inventor discovered that using machine learning to digitize the part where a craftsman judges the degree of doneness by observing the browning of the outer surface of the Baumkuchen dough or controls the baking is particularly effective in reproducing the craftsman's skills using data. The following embodiments were conceived based on this knowledge.

[0012] (Configuration 1) A Baumkuchen baking system according to an embodiment of the present invention includes a communication unit that communicates data with a server, a Baumkuchen baking machine having an oven, a dough container, a rotating roll that is movable between a baking position in the oven and the dough container, and a camera that photographs a portion of the outer periphery of the Baumkuchen dough layered on the rotating roll, and a control unit that controls the Baumkuchen baking machine. The server can access a memory unit that stores a trained model obtained by learning to determine the degree of doneness or to control baking based on images of the outer periphery of the Baumkuchen dough layered on the rotating roll and being baked. The control unit includes an automatic control unit that uses the trained model provided by the server to determine the degree of doneness or to control baking based on images of the outer periphery of the Baumkuchen dough being baked at the baking position in the oven photographed by the camera, and automatically controls the baking of each layer of dough of the Baumkuchen using the determination result.

[0013] In the above-described configuration 1, the Baumkuchen baking machine of the Baumkuchen baking system includes a dough container, a rotating roll, a camera, and a control unit. The server provides the Baumkuchen baking system with a trained model. The trained model is data obtained by learning the relationship between images of the outer surface of the Baumkuchen dough layered on the rotating roll and baked, and the determination of doneness or baking control. The Baumkuchen baking system can determine the doneness or baking control using the trained model based on the camera images. This allows the server and the Baumkuchen baking system to share the method of determining the doneness or baking control based on the color of the outer surface of the dough when baking each layer of Baumkuchen dough. The inventors discovered that sharing this part facilitates the production of high-quality Baumkuchen in the Baumkuchen baking system. The Baumkuchen baking system can produce high-quality Baumkuchen without, for example, introducing a Baumkuchen baking machine with a complex mechanism or skilled craftsmen. As a result, the business of providing Baumkuchen becomes easier.

[0014] The automatic control unit may automatically control the timing to finish baking one layer by moving the rotating roll on which the Baumkuchen dough is stacked from the baking position in the oven to another position based on the determination result, thereby automatically controlling the baking time of each layer using the trained model.

[0015] The memory unit accessible by the server may store multiple trained models. Baking conditions may be associated with each of the multiple trained models. The control unit may further include a user interface unit that accepts input of baking conditions for the Baumkuchen to be manufactured from the operator. The automatic control unit may perform the automatic control using a trained model provided by the server that corresponds to the baking conditions input by the operator. This enables appropriate automatic control according to the baking conditions. The baking conditions may include, for example, at least one of conditions related to the craftsman who contributed to the generation of the training data used to train the trained model, conditions related to the dough (such as the dough ingredients or the dough's physical properties), the size of the rotating roll (core rod), or the number of dough layers.

[0016] The learning (machine learning) for generating the trained model may be, for example, deep learning using a neural network. The trained model may be, for example, a dataset for inputting an image and outputting a determination result of the degree of doneness (e.g., a value indicating the degree of doneness, or whether the degree of doneness is appropriate, etc.) or baking control (e.g., a value for controlling the baking time, etc.). The dataset includes, for example, parameters indicating the weights between layers of a neural network, parameters whose values ​​have been adjusted by learning. Note that learning is not limited to using a neural network. For example, the trained model may be generated by machine learning using regression analysis or a decision tree. Examples of such machine learning techniques include linear regression, support vector machines, support vector regression, elastic nets, logistic regression, and random forests. The trained model may be generated by training in a baumkuchen baking machine that controls baking using the trained model, or by training in a baumkuchen baking machine different from the baumkuchen baking machine.

[0017] (Configuration 2) In the above configuration 1, the memory unit accessible by the server may store a plurality of the trained models. Each of the plurality of trained models may be stored in association with artisan data indicating the artisan who contributed to the generation of the training data used in training each trained model. The control unit may further include a user interface unit that accepts the designation of the artisan from the operator. The automatic control unit can determine the doneness or baking control based on the image captured by the camera using the trained model provided by the server and associated with the artisan data indicating the designated artisan.

[0018] The inventors discovered that the judgment of doneness or baking control based on the browning of the outer surface of the dough varies slightly from baker to baker, and this affects the quality of each Baumkuchen. According to the above configuration 2, the server can provide a trained model for each baker. The Baumkuchen baking system can reproduce baking similar to that performed by the designated baker through automatic control using a trained model corresponding to the designated baker.

[0019] A craftsman is a person or group of people who has the skills to operate a Baumkuchen baking machine and bake Baumkuchen. The higher the level of skill possessed by a craftsman, the higher the level is, but this is not particularly limited. The craftsman data indicating the craftsman may be data that identifies the individual craftsman, or may be data that identifies a group of craftsmen (an organization, group, team, etc.). A craftsman may be, for example, an individual called a meister or pastry chef, or may be a Baumkuchen manufacturing shop, a Baumkuchen manufacturing company, or other organization that manufactures Baumkuchen.

[0020] (Configuration 3) In the above configuration 1 or 2, the control unit may further include a remote control unit that provides a remote terminal capable of data communication via the communication unit with an image of the outer surface of the Baumkuchen batter taken by the camera in real time and controls the baking of the batter for each layer of the Baumkuchen in accordance with operation instructions received from the remote terminal. Note that the control unit may be configured to omit the automatic control unit and instead include a remote control unit.

[0021] This allows an operator to control the baking from a remote location away from the Baumkuchen baking machine while checking the Baumkuchen batter's doneness on an image on the remote terminal. For example, an instruction to move the rotating roll of the Baumkuchen baking machine from the baking position to the batter application position during baking of each layer of batter can be received on the remote terminal as an operation instruction.

[0022] The remote control unit may set a range of baking time that can be controlled by operation instructions from the remote terminal for each layer of the Baumkuchen dough, thereby allowing the degree of doneness of each layer of the Baumkuchen to be adjusted according to the preference, skill, etc. of the operator who remotely controls the device while maintaining the degree of doneness of each layer of the Baumkuchen within a certain range.

[0023] (Configuration 4) In the above configuration 3, the remote control unit may use the trained model provided by the server to determine the degree of doneness or baking control of the Baumkuchen batter based on an image of the outer periphery of the Baumkuchen batter being baked in the Baumkuchen baking machine taken by the camera, and provide the determination result together with the image to the remote terminal in real time. This allows an operator to operate the Baumkuchen baking machine from a remote location while checking in real time information on the degree of doneness or baking control determined using the trained model.

[0024] In the above-mentioned configuration 3, the remote control unit may use the trained model provided by the server to determine the degree of doneness or baking control of the Baumkuchen batter based on an image of the outer periphery of the Baumkuchen batter being baked in the Baumkuchen baking machine taken by the camera, and may use the determination result to automatically control the baking of each layer of the Baumkuchen batter and provide the image to the remote terminal in real time. In addition to the automatic control, the remote control unit may further control the baking of each layer of the Baumkuchen batter based on operation instructions received from the remote terminal. With this configuration, it is possible to automatically control the baking of each layer of batter using the trained model while accepting operation instructions from the remote terminal. For example, it is possible to provide the remote operator with freedom to adjust the baking while ensuring a certain level of quality through automatic control.

[0025] The remote control unit may receive a designation of a craftsman from the remote terminal and determine the degree of doneness or baking control using a trained model corresponding to the designated craftsman, thereby providing information indicating the designated craftsman's determination of the degree of doneness or baking control to the remote terminal in real time.

[0026] The remote control unit may provide the remote terminal with an image of the outer surface of the Baumkuchen dough taken by the camera in real time, receive from the remote terminal a determination result on the degree of doneness or baking control determined at the remote terminal based on the image using a trained model provided by the server, and control the baking of the dough of each layer of the Baumkuchen using the received determination result.

[0027] (Configuration 5) In any of the above configurations 1 to 5, the control unit may further include a learning unit that generates, as training data for learning, data indicating the baking control or the determination result of the degree of doneness of the batter for each layer estimated from the operation of the Baumkuchen baking machine during baking under manual control by an operator, and an image of the outer surface of the Baumkuchen batter during the manually controlled baking, taken by the camera. The control unit may provide the training data generated by the learning unit or a trained model generated by learning using the training data to the server via the communication unit.

[0028] This allows the Baumkuchen baking system to learn how an operator makes decisions and controls the baking machine to bake Baumkuchen, and generate a trained model. The generated trained model is provided to the server. This makes it possible for the operator's skills learned in the Baumkuchen baking system to be provided from the server. For example, in an environment where multiple Baumkuchen baking systems can communicate with the server, the artisan's skills learned in one Baumkuchen baking system can be realized in another Baumkuchen baking system.

[0029] The control unit may associate the trained model generated by the learning unit with craftsman data indicating, as a craftsman, the operator who performed the operation on which the judgment result of the teacher data used in learning the trained model is based, and provide the trained model to the server via the communication unit. This allows the server to record the trained model and the craftsman data in association with each other in the storage unit.

[0030] The learning unit may generate training data for baking each of a plurality of Baumkuchen. In this case, the control unit may provide the server with training data for a Baumkuchen specified by the operator from among the training data for the plurality of Baumkuchen, or a trained model generated by learning using the specified training data. This allows the operator to specify, for example, training data for a Baumkuchen that has been baked to a good degree of doneness, or a trained model based on this training data, as the data to be provided to the server. Furthermore, the control unit may provide the server with training data for a predetermined number or more of a plurality of Baumkuchen, or a trained model generated by learning using the training data for the plurality of Baumkuchen.

[0031] The control unit may associate the trained model generated by the learning unit with dough recipe data indicating the composition of ingredients and the manufacturing procedure of the Baumkuchen dough used in training the trained model, and provide the model to the server via the communication unit. This allows the server to record the trained model and the dough recipe data in association with each other in a storage unit. The control unit may also associate the trained model with the craftsman data and the dough recipe data and provide the trained model to the server.

[0032] (Configuration 6) In any of the above configurations 1 to 6, the control unit may acquire from the server dough recipe data indicating the dough ingredient combination and dough preparation procedure associated with the trained model provided by the server, and output the dough recipe data to an operator of the Baumkuchen baking machine. This enables the operator of the Baumkuchen baking machine to prepare dough suitable for baking using the trained model. As a result, it becomes possible to provide higher quality Baumkuchen. Note that the storage unit accessible by the server may store dough recipe data in association with each trained model. In this case, dough recipe data indicating the dough ingredient combination and dough preparation procedure at the time of generating the training data used to train each trained model is stored in association with each trained model.

[0033] In the above configuration 2, the control unit may acquire dough recipe data associated with the trained model associated with craftsman data indicating the craftsman designated by the operator, thereby enabling the baking of a Baumkuchen of quality closer to that of the Baumkuchen made by the designated craftsman.

[0034] The dough recipe data may include, as data indicating the composition of the dough ingredients, data indicating the ingredients (contents) of the dough or the amount of each ingredient contained. Furthermore, the dough recipe data may include, as data indicating the dough preparation procedure, data indicating the order in which the ingredients are added and the conditions for mixing the added ingredients (mixing conditions). Furthermore, the data indicating the dough preparation procedure may further include data indicating the temperature of the ingredients to be added or data indicating the physical properties of the dough, such as the specific gravity.

[0035] In any of the above configurations 1 to 6, the Baumkuchen baking machine may further include a mixer that mixes the ingredients of the dough. The automatic control unit may acquire from the server dough recipe data that indicates the combination of ingredients and a dough preparation procedure, and that is associated with the trained model provided by the server, and control the mixer based on the dough recipe data. This makes it possible to automate at least a portion of the preparation procedure indicated by the dough recipe data. For example, the control unit may control mixing by the mixer in accordance with the mixing conditions for each ingredient indicated by the dough recipe data.

[0036] In any of the above configurations 1 to 6, the Baumkuchen baking machine may further include a light that irradiates a region included in the camera's imaging range. That is, the Baumkuchen baking machine may be equipped with a dedicated light. This stabilizes the imaging environment of the batter during baking by the camera. As a result, the accuracy of the determination by the automatic control unit is improved. For example, the light is supported in the Baumkuchen baking machine at a position that allows it to irradiate the camera's imaging range. The light source of the light is not limited to, but may be configured to have a brightness of 3000 lm or more and be positioned within 1.5 m of the Baumkuchen batter during baking. Note that the trained model provided by the server may be a trained model obtained by training using images of the outer surface of the Baumkuchen batter photographed under the same illumination conditions as those of the light from the light, and the breakage or baking control of the baked batter as training data.

[0037] (Configuration 7) A Baumkuchen baking support system according to an embodiment of the present invention is capable of accessing a storage unit that records a trained model obtained by learning to determine doneness or control baking based on an image of the outer periphery of the Baumkuchen dough that is stacked on a rotating roll and baked. The Baumkuchen baking support system includes: a model providing unit that provides the trained model to a Baumkuchen baking system that has a Baumkuchen baking machine, a camera, and a control unit; and a baking performance receiving unit that receives from the Baumkuchen baking system performance data indicating the performance of baking Baumkuchen by automatically controlling the Baumkuchen baking machine based on an image captured by the camera using the trained model provided by the model providing unit.

[0038] According to the above-mentioned configuration 7, the Baumkuchen baking system uses the provided trained model to determine the degree of doneness or baking control of each layer of Baumkuchen dough based on images captured by a camera, enabling automatic control of the baking of each layer. The Baumkuchen baking system can produce high-quality Baumkuchen. In addition, performance data showing the results of baking using the trained model is provided to the Baumkuchen baking support system. This allows the Baumkuchen baking support system to understand the usage status of the trained model. Therefore, both the provider and user of the trained model can easily enjoy appropriate benefits. As a result, the business of providing Baumkuchen is made easier.

[0039] (Configuration 8) The storage unit may store a plurality of the trained models. In the storage unit, each of the plurality of trained models may be stored in association with artisan data indicating the artisan who contributed to generating the training data used in training the trained model. The model providing unit may provide the Baumkuchen baking system with a trained model associated with artisan data indicating the artisan input by an operator in the Baumkuchen baking system. This allows the Baumkuchen baking support system to provide the Baumkuchen baking system with trained data corresponding to the specified artisan.

[0040] (Configuration 9) In the above configuration 7 or 8, the Baumkuchen baking support system may further include a model registration unit that receives a trained model from the Baumkuchen baking system and records it in the storage unit. The trained model is a trained model generated by learning using, as training data, the results of the determination of the degree of doneness of each layer of dough or the baking control estimated from the operation of the Baumkuchen baking machine during baking by manual control by an operator, and an image of the outer surface of the Baumkuchen dough during the manual baking taken by the camera. This makes it possible to use a trained model trained in one Baumkuchen baking system in another Baumkuchen baking system.

[0041] In the above configuration 9, the model registration unit may receive, in addition to the trained model, craftsman data indicating, as a craftsman, the operator who performed the operation based on the judgment result of the training data used in training the trained model from the Baumkuchen baking system, and may associate the craftsman data with the trained model and record it in the memory unit.

[0042] (Configuration 10) The Baumkuchen baking support system of configuration 8 may further include an accounting unit that uses the performance data received by the baking performance receiving unit to calculate the usage fee for the trained model used in the Baumkuchen baking system and the compensation for the craftsman indicated by the craftsman data corresponding to the trained model. This enables the provision of trained models and a settlement that promotes the contributions of craftsmen to the trained models.

[0043] A program according to an embodiment of the present invention is a program capable of data communication with a server and causing a computer that controls a Baumkuchen baking machine to execute processing. The server can access a storage unit that stores a trained model obtained by learning to determine the degree of doneness or control baking based on an image of the outer periphery of the Baumkuchen dough layered on a rotating roll and baked. The Baumkuchen baking machine includes an oven, a dough container, a rotating roll that can move between a baking position in the oven and the dough container, and a camera that captures an image of a portion of the outer periphery of the Baumkuchen dough layered on the rotating roll. The program causes a computer to execute the following processes: receiving an instruction for automatic control using the trained model from an operator; determining the degree of doneness or control baking using the trained model provided by the server based on an image of the outer periphery of the Baumkuchen dough being baked at the baking position in the oven captured by the camera; and automatically controlling the baking of each layer of the Baumkuchen dough using the determination result.

[0044] A program according to an embodiment of the present invention is a program for causing a computer capable of communicating with a Baumkuchen baking system having a Baumkuchen baking machine, a camera, and a control unit to execute processing. The program causes the computer to execute the following processes: accessing a storage unit that records a trained model obtained by learning how to determine the degree of doneness or control baking based on an image of the outer periphery of the Baumkuchen dough that is layered on a rotating roll and baked; providing the trained model to the Baumkuchen baking system; and receiving from the Baumkuchen baking system performance data indicating the performance of baking Baumkuchen by automatically controlling the Baumkuchen baking machine based on images captured by the camera using the provided trained model.

[0045] In one embodiment of the present invention, a manufacturing method is a method for manufacturing Baumkuchen by controlling a Baumkuchen baking machine using a computer capable of communicating with a server. The server can access a memory unit that stores multiple trained models obtained by learning to determine the degree of doneness or control baking based on images of the outer periphery of the Baumkuchen dough layered on a rotating roll and baked. The Baumkuchen baking machine includes an oven, a dough container, a rotating roll that can move between a baking position in the oven and the dough container, and a camera that captures images of a portion of the outer periphery of the Baumkuchen dough layered on the rotating roll. The manufacturing method includes the steps of: the computer receiving an instruction for automatic control using the trained model from an operator; and the computer determining the degree of doneness or control baking using the trained model provided by the server based on images of the outer periphery of the Baumkuchen dough being baked at the baking position in the oven captured by the camera, and automatically controlling the baking of each layer of dough of the Baumkuchen using the determination result.

[0046] [Embodiment] Hereinafter, embodiments will be described with reference to the drawings. The same or corresponding components in the drawings are designated by the same reference numerals, and the same description will not be repeated. For ease of understanding, the drawings referred to below show simplified or schematic configurations, and some components are omitted.

[0047] (Overall configuration example) Fig. 1 is a diagram showing an example of the configuration of the entire system including the Baumkuchen baking system and Baumkuchen baking support system of this embodiment. In the example shown in Fig. 1, a Baumkuchen baking system 10, a Baumkuchen baking support system 20, and a remote terminal 40 are connected via a network so as to be able to communicate data with each other.

[0048] The Baumkuchen baking support system 20 is configured with a server. Hereinafter, the Baumkuchen baking support system 20 may also be referred to as the server 20. The Baumkuchen baking support system 20 can access a storage unit 30. The storage unit 30 stores a plurality of trained models in association with baking conditions (artisan data, as an example).

[0049] The Baumkuchen baking system 10 includes a Baumkuchen baking machine 1 and a control unit 8. The control unit 8 is configured by a computer. The control unit 8 automatically controls the baking of Baumkuchen by the Baumkuchen baking machine 1 using a trained model provided from the server 20. The Baumkuchen baking system 10 also learns the operation of the Baumkuchen baking machine under manual control by an operator and generates a trained model. Furthermore, the Baumkuchen baking system 10 receives operation instructions from a remote terminal and controls the Baumkuchen baking machine 1 to bake Baumkuchen.

[0050] Baumkuchen baking machine 1 has oven 2, dough container 4, rotating roll 3 that can move between baking position P1 of oven 2 and dough container 4, and camera 7. Camera 7 photographs a portion of the outer periphery of the Baumkuchen dough layered on rotating roll 3 at baking position P1. Control unit 8 controls the operation of moving rotating roll 3, on which Baumkuchen dough K is layered, from dough application position P2, where dough from dough container 4 is applied to the dough on rotating roll 3, to baking position P1 of oven 2, and the operation of moving from baking position P1 to dough application position P2.

[0051] The Baumkuchen baking system 10 is installed in a facility such as a pastry shop, bakery, or confectionery factory, for example. A plurality of Baumkuchen baking systems 10 installed in a plurality of facilities may be connected to the server 20 via a network. The remote terminal 40 is a terminal operated by a remote operator such as a craftsman or a consumer who is located away from the Baumkuchen baking machine 1. The remote terminal 40 is composed of a display device, an input device (touch panel, keyboard, buttons, mouse, etc.), and a computer equipped with a communication function.

[0052] The trained model is a trained model obtained by learning baking control based on images of the outer surface of the Baumkuchen dough that is layered on a rotating roll and baked. The training data used to train the trained model may include, for example, images of the outer surface of the Baumkuchen dough when a baker operates a baking machine similar to Baumkuchen baking machine 1 to bake the Baumkuchen, and a determination of the degree of doneness or baking control based on the baker's operation. The trained model is, for example, a dataset for executing a process that takes as input an image of the outer surface of the Baumkuchen dough that is layered on a rotating roll and baked, and outputs a value indicating the determination of the degree of doneness or baking control.

[0053] The trained model may be a model that inputs, in addition to the image of the outer periphery of the baumkuchen dough, other data detected during the baking of the baumkuchen. That is, the trained model may be a model obtained by learning to determine the degree of doneness or control the baking based on the image of the outer periphery of the dough and the data detected during baking. For example, the trained model may be a model that inputs, in addition to the image, at least one of the rotation speed of the baumkuchen dough stacked on the rotating roll, the baking time for the outer periphery of the baumkuchen dough, or the oven temperature. For example, at least one of the temperature inside the oven and the surface temperature of the dough may be input as the oven temperature.

[0054] The server 20 stores multiple trained models. Each of the multiple trained models is associated with baking conditions (for example, craftsman data). The craftsman data is data indicating the craftsman who contributed to the generation of the training data used to train the trained model. The craftsman who contributed to the generation of the training data can be, for example, the craftsman who performed the operation that formed the basis of the value indicating the doneness or baking control included in the training data. Note that the baking conditions are not limited to the craftsman data. Furthermore, the trained model does not have to be recorded in association with the baking conditions.

[0055] The server 20 provides the trained model to the Baumkuchen baking system 10. For example, the server 20 provides a trained model corresponding to baking conditions specified in the Baumkuchen baking system 10. The server 20 also acquires performance data indicating the performance of baking performed using the provided trained model from the Baumkuchen baking system 10. The performance data is recorded in the storage unit 30, for example. The server 20 can calculate various fees using the performance data. The server 20 is a system that enables the Baumkuchen baking system 10 to use the trained model on the cloud.

[0056] (Configuration example of Baumkuchen baking system 10) 1, the control unit 8 of the Baumkuchen baking system 10 has a communication unit 81, an automatic control unit 82, a user interface unit 83 (hereinafter referred to as the UI unit 83), a learning unit 84, and a remote control unit 85. The communication unit 81 is a functional block for communicating data with the outside via a network. The communication unit 81 is configured by, for example, a communication module included in the computer of the control unit 8.

[0057] The UI unit 83 is a user interface for inputting data from the operator and outputting data to the operator. The UI unit 83 is configured, for example, by an input / output interface included in the computer of the control unit 8. The UI unit 83 controls the information to be output to the operator and the information to be input from the operator. The UI unit 83 receives baking conditions for the Baumkuchen to be produced from the operator. The UI unit 83 receives the baker's specifications as an example of baking conditions.

[0058] The automatic control unit 82 uses a trained model provided by the server 20 to automatically control the baking of the dough for each layer of the Baumkuchen based on an image of the outer surface of the Baumkuchen dough K during baking taken by the camera 7. The automatic control unit 82 performs automatic control using a trained model corresponding to the baking conditions accepted by the UI unit 83. As an example, the automatic control unit 82 performs automatic control using a trained model of a craftsman designated by the operator.

[0059] The learning unit 84 generates a trained model by learning the doneness judgment or baking control based on the operator's operation of the Baumkuchen baking machine 1 and the images taken by the camera 7 during the operation. For example, the learning unit 84 estimates the doneness judgment result or baking control from the operator's operation during baking by manually controlling the Baumkuchen baking machine 1. The learning unit 84 can generate a trained model by learning using as training data the estimated judgment result or baking control and images taken of the outer surface of the dough at the baking position in the oven 2 during a period including the time of the operation that served as the basis for the estimation.

[0060] The remote control unit 85 makes it possible to control the baking operation of the Baumkuchen baking machine 1 in real time from the remote terminal 40. The remote control unit 85 transmits an image of the outer surface of the Baumkuchen dough during baking, taken by the camera 7, to the remote terminal 40 in real time. That is, the remote control unit 85 relays the image of the camera 7 during baking to the remote terminal 40. The remote control unit 85 controls the baking of the dough for each layer of the Baumkuchen in accordance with the operation instructions received from the remote terminal 40 while the image is being transmitted.

[0061] For example, the remote control unit 85 can receive an instruction to end baking from the remote terminal 40 while transmitting an image of the outer periphery of the baumkuchen dough being baked in the oven 2. The instruction to end baking is, for example, an instruction to remove the rotating roll 3 from the oven 2. This allows the operator to control the baking time of each layer from the remote terminal 40.

[0062] The remote control unit 85 may set a range of baking time that can be controlled by operation instructions from the remote terminal 40 for each layer of the Baumkuchen dough. For example, a lower limit and an upper limit of the baking time may be set. In this case, the lower limit may be the minimum time required to bake one layer of dough. The upper limit may be the time beyond which baking will result in overcooking.

[0063] The controllable range of baking time may be, for example, a predetermined period. Alternatively, the controllable range of baking time may be determined depending on the baking conditions. The baking conditions may be, for example, at least one of the oven temperature, the dough rotation speed, the dough ingredient composition, the dough physical properties, the core rod size, or the number of dough layers. Alternatively, the controllable range of baking time may be determined depending on the degree of doneness determined based on the image captured by camera 7 using a trained model.

[0064] The remote control unit 85 may use the trained model provided by the server 20 to determine the degree of doneness or baking control based on an image of the outer periphery of the baumkuchen dough during baking taken by the camera 7, and provide the determination result together with the image to the remote terminal 40 in real time. For example, information indicating the degree of doneness determined using the trained model or the time to stop baking can be provided to the remote terminal 40 together with the image.

[0065] The Baumkuchen baking machine 1 may include at least one of a rotation sensor that detects the rotation of the rotating roll 3 around its axis, a temperature sensor that detects the temperature of the oven 2, or a timer that measures the baking time.

[0066] The temperature sensor may be, for example, a thermometer that measures the temperature of the air inside the oven 2, a radiation thermometer that measures the temperature of the outer surface of the Baumkuchen, or a sensor that detects the temperature from output values ​​such as the temperature, current, and voltage of the heater of the oven 2. As an example, the temperature sensor may acquire both the temperature of the air inside the oven 2 and the temperature of the outer surface of the Baumkuchen dough.

[0067] The rotation sensor may be, for example, a detector that optically, magnetically, or mechanically detects the movement of a detected element that rotates together with the axis of the rotating roll 3. Alternatively, the rotation sensor may be configured to detect the rotation of the rotating roll 3 from the output value of a motor that controls the rotation of the rotating roll 3.

[0068] The timer may be, for example, a part of the control unit 8. The timer can measure the baking time, for example, by measuring the time that has elapsed since the rotating roll 3 was placed in the baking position.

[0069] The camera 7 of the Baumkuchen baking machine 1 may be installed so that its imaging range covers a part in the axial direction of the outer peripheral surface of the dough K stacked on the rotating roll 3. This makes it possible to obtain, with a simple configuration, an image suitable for determining the degree of doneness of the outer peripheral surface of the rotating Baumkuchen dough.

[0070] The camera 7 may be installed, for example, to capture an image showing the entire radial direction of the dough K stacked on the rotating roll 3 at the baking position. The control unit 8 may acquire an image of a portion of the dough in the radial direction that is cut out from the image captured by the camera 7. In this case, the control unit 8 performs automatic control processing or learning processing using the image of the portion of the dough K in the radial direction. By using the image of the portion of the dough K in the radial direction, an image of a portion of the entire image of the dough in the radial direction that clearly shows the degree of doneness in color can be used for the automatic control processing or learning processing.

[0071] The number of cameras 7 may be one or more. The optical axis of the camera 7 may be positioned so as to intersect with the axial direction of the rotating roll 3. The camera 7 may be positioned, for example, outside the oven 2 to capture an image of the outer periphery of the dough inside the oven 2 through a window of the oven 2. The camera 7 and the rotating roll 3 may also be configured so that the relative position of the optical axis of the camera 7 and the rotating roll 3 at the baking position P1 of the oven 2 is fixed. This allows the conditions for capturing images of the dough at the baking position by the camera 7 to be fixed. The relative position of the optical axis of the camera 7 and the heater of the oven 2 may also be fixed at the baking position of the oven 2. The light 72 is supported so as to be positionable in a position where it can irradiate the image capturing range of the camera 7 with light. For example, the light 72 may be positioned outside the oven 2 to irradiate the outer periphery of the dough inside the oven 2 through a window of the oven 2.

[0072] The control unit 8 has a processor and a memory. The control unit 8 may be composed of two or more computers. The control processing of the Baumkuchen baking machine by the control unit 8 can be realized by the processor executing a predetermined program. Programs that cause the control unit 8 to execute processing and non-transitory recording media on which the programs are recorded are also included in embodiments of the present invention. The control unit 8 may be built into the Baumkuchen baking machine 1, or may be communicably connected via a network to parts of the baking machine including the oven 2, rotating roll 3, and moving mechanism of the Baumkuchen baking machine 1.

[0073] The configuration of the control unit 8 is not limited to the example in Fig. 1. For example, one or two of the automatic control unit 82, the learning unit 84, and the remote control unit 85 may be omitted. For example, the control unit 8 may be configured to include a communication unit 81, a UI unit 83, and a remote control unit 85.

[0074] (Example of control unit configuration) 2 is a diagram showing an example of the configuration of the control unit 8 of the Baumkuchen baking system 10. In the example shown in Fig. 2, the Baumkuchen baking machine 1 includes a rotation sensor, a temperature sensor, and a timer. The temperature sensor measures the temperature of the oven 2.

[0075] When acquiring the image captured by the camera 7, the control unit 8 may further acquire at least one of the rotation speed of the Baumkuchen dough K stacked on the rotating roll 3, the temperature of the oven 2, or the baking time of the outer circumferential surface of the Baumkuchen dough K. That is, the control unit 8 may be configured to acquire at least one of the rotation speed detected by the rotation sensor, the temperature detected by the temperature sensor, or the baking time measured by the timer.

[0076] The rotation speed may be, for example, the rotation speed of the rotating roll 3, or the speed at which the outer surface of the dough moves in the circumferential direction. The temperature of the oven 2 may be, for example, the surface temperature of the outer surface of the dough, the temperature of the air inside the oven 2, or the temperature of the heat source of the oven 2. The baking time of the outer surface of the dough is the baking time for one layer of dough. For example, the baking time may be the time elapsed from the time the rotating roll moves from the dough application position to the baking position in the oven.

[0077] The control unit 8 also has a control command unit 86 that sends commands to the movement mechanism, the rotating roll 3, and the oven 2. The movement mechanism is a mechanism that moves the rotating roll 3 between the baking position P1 of the oven 2 and the dough container 4. The movement mechanism may include, for example, a support member that rotatably supports the shaft of the rotating roll, and an actuator that moves the shaft of the rotating roll supported by the support member. The support member may be, for example, a movable arm or a guide such as a rail. The actuator may be, for example, a motor, a hydraulic cylinder, or another power source. The control command unit 86 sends commands to the actuator. By controlling the drive of the actuator, the movement of the rotating roll 3 between the baking position P1 and the dough application position P2 of the oven 2 can be controlled.

[0078] The movable arm may be configured such that one end is rotatably supported by a pivot shaft relative to the Baumkuchen baking machine and the other end rotatably supports the rotation shaft of the rotating roll. In this case, the actuator may include a motor that rotates the movable arm about the pivot shaft. For example, a pair of movable arms may be provided that rotatably support both ends of the rotating roll in the axial direction.

[0079] The operation of the moving mechanism to move the rotating roll from the baking position of the oven to the dough application position may be an operation of moving at least one of the rotating roll and the dough container to bring them closer to each other. For example, the rotating roll may be moved closer to the dough container, or the dough container may be moved closer to the rotating roll.

[0080] The control command unit 86 controls the rotation of the rotating roll 3. For example, the rotation around the axis of the rotating roll 3 and the movement in a direction perpendicular to the axis of the rotating roll 3 are controlled. The control command unit 86 may also control the heater of the oven 2. That is, the control unit 8 can also control the temperature of the oven 2.

[0081] The UI unit receives data input from an operator and outputs information to the operator via an input / output device provided in the Baumkuchen baking machine 1. The input / output device may include, for example, a touch panel, a button, a lever, a keyboard, or a mouse.

[0082] In the example of FIG. 2, the control unit 8 is connected to an operation reception unit included in the Baumkuchen baking machine 1. The operation reception unit receives operations for the Baumkuchen baking machine 1 from an operator. The operation reception unit is composed of, for example, an operation panel and operators such as operation buttons. The operation reception unit can receive, from the operator, operations for, for example, the operation of the rotating roll 3 and the temperature of the oven 2. As operations for the rotating roll 3, the operation reception unit receives from the operator an operation to move the rotating roll 3 to the batter application position, an operation to move the rotating roll 3 from the batter application position to the baking position of the oven 2 (a baking start operation), an operation to move the rotating roll 3 from the baking position to the batter application position (a baking end operation), and an operation to control the rotation speed of the rotating roll 3. When baking is performed under manual control by the operator, the rotating roll 3 is controlled in accordance with the operation on the operation reception unit.

[0083] <Configuration example of automatic control unit> 2, the automatic control unit 82 includes an image acquisition unit 821 and a determination unit 822. The image acquisition unit 821 acquires from the camera 7 a group of images of at least one circumference of the outer periphery of the dough K that rotates with the rotation of the rotating roll 3 at the baking position P1 of the oven 2. The image acquisition unit 821 acquires, for example, from the camera 7, a group of images of a portion of the outer periphery of the dough that are captured at a predetermined interval. Each of the acquired images may be an image in which a portion of the dough in the radial direction is cut out from an image that captures the entire dough in the radial direction.

[0084] The determination unit 822 determines the timing for moving the rotating roll 3 from the baking position of the oven 2 to the dough application position P2 based on the browning of the outer peripheral surface of the dough K shown in the group of multiple images of at least one circumference of the outer peripheral surface of the dough K at the baking position P1 of the oven 2. This controls the baking time for one layer.

[0085] The determination unit 822 determines the timing using a trained model provided by the server 20. For example, the determination unit 822 uses the trained model to execute a process in which an image of the outer periphery of the dough is input and an evaluation value of the degree of doneness is output. For example, the determination unit 822 can sequentially determine the degree of doneness of a group of multiple images, and when the degree of doneness determined from the images satisfies a predetermined condition, determine that it is time to move the rotating roll from the baking position to the dough application position.

[0086] The automatic control unit 82 determines a trained model according to the baking conditions received by the UI unit 83. The automatic control unit 82 determines the trained model corresponding to the baking conditions input by the operator as the model to be used for automatic control, from among multiple trained models that the server 20 can provide. The automatic control unit 82 may download multiple trained models from the server 20 in advance, or may download from the server 20 a trained model determined according to the baking conditions.

[0087] For example, if craftsman data as baking conditions are associated with a trained model, the UI unit 83 can present selectable craftsmen to the operator and accept the craftsman's selection input. In this case, a learning process is performed in advance for each craftsman, a trained model is generated for each craftsman, and the trained model is recorded in the storage unit 30. This prepares a trained model that enables the reproduction of doneness judgment or baking control that reflects the craftsman's individuality.

[0088] The automatic control unit 82 may determine the degree of doneness or control the baking process based on at least one of the rotation speed of the dough K, the temperature of the oven 2, or the baking time, in addition to the image captured by the camera 7. In this way, using at least one of the rotation speed, temperature, and baking time allows for a determination that takes into account the influence of at least one of the rotation speed, temperature, and baking time on the degree of doneness. As a result, it is possible to control the timing of moving the rotating roll from the baking position to the dough application position to achieve a more appropriate degree of doneness. In this case, the trained model may be a trained model obtained by learning to determine the degree of doneness or control the baking process based on at least one of the rotation speed, temperature, and baking time, in addition to the image of the outer surface of the dough.

[0089] The automatic control unit 82 may acquire at least one of the moving speed in the circumferential direction of the outer surface of the Baumkuchen dough K layered on the rotating roll 3, or the outer diameter of the Baumkuchen dough K layered on the rotating roll 3. The automatic control unit 82 may determine the degree of doneness or the baking control based on at least one of the moving speed and the diameter acquired in addition to the image captured by the camera 7. In this case, the trained model may be a trained model obtained by learning the degree of doneness determination or the baking control based on at least one of the moving speed and the diameter in addition to the image of the outer surface of the dough.

[0090] As the number of dough layers on the rotating roll 3 increases, the diameter of the dough also increases. As the diameter of dough K increases, the circumferential movement speed of the outer surface of the dough increases, even if the rotation speed of the shaft of the rotating roll 3 remains the same. By including the circumferential movement speed of the outer surface of the dough or the diameter of the outer surface of the dough in the input data of the trained model and using this for judgment, it becomes possible to make judgments that take into account differences in baking conditions due to the layering of dough. As a result, automatic control can be performed to more appropriately bake each layer.

[0091] The control unit 8 may calculate the circumferential movement speed of the outer circumferential surface of the fabric based on the outer diameter of the fabric stacked on the rotating roll 3 and the rotation speed of the rotating roll. The outer diameter of the fabric can be obtained, for example, by measuring the radial dimension of the fabric in an image that shows the entire radial view of the fabric stacked on the rotating roll 3.

[0092] The configuration of the automatic control unit 82 is not limited to the example shown in Fig. 2. In Fig. 2, the trained model is acquired from the server 20 via the communication unit 81 and used in the automatic control process. In contrast, for example, the automatic control unit 82 may be configured to transmit an image of the camera 7 to the server 20 and receive from the server 20 a result of a determination made by the server 20 using the trained model based on the image.

[0093] <Example of learning section configuration> In the example shown in FIG. 2, the learning unit 84 has a judgment and estimation unit 841. The judgment and estimation unit 841 estimates the operator's judgment of the degree of doneness of the Baumkuchen dough or baking control based on the operator's manual control of the baking operation on the Baumkuchen baking machine 1. The learning unit 84 generates, as training data, the estimated operator's judgment of the degree of doneness or baking control and an image of the outer periphery of the dough when the manual control baking operation is performed. The images used as training data can be, for example, a group of multiple images taken of at least one circumference of the outer periphery of the dough K at the baking position P1 of the oven 2 during a period that includes the time of manual control operation.

[0094] The judgment and estimation unit 841 may infer the operator's judgment of the degree of doneness, for example, based on whether or not the operator has performed an operation to move the rotating roll 3 on which the Baumkuchen batter K has been layered from the baking position P1 to the batter application position P2. The operator performs an operation to move the rotating roll 3 on which the Baumkuchen batter K has been layered from the baking position P1 to the batter application position P2 when the operator has determined that the degree of doneness of the batter is appropriate. Therefore, the movement operation performed by the operator can indicate the result of the judgment of the degree of doneness.

[0095] For example, if the operator does not move the rotating roll 3 from the baking position to the batter application position P2 while the rotating roll 3 on which the dough is layered is rotating at the baking position P1 of the oven 2, the judgment and estimation unit 841 can infer that the operator has determined that the dough is not baked sufficiently, that is, that the dough is not cooked enough. In this case, an image of the outer surface of the dough taken by the camera 7 while the rotating roll 3 is rotating is recorded as training data in association with the result of the determination that the dough is not cooked enough.

[0096] Furthermore, for example, when the operator moves the rotating roll 3 from the baking position P1 to the batter application position P2 while the rotating roll 3 on which the dough is layered is rotating at the baking position P1 of the oven 2, the judgment and estimation unit 841 can infer that the operator has judged that the dough is well-baked. In this case, during the period including the time of the operation, an image of the outer surface of the dough taken by the camera 7 while the rotating roll 3 makes at least one rotation is recorded as training data in association with the result of the judgment of well-baked dough.

[0097] The learning unit 84 may generate the trained model by further using at least one of the rotation speed of the Baumkuchen dough K layered on the rotating roll 3, the temperature of the oven, or the baking time of the outer peripheral surface of the Baumkuchen dough during a period including the time when the operator determines the operation. The learning unit 84 may generate the trained model by further using at least one of the circumferential movement speed of the outer peripheral surface of the Baumkuchen dough K layered on the rotating roll 3, or the outer diameter of the Baumkuchen dough K during a period including the time when the operator determines the degree of doneness.

[0098] The learning unit 84 provides the generated trained model to the server 20 via the communication unit 81. The learning unit 84 may associate the generated trained model with craftsman data and provide them to the server 20. The craftsman data is data that indicates, as a craftsman, the operator who performed the operation that served as the basis for the training data in generating the trained model. The learning unit 84 may also provide other baking conditions of the craftsman data to the server 20 in association with the trained model. For example, the baking conditions for baking Baumkuchen that served as the basis for generating the training data for the trained model may be provided to the server 20 in association with the trained model.

[0099] Here, the process of generating a trained model by learning using the training data may be executed outside the learning unit 84. The learning unit 84 may generate the training data and provide it to the server 20. Note that the learning unit 84 may determine parameters of the training model using the training data to generate the trained model.

[0100] The learning unit 84 can generate training data for multiple Baumkuchen. That is, the learning unit 84 can generate training data for multiple Baumkuchen baking runs. Note that the training data for one Baumkuchen (i.e., one Baumkuchen baking run) may include image and operation data for multiple layers. The UI unit 83 may receive, from the operator, designation of training data to be used for learning from the training data for multiple Baumkuchen. In this case, the training data for the Baumkuchen designated by the operator is provided to the server 20. Alternatively, a trained model generated using the training data designated by the operator is provided to the server 20. Note that the designation of training data received by the UI unit 83 may be designation of training data for a Baumkuchen to be excluded from that used for learning.

[0101] <Example of remote control unit configuration> The remote control unit 85 transmits the image captured by the camera 7 to the remote terminal 40 in real time via the communication unit 81. In addition to the image captured by the camera 7, the remote control unit 85 may provide the remote terminal 40 in real time with at least one of the rotation speed of the dough K detected by the rotation sensor, the baking time detected by the timer, and the temperature of the oven 2.

[0102] Furthermore, the control instructions that the remote control unit 85 receives from the remote terminal 40 may include instructions to control the rotation of the rotating roll 3 or the temperature of the oven 2, in addition to instructions to control the baking time.

[0103] For example, the remote control unit 85 can receive an instruction from the remote terminal 40 to change the rotation speed of the rotating roll 3 while transmitting an image of the outer circumferential surface of the baumkuchen dough being baked in the oven 2. In this case, the remote control unit 85 may set a range of the rotation speed of the rotating roll 3 that can be controlled by an instruction from the remote terminal 40.

[0104] The range of the controllable rotation speed may be, for example, a predetermined period. Alternatively, the range of the controllable rotation speed may be determined depending on the baking conditions. Alternatively, the range of the controllable rotation speed may be determined depending on the degree of doneness determined based on the image captured by the camera 7 using a trained model.

[0105] The remote control unit 85 may use the trained model provided by the server 20 to determine a preferred rotation speed based on an image of the outer surface of the Baumkuchen dough during baking taken by the camera 7, and provide the same to the remote terminal 40 in real time.

[0106] For example, the remote control unit 85 can receive an instruction from the remote terminal 40 to change the temperature of the oven 2 while transmitting an image of the outer periphery of the dough of the Baumkuchen being baked in the oven 2. In this case, the remote control unit 85 may set a range of the temperature of the oven 2 that can be controlled by an instruction from the remote terminal 40.

[0107] The temperature range of the oven 2 that can be remotely controlled may be, for example, a predetermined period. Alternatively, the controllable temperature range may be determined according to the baking conditions. Alternatively, the controllable temperature range may be determined according to the degree of doneness determined based on the image of the camera 7 using a trained model.

[0108] The remote control unit 85 may use the trained model provided by the server 20 to determine the preferred temperature of the oven 2 based on an image of the outer surface of the Baumkuchen dough during baking taken by the camera 7, and provide the temperature to the remote terminal 40 in real time.

[0109] The remote control unit 85 may switch between permission and denial of control of the Baumkuchen baking machine 1 from the remote terminal 40 in accordance with a switching operation by the operator input from the UI unit 83 or the operation reception unit. This makes it possible to avoid unintended remote control by the operator while the operator is preparing to bake the Baumkuchen baking machine 1, for example.

[0110] In addition, the remote control unit 85 may output information indicating that the Baumkuchen baking machine 1 is in a state where it can be controlled from the remote terminal 40 to the operator via a display device of the input / output device or another notification device (speaker, lamp, etc.) provided in the Baumkuchen baking machine.

[0111] The remote control unit 85 may automatically control the baking of each layer of baumkuchen dough using a trained model provided by the server 20, and may also transmit an image of the outer surface of the dough to the remote terminal 40 in real time. In this case, the remote control unit 85 also receives control instructions from the remote terminal 40. In this manner, the remote control unit 85 may control the baking of each layer of dough using both automatic control using the trained model and control instructions from the remote terminal 40. Even in this form, the range of the baking time and other baking conditions that can be controlled by the control instructions from the remote terminal 40 may be set. The controllable range may be determined in advance, or may be set based on a detected image and other baking conditions. The controllable range may also be determined based on the determination result using the trained model.

[0112] The remote control unit 85 may accept the designation of a craftsman from the remote terminal 40 and perform automatic control based on the judgment results using a trained model corresponding to the designated craftsman, or provide the judgment results to the remote terminal 40 in real time.

[0113] As a modified example, the remote terminal 40 may determine the degree of doneness or the baking control based on the image using a trained model provided by the server 20. In this case, the remote control unit 85 provides the image of the camera 7 to the remote terminal 40 in real time. The remote terminal 40 determines the degree of doneness or the baking control based on the image provided in real time using the trained model provided by the server 20. The remote control unit 85 receives the determination result made in the remote terminal 40 using the trained model, and automatically controls the baking of each layer of the Baumkuchen dough using the determination result.

[0114] (Server (Baumkuchen baking support system) configuration example) In the example shown in FIG. 1 , the server 20 has a model providing unit 11, a baking result receiving unit 12, a model registration unit 13, and an accounting unit 14. The model providing unit 11 provides a trained model to the Baumkuchen baking system 10. For example, the trained model may be provided in response to a request from the Baumkuchen baking system 10, or multiple trained models may be provided in advance to the Baumkuchen baking system 10 all at once. Furthermore, training data may be provided as the trained model. In this case, the control unit 8 of the Baumkuchen baking system 10 generates a trained model based on the training data.

[0115] The baking result receiving unit 12 receives result data from the Baumkuchen baking system 10. The result data is data indicating the usage result of the trained model provided to the Baumkuchen baking system 10. In other words, the result data indicates the result of baking Baumkuchen by automatically controlling the Baumkuchen baking machine 1 using the trained model provided by the model providing unit 11. The result data includes, for example, information specifying the trained model used and information about the Baumkuchen manufactured using the trained model. The information about the manufactured Baumkuchen may include, for example, at least one of the amount of Baumkuchen manufactured, the type of Baumkuchen, the time of manufacture, or the baking conditions.

[0116] The settlement unit 14 uses the performance data to calculate fees related to the use of the trained model. As an example, a usage fee to be charged to an operating entity that manufactured baumkuchen using the trained model and a remuneration amount to a craftsman who contributed to the generation of the used trained model are calculated. The usage fee for the trained model may be, for example, a fee based on the amount or type of baumkuchen manufactured using the trained model. Furthermore, when the trained model is associated with craftsman data, the usage fee may be calculated based on a basic fee preset for the craftsman. The remuneration amount to the craftsman may be an amount based on the amount or type of baumkuchen manufactured using the trained model. Furthermore, the remuneration amount may be calculated based on a basic fee preset for the craftsman.

[0117] The model registration unit 13 receives the trained model generated by the Baumkuchen baking system 10 and records it in the storage unit 30. The model registration unit 13 may receive the training data generated by the Baumkuchen baking system 10 as a trained model. In this case, the model registration unit 13 may record the training data as a trained model in the storage unit 30, or may record the trained model generated based on the training data in the storage unit 30.

[0118] The model registration unit 13 can receive the trained model and the corresponding craftsman data from the Baumkuchen baking system 10. In this case, the model registration unit 13 associates the received trained model and craftsman data with each other and records them in the storage unit 30. This allows a trained model to be prepared for each craftsman.

[0119] The server 20 is composed of one or more computers. The functions of the model providing unit 11, firing result receiving unit 12, model registration unit 13, and settlement unit 14 can be realized by the computer's processor executing a predetermined program. Programs that cause the server 20 to execute processes and non-transitory recording media on which the programs are recorded are also included in embodiments of the present invention. The server 20 may be composed of multiple computers connected to each other via a network. The memory unit 30 is composed of a storage device that can be accessed by the computers that make up the server 20.

[0120] (Example of automatic control using a trained model) FIG. 3 is a flowchart showing an example of a process for baking Baumkuchen by automatic control using a trained model provided by a server in the Baumkuchen baking system 10.

[0121] In the example shown in FIG. 3, as a preparatory step for baking, an operator puts baumkuchen dough into the dough container 4 (Op1). The dough can be dough with a predetermined composition (ingredients). Here, the composition (ingredients combination) of the dough put into the dough container 4 may be the same as the composition (ingredients combination) of the dough used in the baking of the baumkuchen performed to generate the trained model provided by the server 20. This can further improve the quality of the baumkuchen baked by automatic control using the trained model. The operator turns on the heater switch of the oven 2 to start heating (Op2).

[0122] The UI unit 83 accepts input of baking conditions from the operator (S101). The UI unit 83 accepts input of baking conditions such as the type of dough, the size of the rotating roll 3 (core rod), the number of layers of dough to be baked, and the chef's designation. The UI unit 83 presents the operator with chefs whose chef data is recorded in association with multiple trained models in the memory unit 30, and accepts their selection. The operator can select a desired chef from among multiple chefs who have trained models.

[0123] The automatic control unit 82 determines a trained model corresponding to the baking conditions input by the operator in S101 (S102). The automatic control unit 82 can, for example, acquire, from the storage unit 30, a trained model associated with craftsman data indicating the craftsman specified by the operator.

[0124] The automatic control unit 82 automatically controls the Baumkuchen baking machine 1 using the trained model determined in S102 to bake the Baumkuchen (S103). A detailed example of the processing in S103 will be described later.

[0125] When the baking of the Baumkuchen is completed, the automatic control unit 82 transmits performance data indicating the performance of the Baumkuchen baking using the trained model to the server 20. The performance data includes, for example, an identifier for identifying the Baumkuchen baking system 10, the date and time of baking, the number of times, the type of dough, the size of the core rod, the number of layers of dough, and data for specifying the trained model.

[0126] When baking is complete, the operator removes the baked Baumkuchen together with the rotating roll 3 from the oven 2 (Op3). This allows high-quality Baumkuchen to be produced through baking under automatic control using the trained model.

[0127] (Remote control example) Fig. 4 is a flowchart showing a processing example when the Baumkuchen baking system 10 bakes Baumkuchen by remote control. In the example shown in Fig. 4, as a preparatory work for baking, an operator (hereinafter referred to as an on-site operator) at the installation location of the Baumkuchen baking machine 1 puts Baumkuchen batter into the batter container 4 (Op1). In addition, the on-site operator turns on the heater switch of the oven 2 to start heating (Op2).

[0128] The remote control unit 85 establishes a connection with the remote terminal 40, and sets the control unit 8 in a state where data communication with the remote terminal 40 is possible (S201). The remote control unit 85 accepts input of baking conditions from the remote terminal 40 (S202). A remote operator at a remote location inputs the baking conditions via the remote terminal 40. Note that the input of baking conditions may also be accepted by the UI unit 83 from an on-site operator. For example, the remote control unit 85 can accept a designation of a craftsman from the remote terminal 40, and the UI unit 83 can accept input of baking conditions related to the dough, number of layers, and core rod size from the on-site operator.

[0129] The remote control unit 85 determines the trained model to be used for remote control based on the baking conditions input in S202 (S203). For example, in S202, the trained model corresponding to the craftsman specified by the remote operator via the remote terminal 40 is determined as the trained model to be used for remote control. The remote control unit 85 acquires the determined trained model from the server 20 and makes it available to the control unit 8.

[0130] The remote control unit 85 starts real-time transmission of images from the camera 7 (S204). The remote control unit 85 executes baking control processing in accordance with control instructions from the remote terminal 40 (S205). Fig. 5 is a flowchart showing an example of the control processing of S205. Fig. 6 is a diagram showing an example of a screen displayed on the remote terminal 40 in the processing of S205.

[0131] In the example of FIG. 5, the remote control unit 85 initializes the layer counter n (n=1) (S301). The remote control unit 85 applies the nth layer of dough to the rotating roll 3. The remote control unit 85 instructs the control command unit 86 to perform the application operation. The control command unit 86 moves the rotating roll 3 to the dough application position P2 and rotates the rotating roll 3.

[0132] After the nth layer of batter is applied, the remote control unit 85 starts baking the nth layer of batter (S303). The remote control unit 85 instructs the control command unit 86 to start baking. The control command unit 86 moves the rotating roll 3 from the batter application position P2 to the baking position P1 and rotates the rotating roll 3.

[0133] The remote control unit 85 starts the nth layer's doneness determination using the trained model (S304). Using the trained model, the remote control unit 85 calculates a doneness determination value based on an image of the outer surface of the dough during baking taken by the camera 7. While the rotating roll 3 makes one rotation at the baking position, multiple images are taken by the camera 7. The remote control unit 85 calculates a determination value for the multiple images. For example, a determination value may be calculated for each of the multiple images.

[0134] In the example shown in FIG. 6, the screen RG displayed on the remote terminal 40 displays the current image R1 from the camera 7, the current level of doneness R2, the result of the doneness determination using the trained model R3 (including the name of the chef who contributed to the generation of the trained model), the current layer of dough R4, the operation status (whether baking or batter application is in progress) R5, and a button BN1 for instructing the remote terminal 40 to end baking. The remote control unit 85 provides this current information to the remote terminal 40 in real time and causes it to be displayed. Note that the display items on the screen RG of the remote terminal 40 are not limited to these. For example, the rotation speed of the rotating roll (core rod), temperature (e.g., at least one of the temperature inside the oven and the temperature on the surface of the dough), or the baking time of the current layer (the time elapsed since the start of baking), etc. may also be displayed on the screen RG.

[0135] The current level of doneness can be determined, for example, based on the elapsed time since the start of baking, i.e., the baking time measured by a timer. Alternatively, the current level of doneness can be determined based on the results of a trained model's determination of the degree of doneness based on the current image. Note that the trained model's determination of the degree of doneness can be based on at least one of the rotation speed, temperature, and baking time in addition to the image.

[0136] In the example shown in FIG. 6, the upper and lower limit levels of the controllable range are indicated by arrows. The upper and lower limit levels of the controllable range may be determined, for example, according to the upper and lower limits of a predetermined elapsed time. Alternatively, at least one of the upper and lower limit levels may be determined based on the current image. This determination may be based on the rotation speed, temperature, or baking time in addition to the image. This determination may also use the results of a judgment using a trained model. As an example, the predicted time until the dough burns can be calculated from the results of the doneness determination based on the image and temperature, and the upper limit level can be determined based on the predicted time.

[0137] In the example shown in Figure 6, the recommended level of doneness is indicated by an arrow. The recommended level may be determined using a judgment result based on an image using a trained model. For example, the remaining baking time for the current image may be calculated using a trained model that takes an image as input and outputs the remaining baking time. The recommended level of doneness can be determined based on the remaining baking time.

[0138] In S305 of Figure 5, it is determined whether the current baking time, i.e., the time elapsed since the start of baking, has reached the lower limit T1 of the preset controllable range. After the baking time has reached the lower limit T1, the processes of S306 to S308 are executed. If an instruction to end baking is input from the remote terminal 40 (S306), the remote control unit 85 controls the end of baking (S310). In S310, the remote control unit 85 instructs the control command unit 86 to perform the baking end operation. The control command unit 86 moves the rotating roll 3 from the baking position P1 to the batter application position P2.

[0139] If there is no instruction to end baking in S307, and there is an update to the baking degree judgment value based on the image using the trained model (YES in S307), the display of the judgment result on the remote terminal 40 is updated. The displayed judgment result may be, for example, information indicating preferable control based on the baking degree, such as continuing baking or ending baking, or information indicating the level of baking. An example of calculating the judgment value will be described later.

[0140] If the baking time exceeds the upper limit T2 of the controllable range (YES in S309) without receiving an instruction to end baking (NO in S306), the remote control unit 85 controls the end of baking (S310), thereby preventing the baking time for one layer from exceeding the upper limit T2 of the controllable range.

[0141] When S310 is executed and the firing of one layer is completed, it is determined whether the current layer number n has reached the target layer number N1 (S311). If n < N1, 1 is added to n, and the processes of S302 to S311 are repeated. As a result, the processes of S302 to S310 are repeated the number of times of the target layer number N1. That is, the control of the firing of the target layer number N1 is executed. When the firing process of the target layer number N1 is completed, the on-site operator takes out the rotary roll 3 of Baumkuhen from the oven 2 (Op3 in FIG. 5).

[0142] By the above remote control, the remote operator can control the firing time while checking the current baking color of the dough. Since the lower limit T1 and the upper limit T2 are set as the controllable range of the firing time, for example, even if the remote operator is inexperienced, it is possible to avoid each layer from being significantly underbaked or overbaked. In addition, since the image determination result using the learned model is displayed on the remote terminal in real time, the remote operator can grasp the appropriate timing of the end of firing to some extent.

[0143] Note that the remote control operation is not limited to the above example. The display of the image determination result using the learned model and the setting of the controllable range may be omitted. Also, the remote control is not limited to the control of the firing time. Instead of or in addition to the control of the firing time, at least one of the rotation speed of the rotary roll 3 or the temperature of the oven 2 may be controlled from the remote terminal 40.

[0144] (Operation example of server) FIG. 7 is a diagram showing an example of the operation of the server 20. In the example shown in FIG. 7, the server 20 receives baking conditions from the Baumkuchen baking system 10 (S401). The baking conditions include, for example, a baker designated by an operator in the Baumkuchen baking system 10. In addition to the baker, the baking conditions may also include, for example, conditions related to the Baumkuchen dough or conditions related to the size of the Baumkuchen. Examples of conditions related to the dough include conditions for the dough composition (ingredients combination) or physical properties. Examples of conditions for the dough composition (ingredients combination) include the types or proportions of eggs, flour, and butter contained in the dough, and conditions for toppings contained in the dough. Examples of conditions for the physical properties of the dough include the dough temperature, dough specific gravity, and dough viscosity.

[0145] The server 20 provides the Baumkuchen baking system 10 with a trained model determined based on the baking conditions received in S401 (S402). For example, in the example shown in FIG. 1, a plurality of trained models are recorded in the storage unit 30 accessible by the server 20. Each of the plurality of trained models is recorded in association with the craftsman data. In this case, the server 20 obtains the trained model corresponding to the craftsman received in S401 from the storage unit 30 and provides it to the Baumkuchen baking system 10.

[0146] The server 20 receives performance data indicating the usage performance of the trained model provided in S402 from the Baumkuchen baking system 10 that provided the trained model. The performance data includes, for example, information identifying the trained model and information indicating the Baumkuchen baked (manufactured) using the trained model.

[0147] The server 20 uses the performance data received in S403 to calculate the amount of fees or remuneration incurred by using the provided trained model (S404). For example, the usage fee for the provided trained model and the remuneration amount to the craftsman who contributed to the generation of the trained model are calculated. The server 20 may provide the usage fee for the trained model and the remuneration amount to the craftsman calculated in S404 to a payment system. The payment system executes a process for charging the usage fee for the trained model and a process for paying the remuneration to the craftsman.

[0148] In this embodiment, the trained model is stored in the server 20. The trained model is generated by machine learning and used to digitize the part where a baker judges the degree of doneness or controls the baking by looking at the color of the baumkuchen dough. By providing this trained model to a baumkuchen baking system that has a camera and a control unit, the artisan's skills can be efficiently reproduced by automatic control. In other words, the baumkuchen artisan's skills are stored in the form of a trained model in the server 20, and the baumkuchen baking system can be used.

[0149] The server 20 stores the trained model and manages its use. This allows for appropriate protection and use of the trained model. If the skills of a Baumkuchen craftsman were to be used in the form of a trained model, many facilities would be able to bake high-quality Baumkuchen without the need for a craftsman. On the other hand, if the craftsman's skills were to be widely reproduced, the value of the skills cultivated over many years of experience could be reduced. Therefore, in this embodiment, as an example, the server 20 stores the craftsman data in association with the trained model. In this case, the server 20 provides the trained model of the specified craftsman, so it can grasp and manage its usage status. It can also determine an appropriate remuneration amount for the craftsman depending on usage.

[0150] (Example of Baumkuchen baking machine configuration) Figure 8 is a front view of a Baumkuchen baking machine in this embodiment. Figure 9 is a side view of the Baumkuchen baking machine shown in Figure 8. The Baumkuchen baking machine 1 shown in Figures 8 and 9 includes an oven 2, a rotating roll 3 that rotates while stacking Baumkuchen batter, a dough container 4 that stores the Baumkuchen batter before baking, a movement mechanism (5, 6) that moves the rotating roll 3 between a baking position in the oven and a batter application position, and a control unit 8 that controls the operation of the Baumkuchen baking machine.

[0151] The Baumkuchen baking machine 1 further includes a camera 7, lighting 72, and various sensors (not shown in FIG. 8). The various sensors may include, for example, at least one of a temperature sensor that measures the temperature of the oven, a rotation sensor that detects the rotation speed of the Baumkuchen batter stacked on the rotating roll 3, and a timer that measures the baking time of the Baumkuchen batter.

[0152] The camera 7 is positioned at a position where it can photograph a portion of the outer peripheral surface of the Baumkuchen dough K layered on the rotating roll 3 at the baking position. The optical axis of the camera 7 intersects with the outer peripheral surface of the Baumkuchen dough K. The camera 7 is supported by a support member 71. The support member 71 fixes the relative position of the optical axis of the camera 7 and the rotating roll 3 at the baking position. The light 72 irradiates light onto an area included in the photographing range of the camera 7. The light 72 is supported by the support member.

[0153] The camera 7 takes a plurality of images of the outer circumferential surface while the rotating roll 3 on which the Baumkuchen dough K is layered makes at least one revolution. The camera 7, for example, takes a video of the rotating Baumkuchen dough K. This allows a group of a plurality of images of the outer circumferential surface of the Baumkuchen dough taken for at least one revolution to be obtained.

[0154] The oven 2 is a heating furnace and is equipped with a heater 22 inside. The oven 2 has an openable and closable window 21. A dough container 4 is placed in front of the window 21. The dough container 4 is placed on a stand 41.

[0155] In the example shown in Figs. 8 and 9, Baumkuchen dough layered on a rotating roll 3 is placed at a baking position inside an oven 2. Both ends of the rotating roll 3 are rotatably supported by a pair of arms 5. The rotating roll 3 is rotated by, for example, a motor (not shown). The control unit 8 can control the rotation of the rotating roll 3 by controlling this motor.

[0156] The pair of arms 5 are attached to the Baumkuchen baking machine 1 so as to be rotatable around a pivot axis PA. An actuator 6 is connected to the arms 5. The arms 5 rotate when the actuator 6 is driven. The actuator 6 is, for example, a motor. The drive of the actuator 6 is controlled by a control unit 8. The control unit 8 controls the rotation of the arms 5 by controlling the drive of the actuator 6. The position of the rotating roll 3 is controlled by controlling the rotation of the arms 5. In this example, the arms 5 and the actuator 6 form a movement mechanism for the rotating roll 3.

[0157] The control unit 8 controls the position of the rotating roll 3 to move the rotating roll 3, on which the Baumkuchen dough K is layered, between a dough application position and a baking position in the oven 2. The dough application position is a position where the dough from the dough container 4 is applied to the dough on the rotating roll 3. FIG. 10 is a diagram showing the rotating roll 3 at the dough application position. The dough application position is a position above the dough container 4. As the rotating roll 3 rotates at the dough application position, more dough is applied to the outer peripheral surface of the Baumkuchen dough K layered on the rotating roll 3. The configuration for detecting the position of the rotating roll 3 by the control unit 8 is not particularly limited. For example, a position detection sensor that detects the position of the rotating roll 3 or the arm 5 may be provided in the Baumkuchen baking machine 1. Alternatively, the control unit 8 may be configured to detect the position of the rotating roll 3 based on the operation of the actuator 6.

[0158] The control unit 8 rotates the rotating roll 3 at least one revolution at the batter application position, thereby applying one layer of Baumkuchen batter K to the rotating roll 3. The control unit 8 moves the rotating roll 3, on which the Baumkuchen batter K has been applied, from the batter application position to the baking position in the oven 2. This starts baking the applied batter for one layer.

[0159] The control unit 8 acquires from the camera 7 a group of multiple images capturing at least one circumference of the outer periphery of the Baumkuchen batter that rotates with the rotation of the rotating roll 3 at the baking position of the oven 2. The control unit 8 determines the degree of doneness based on the browning of the outer periphery of the Baumkuchen batter shown in the group of multiple images captured by the camera 7. The control unit 8 determines the timing for moving the rotating roll 3 from the baking position of the oven 2 to the batter application position based on the result of the determination of the degree of doneness. As a result, if the degree of doneness is determined to be satisfactory, the control unit 8 can move the rotating roll 3 from the baking position of the oven 2 to the batter application position. Baking ends when the rotating roll 3 moves from the baking position of the oven 2 to the batter application position. In other words, the control unit 8 determines the degree of doneness of one layer of Baumkuchen batter and controls the baking time for one layer of batter to achieve the appropriate degree of doneness.

[0160] The control unit 8 controls the position of the rotating roll 3, and the operation of spreading the baumkuchen batter and baking it in the oven 2 is repeated multiple times. In this way, multiple layers of baumkuchen batter are baked. When baking each layer, the baking time is controlled based on the image from the camera 7 to ensure that the layer is baked to the appropriate degree.

[0161] (Example of control processing) FIG. 11 is a flowchart showing an example of automatic control processing of the Baumkuchen baking machine 1 by the control unit 8. In the example shown in FIG. 11, the control unit 8 causes the Baumkuchen baking machine 1 to apply one layer of batter to the rotating roll 3 and bake it. The control unit 8 rotates the rotating roll 3 at the batter application position, and applies one layer of batter to the outer circumferential surface of the batter stacked on the rotating roll 3 (S1). After application, the control unit 8 moves the rotating roll 3 from the batter application position to the baking position of the oven 2 (S2). This starts baking. In the baking process, the rotating roll 3 on which the Baumkuchen batter is stacked rotates at the baking position of the oven 2.

[0162] The control unit 8 acquires from the camera 7 an image of the outer circumferential surface of the dough rotating with the rotation of the rotating roll 3 (S3). FIG. 12 is a diagram showing an example of an image acquired by the camera 7. In the example shown in FIG. 12, the camera 7 captures an image of a portion of the dough K stacked on the rotating roll 3 in the axial direction, covering the entire area in the radial direction. The control unit 8 then cuts out and acquires an image of the radial center portion A1 of the dough K from this image. That is, an image of the dough region that does not include the radial end Ke of the dough K shown in the image is cut out. This makes it possible to acquire an image of a portion of the dough that clearly shows the degree of doneness on the outer circumferential surface. For example, the color of the dough near the radial end Ke of the dough K shown in the image is easily affected by light from the heater 22, etc. By cutting out the image of the radial center portion A1 of the dough K, it is possible to acquire an image of a portion that is less affected by light from the heater 22, etc.

[0163] The control unit 8 acquires sensor data in synchronization with the acquisition of the image (S4). The sensor data includes, for example, the surface temperature of the outer surface of the Baumkuchen detected by a temperature sensor (radiation thermometer). In addition, the baking time measured by a timer is acquired as sensor data. The baking time is the elapsed time from the start of baking.

[0164] The control unit 8 uses the trained model provided by the server 20 to determine a value for the degree of doneness based on the image acquired in S3 and the sensor data acquired in S4 (S5). That is, the control unit 8 determines the degree of doneness based on the browning of the outer periphery of the dough shown in the image, the surface temperature of the dough, and the baking time. The trained model can be data generated by deep learning using a neural network. That is, the control unit 8 can determine the degree of doneness from the image and sensor data using artificial intelligence technology using a neural network.

[0165] FIG. 13 is a diagram showing an example of the configuration of a neural network used in the determination process. In the example shown in FIG. 13, a portion of the surface of a baumkuchen is cut out from a color camera image. The cut-out image is input to a convolutional neural network LS1. The convolutional neural network LS1 outputs 32 parameters (features). Furthermore, the values ​​of the baking time and the surface temperature of the baumkuchen are input to a fully connected layer L1, which has, for example, five units. This fully connected layer L1 outputs five parameters. The 32 parameters and the five parameters are concatenated and further input to a fully connected layer L2. The output from this fully connected layer L2 is input to the next fully connected layer L3, which outputs a determination value (for example, 0 to 1).

[0166] In the example shown in FIG. 13, a camera image is input to a convolutional neural network, and sensor data is input to a fully connected layer. The image features that have passed through the convolutional neural network and the sensor data parameters that have passed through the fully connected layer are concatenated and further input to a fully connected layer. A doneness determination value is output through this fully connected layer and another fully connected layer. In this way, a machine learning model can be configured to include a convolutional neural network that converts the input image into features, a first input layer that inputs sensor data, a second input layer that inputs parameters that combine the image features and the output of the sensor data input layer, and a layer that further converts the output of the second input layer. In this way, by using a neural network configuration that combines the image and sensor data, it is possible to determine the doneness based on the image and sensor data. Note that the configuration of the neural network used in the determination process is not limited to the example shown in FIG. 13. For example, the number of layers in the fully connected layer and the number of parameters can be set as needed. Furthermore, the sensor data input to the fully connected layer L1 is not limited to the example shown in FIG. 13. For example, at least one value of the rotation speed of the rotating roll 3, the temperature of the oven, or the baking time may be input to the full bonding layer L1.

[0167] If the judgment value of the degree of doneness determined in S5 satisfies a predetermined condition (YES in S6), the control unit 8 moves the rotating roll 3 from the baking position in the oven 2 to the batter application position, and ends baking. For example, if the judgment value is equal to or greater than a predetermined threshold value, the control unit 8 determines that the baumkuchen is done, and causes the movement mechanism to remove the baumkuchen from the oven.

[0168] If the determination value of the degree of doneness determined in S5 does not satisfy the predetermined condition (NO in S6), the control unit 8 returns to S3, acquires images, and repeats the processes of S4 to S6. In the example shown in Fig. 11, the process of determining the degree of doneness is performed for each of a group of multiple images. As a result, the process of determining the degree of doneness and judging whether baking is complete based on the determination is performed for a group of multiple images taken while the rotating roll 3 makes at least one rotation.

[0169] FIG. 14 is a diagram showing an example of a group of images acquired by the control unit 8 from the start of baking to the end of baking. For example, after baking starts, one image is captured by the camera 7 at a predetermined interval (for example, every 0.5 seconds). The control unit 8 sequentially acquires the images captured by the camera 7. In the example shown in FIG. 14, n images G1 to Gn are acquired. For images G1 to G(n-1), the determination value for the degree of doneness does not satisfy the condition, but for the nth image Gn, the determination value for the degree of doneness does satisfy the condition. Baking is terminated when the nth image Gn is acquired.

[0170] In the above example, the degree of doneness and the end of baking are determined for each image, but the degree of doneness and the end of baking may be determined for multiple images.

[0171] In the above example, the baking time and temperature are acquired as sensor data. The control unit 8 may also acquire the rotation speed of the Baumkuchen dough layered on the rotating roll as sensor data. The control unit 8 can acquire the rotation speed of the rotating roll 3 detected by the rotation sensor in S4 of FIG. 11. The control unit 8 may determine the degree of doneness based on the image and the rotation speed. The control unit 8 can also acquire the circumferential movement speed of the outer surface of the Baumkuchen dough from the image and rotation speed of the camera 7.

[0172] For example, the diameter D1 of the outer periphery of the dough layered on the rotating roll 3 can be measured from the image shown in FIG. 12. The diameter D1 obtained from the image and the rotation speed of the rotating roll 3 obtained from the rotation sensor can be used to calculate the circumferential movement speed of the outer periphery of the dough. The control unit 8 may determine the degree of doneness using the circumferential movement speed of the outer periphery of the dough and the image. This makes it possible to make a determination that takes into account changes in baking conditions due to the amount of dough layered. The control unit 8 may also determine the degree of doneness using the diameter D1 and the image. In this case, it is also possible to make a determination that takes into account changes in baking conditions due to the amount of dough layered.

[0173] (Learning process example) FIG. 15 is a flowchart showing an example of processing for collecting training data for learning processing based on the baking operation of the Baumkuchen baking machine 1. The example shown in FIG. 15 is an example in which the Baumkuchen baking machine 1, in accordance with the operation of the operator, applies batter for one layer to the rotating roll 3 and bakes the batter. In accordance with the operation of the operator, the Baumkuchen baking machine 1 rotates the rotating roll 3 at the batter application position and applies batter for one layer to the outer peripheral surface of the batter stacked on the rotating roll 3 (S11). After application, the operator operates the rotating roll 3 to move from the batter application position to the baking position of the oven 2 (S12). This starts baking. In the baking process, the rotating roll 3, on which the Baumkuchen batter is stacked, rotates at the baking position of the oven 2.

[0174] The control unit 8 acquires an image of the outer circumferential surface of the fabric rotating with the rotation of the rotating roll 3 from the camera 7 (S13). The image acquisition process can be executed, for example, in the same manner as S3 in FIG. 11. The control unit 8 acquires sensor data in synchronization with the image acquisition (S14). The sensor data acquired is data from the same sensor as the data acquired in S5 in FIG. 11.

[0175] The Baumkuchen baking machine 1 accepts an operation to end baking while the Baumkuchen is being baked (S15). Specifically, while the rotating roll 3 on which the batter is layered is rotating at the baking position of the oven 2, the operator can at any time operate the Baumkuchen baking machine 1 to move the rotating roll 3 from the baking position to the batter application position. When the operator operates to move the rotating roll 3 from the baking position to the batter application position, baking ends.

[0176] If the operator does not issue an operation to end baking for a certain period of time during baking (NO in S16), the control unit 8 assumes that the operator has determined that the bread is not yet baked. In this case, the control unit 8 associates the determination result of "not yet baked" with the image acquired in S13 and the sensor data acquired in S14, and records these as training data in the recording device. The control unit 8 then executes the image acquisition process of S13 again, and repeats the processes of S14 to S16. For example, the processes of S13 to S16 are executed for each of a group of multiple images taken during the period when the rotating roll 3 makes at least one revolution.

[0177] If the operator operates to end baking during baking (YES in S16), the control unit 8 assumes that the operator has judged the degree of doneness to be satisfactory. In this case, the control unit 8 associates the judgment result of satisfactory doneness with the image acquired in S13 and the sensor data acquired in S14, and records them as training data in the recording device. The operation to end baking is an operation to move the rotating roll 3 from the baking position to the batter application position. Baking ends when the rotating roll 3 moves from the baking position (S18).

[0178] 15, a group of multiple images of at least one circumference of the outer periphery of dough stacked on rotating roll 3 is recorded in association with the assessment result of doneness. Learning unit 84 of control unit 8 performs machine learning using the assessment result of doneness associated with the group of multiple images as training data to generate a trained model. Machine learning is not limited to this, but can be performed, for example, by deep learning using a neural network configured as shown in FIG. 13.

[0179] For example, an example of learning processing when a neural network model is used will be described. The learning unit 84 compares the output (determination result) of the pre-learning model obtained by inputting images and sensor data of teacher data into the pre-learning model with the determination result of the teacher data, and adjusts the weighting between layers to improve the degree of match. For example, in the case of a model configured as shown in FIG. 13, an image recorded as teacher data is input to the convolutional neural network LS1, and sensor data (e.g., baking time and surface temperature) recorded in association with the image is input to the fully connected layer L1. The output result (determination value) of the model in response to this input is compared with the determination result of the teacher data associated with the input image. The weighting parameters between each layer of the neural network are adjusted to improve the degree of match between the model output and the teacher data. A learning process is performed in which the weighting parameters of the model are adjusted using teacher data of a large number of images. The model whose weights have been adjusted by the learning process becomes the trained model.

[0180] The learning process of the control unit 8 is not limited to machine learning using a neural network. For example, other machine learning methods using regression analysis or decision trees may be used.

[0181] For example, when a skilled craftsman operates the Baumkuchen baking machine 1 to bake Baumkuchen, the control unit 8 can record the judgment result estimated from the operation, the image at the time of the judgment, and the sensor data in association with each other as training data. By performing machine learning using the training data recorded about the baking operation by such a craftsman, it is possible to generate a trained model that enables the same control of the baking time as that of this craftsman.

[0182] The process shown in FIG. 15 is an example of a process for generating training data for baking one layer. The process of FIG. 15 is repeatedly executed a predetermined number of layers until one baumkuchen is baked. In this way, training data for one baumkuchen is generated. For example, training data can be generated by repeating the process of FIG. 15 for the actions of one operator (craftsman) baking multiple baumkuchen. Training data for multiple baumkuchen by one operator is generated. The control unit may accept designation of training data to be used for learning from the training data for multiple baumkuchen by one operator.

[0183] FIG. 16 is a diagram showing a modified example of a Baumkuchen baking system and a Baumkuchen baking support system. In the example shown in FIG. 16, a plurality of trained models are stored in association with dough recipe data and baking conditions (artisan data, for example) in a storage unit 30 accessible by the server 20. The dough recipe data is data indicating the combination of ingredients for the Baumkuchen dough and the manufacturing procedure. The dough recipe data indicating the combination of ingredients for the dough and the manufacturing procedure used in baking when the training data for the trained model was generated is associated with the trained model. As a result, data indicating the dough ingredients, the method for making the dough, and the baking control for the Baumkuchen using that dough is recorded in the storage unit 30.

[0184] Note that a trained model and dough recipe data to which no artisan data is associated may be stored in the storage unit 30. For example, a trained model that has learned a standard baking method without being specialized for a particular artisan may be associated with dough recipe data that indicates the basic dough ingredient combination and preparation method and stored in the storage unit 30. Such a trained model and dough recipe data may be recorded as, for example, plain data. When an original dough recipe is devised in a store equipped with the Baumkuchen baking system 10, or when a basic recipe is sufficient, the operator can specify a trained model and dough recipe data to which no artisan data is associated (for example, plain data).

[0185] The Baumkuchen baking system 10 further includes a mixer 31. In the Baumkuchen baking system 10, the control unit 8 acquires the trained model and the dough recipe data associated therewith from the server 20. The control unit 8 outputs the dough recipe data to the operator in the preparation stage for baking using the trained model, i.e., before baking. The dough recipe data may be output, for example, as a video or still image on a display, printed on a printer, output by audio, or transmitted to the operator's terminal, or a combination of at least two of these. The operator can make the dough according to the output information.

[0186] The control unit 8 may, for example, acquire from the server 20 a trained model and dough recipe data corresponding to the craftsman data of a craftsman designated by the operator. The Baumkuchen baking system 10 bakes dough prepared according to the ingredient combination and preparation procedure indicated in the acquired dough recipe data, under control using the acquired trained model. This makes it possible to complete a Baumkuchen of substantially the same quality as a Baumkuchen prepared by a craftsman.

[0187] FIG. 17 is a diagram showing an example of information shown in dough recipe data. The dough recipe data in FIG. 17 includes the composition of dough ingredients. The data on the composition of dough ingredients includes the ingredients and the amounts of each ingredient. The amounts of dough ingredients may be expressed as a percentage (%) as in FIG. 17, or may be expressed in weight (g) or other units. The amounts may be omitted. The dough recipe data also indicates the timing for adding each ingredient to the mixer and the mixing conditions. The mixing conditions include the mixing time and the mixing speed (mixer rotation speed). In the example of FIG. 17, the dough recipe data also includes the temperature of the ingredients added to the mixer. The content of the dough recipe data is not limited to the example shown in FIG. 17. For example, some of the information shown in FIG. 17 may be omitted. The dough recipe data may also include at least one of the specific gravity or temperature of the dough at the end of mixing.

[0188] The control unit 8 may control the mixer 31 based on the dough recipe data. For example, the control unit 8 may notify the operator of the order and timing of adding each ingredient indicated in the dough recipe data, and may accept input from the operator that each ingredient has been added to the mixer 31. The control unit 8 can control the operation timing and speed of the mixer 31 based on the timing of adding each ingredient and the mixing time and mixing speed indicated in the dough recipe data.

[0189] Alternatively, the Baumkuchen baking system 10 may further include a feeder (not shown) that holds dough ingredients and feeds the ingredients into the mixer 31. The control unit 8 may control the ingredient feeding operation by the feeder according to the timing indicated by the dough recipe data for feeding the ingredients into the mixer 31. This makes it possible to automatically control the timing for feeding the ingredients into the mixer 31, i.e., the timing for starting to mix the ingredients, using the dough recipe data.

[0190] Each feeder may also be provided with a temperature regulator (e.g., a heater) to adjust the temperature of the ingredients. In this case, the control unit 8 may control the temperature regulator of each ingredient feeder based on the temperature of the ingredients indicated by the dough recipe data. This allows the temperature of the ingredients to be automatically adjusted using the dough recipe data.

[0191] (Other variations) In the above example, the control target based on the determination of the degree of doneness using the images from camera 7 is the baking time (the time when the rotating roll moves from the baking position), but the control target of control unit 8 is not limited to this. For example, at least one of the rotation speed of rotating roll 3 or the temperature of the oven can be controlled based on a group of multiple images of at least one rotation of the outer periphery of the Baumkuchen dough at the baking position in the oven, captured by camera 7. In this case, control unit 8 can use a trained model generated by machine learning to determine how to control at least one of the rotation speed or the oven temperature based on the group of multiple images. The trained model may be, for example, a dataset for executing a process that uses images of the outer periphery of the dough as input and outputs control information for at least one of the rotation speed or the oven temperature. Control unit 8 can generate the trained model using at least one of the rotation speed or the oven temperature detected during baking by the operator operating the Baumkuchen baking machine and the images from camera 7 as training data.

[0192] For example, the control unit 8 may adjust the rotation speed of the rotating roll 3 according to the change in diameter of the dough over time or the change in diameter in the axial direction (the uneven shape of the outer peripheral surface) shown in the image captured by the camera 7. Alternatively, the control unit 8 may adjust the heat of the heater 22 of the oven 2 according to the degree of doneness determined based on the image.

[0193] The control unit 8 may determine control of at least one of the rotation speed or oven temperature based on at least one of the rotation speed, baking time, and oven temperature acquired when acquiring the image groups, in addition to the multiple image groups. This determination process may use a dataset that uses at least one of the rotation speed, baking time, and oven temperature and an image of the outer periphery of the dough as input and control information as output as a trained model. The control unit 8 may also generate such a trained model based on the operator's operation of the Baumkuchen baking machine. For example, the control unit 8 may detect at least one operation of the rotation speed and the oven temperature by the operator, and generate a trained model using the detected operation and a group of images of the outer periphery of the dough captured during a period including the time of the operation detection as training data. Examples of the detected operation include an operation to adjust the rotation speed of the rotating roll 3 or an operation to adjust the temperature of the oven 2.

[0194] Furthermore, the sensor data used by the control unit 8 in the determination process is not limited to the rotation speed, oven temperature, and baking time listed in the examples above. One or two of these may be used in the determination process. Furthermore, other sensor data may be used in the determination process. For example, by performing the determination process using the rotation speed of the rotating roll 3 in addition to the image from the camera 7, it becomes possible to make a determination that takes into account changes in baking conditions due to the rotation speed. Furthermore, by performing the determination process using the oven temperature in addition to the image from the camera 7, it becomes possible to make a determination that takes into account changes in baking conditions due to the oven temperature. Furthermore, by performing the determination process using the baking time in addition to the image from the camera 7, it becomes possible to make a determination that takes into account changes in baking time.

[0195] Furthermore, dough information about the Baumkuchen dough may be used in the determination process. The dough information may include, for example, at least one of the physical properties of the dough in the dough container before application, or the composition of the dough ingredients (e.g., the composition or ratio of wheat, eggs, and butter, as well as ingredients to be topped on the dough, such as plain, chocolate, matcha, coffee, strawberry, etc.). The physical properties of the dough may be, for example, the dough temperature, the specific gravity of the dough, or the viscosity of the dough. For example, when the control unit 8 acquires multiple images of the outer periphery of the dough from the camera 7, the control unit 8 may also acquire dough information about the dough. In this case, in the determination process, the control unit 8 determines the timing to move the rotating roll 3 from the baking position in the oven to the dough application position based on the acquired dough information in addition to the browning of the outer periphery of the dough shown in the multiple image groups.

[0196] A trained model may be used in this determination process. The trained model may be, for example, data for inputting an image of the outer periphery of the dough and dough information and outputting a judgment result of the degree of doneness based on the browning indicated by the image. The control unit 8 can generate the trained model through machine learning using, as training data, the result of the operator's judgment of the degree of doneness estimated based on the operator's operation of the Baumkuchen baking machine, the dough information, and a group of multiple images captured of at least one circumference of the outer periphery of the dough at the baking position in the oven during a period including the time of the judgment.

[0197] The Baumkuchen baking machine 1 may include an input unit or a sensor that acquires dough information. The control unit 8 can acquire the dough information from the input unit or the sensor. For example, the Baumkuchen baking machine 1 may be provided with at least one of a weight sensor that measures the weight of the dough in the dough container 4, a dough temperature sensor that measures the temperature of the dough in the dough container 4, and a volume sensor that measures the volume of the dough in the dough container 4. Alternatively, the UI unit 83 may accept input of dough information from an operator.

[0198] Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments. [Explanation of symbols]

[0199] 1: Baumkuchen baking machine, 2: Oven, 3: Rotating roll, 4: Dough container, 5: Arm, 6: Actuator, 7: Camera, Control unit 8, 10: Baumkuchen baking system, 20: Server (Baumkuchen baking support system), 30: Memory unit

Claims

1. a communication unit that communicates data with the server; a baumkuchen baking machine including an oven, a dough container, a rotating roll that can move between a baking position of the oven and the dough container, and a camera that photographs a part of the outer periphery of the baumkuchen dough stacked on the rotating roll; a control unit that controls the Baumkuchen baking machine, The server can access a storage unit that stores a trained model obtained by learning to determine the degree of doneness or control baking based on an image of the outer surface of the baumkuchen dough that is stacked on a rotating roll and baked, The control unit has an automatic control unit that determines the degree of doneness or baking control using a trained model provided by the server based on an image of the outer periphery of the baumkuchen dough being baked at the baking position of the oven taken by the camera, and automatically controls the baking of the dough of each layer of the baumkuchen using the determination result, A plurality of the trained models are recorded in the storage unit accessible by the server, and each of the plurality of trained models is associated with and recorded with craftsman data indicating craftsmen who contributed to the generation of teacher data used in training each trained model; The control unit further includes a user interface unit that receives a designation of a craftsman from an operator, The automatic control unit is a Baumkuchen baking system that determines the degree of doneness or baking control based on the image taken by the camera using a trained model provided by the server and associated with craftsman data indicating the craftsman specified by the operator.

2. The Baumkuchen baking system according to claim 1, The Baumkuchen baking system further includes a remote control unit that provides a remote terminal capable of data communication via the communication unit with an image of the outer surface of the Baumkuchen dough being baked, taken by the camera, in real time, and controls the baking of the dough for each layer of the Baumkuchen in accordance with operation instructions received from the remote terminal.

3. The Baumkuchen baking system according to claim 2, The remote control unit uses a trained model provided by the server to determine the degree of doneness of the Baumkuchen dough or baking control based on an image of the outer surface of the Baumkuchen dough being baked at the baking position taken by the camera, and provides the determined information together with the image to the remote terminal in real time.

4. The Baumkuchen baking system according to any one of claims 1 to 3, The control unit further includes a learning unit that generates, as training data for learning, data indicating the baking control or the judgment result of the degree of doneness of the dough of each layer estimated from the operation of the Baumkuchen baking machine during baking by manual control by an operator, and an image of the outer surface of the Baumkuchen dough during baking by the manual control taken by the camera, The control unit provides the training data generated by the learning unit or the trained model generated by learning using the training data to the server via the communication unit.

5. The Baumkuchen baking system according to any one of claims 1 to 4, The control unit acquires from the server dough recipe data indicating the composition of dough ingredients and the procedure for making the dough, which is associated with the trained model provided by the server, and outputs the data to an operator of the Baumkuchen baking machine.

6. A Baumkuchen baking support system, A memory unit is accessible that records a trained model obtained by learning to judge the degree of doneness or control baking based on an image of the outer surface of the baumkuchen dough that is layered on a rotating roll and baked, a model providing unit that provides the trained model to a Baumkuchen baking system having a Baumkuchen baking machine, a camera, and a control unit; a baking result receiving unit that receives, from the Baumkuchen baking system, result data indicating the result of baking Baumkuchen by automatically controlling the Baumkuchen baking machine based on the image taken by the camera using the trained model provided by the model providing unit; The storage unit records a plurality of the trained models, and each of the plurality of trained models is recorded in association with craftsman data indicating craftsmen who contributed to the generation of teacher data used in training each trained model, The model providing unit provides the Baumkuchen baking system with a trained model associated with craftsman data indicating a craftsman input by an operator in the Baumkuchen baking system.

7. The Baumkuchen baking support system according to claim 6, further comprising: A Baumkuchen baking support system comprising: a model registration unit that receives from the Baumkuchen baking system a trained model generated by learning using, as training data, the judgment results or baking control of the degree of doneness of each layer of dough estimated from the operation of the Baumkuchen baking machine by manual control by an operator during baking, and an image taken by the camera of the outer surface of the Baumkuchen dough during the manually controlled baking, and records the trained model in the memory unit.

8. The Baumkuchen baking support system according to claim 6 or 7, further comprising: A Baumkuchen baking support system comprising an accounting unit that calculates, using the performance data received by the baking performance receiving unit, the usage fee for the trained model used in the Baumkuchen baking system and the remuneration to a craftsman indicated by the craftsman data corresponding to the trained model.

9. A program that can communicate with a server and causes a computer that controls a Baumkuchen baking machine to execute processing, The server is capable of accessing a storage unit that stores a plurality of trained models obtained by learning to determine the degree of doneness or control baking based on an image of the outer periphery of the dough of the Baumkuchen that is stacked on a rotating roll and baked, and artisan data associated with each of the plurality of trained models, the artisan data indicating the artisans who contributed to the generation of the teacher data used in training each trained model; The Baumkuchen baking machine includes an oven, a dough container, a rotating roll that is movable between a baking position of the oven and the dough container, and a camera that photographs a part of the outer circumferential surface of the Baumkuchen dough layered on the rotating roll, A process of receiving an instruction for automatic control using the trained model from an operator; A process of receiving a designation of a craftsman from the operator; A program that causes a computer to execute a process of determining the degree of doneness or baking control based on an image of the outer surface of the Baumkuchen dough being baked at the baking position in the oven taken by the camera, using a trained model provided by the server that is associated with craftsman data indicating the craftsman specified by the operator, and automatically controlling the baking of the dough for each layer of the Baumkuchen using the determination result.

10. A program that causes a computer that can communicate with a Baumkuchen baking system having a Baumkuchen baking machine, a camera, and a control unit to execute processing, A process of accessing a memory unit that records a trained model obtained by learning to judge the degree of doneness or control baking based on an image of the outer surface of the baumkuchen dough that is layered on a rotating roll and baked; A process of providing the trained model to the Baumkuchen baking system; and receiving, from the Baumkuchen baking system, performance data indicating the performance of baking Baumkuchen by automatically controlling the Baumkuchen baking machine based on the image captured by the camera using the provided trained model; The storage unit records a plurality of the trained models, and each of the plurality of trained models is recorded in association with craftsman data indicating craftsmen who contributed to the generation of teacher data used in training each trained model, The process of providing the trained model includes a process of providing the trained model associated with craftsman data indicating a craftsman input by an operator in the Baumkuchen baking system to the Baumkuchen baking system.

11. A method for producing Baumkuchen by controlling a Baumkuchen baking machine using a computer that can communicate with a server, comprising: The server is capable of accessing a storage unit that stores a plurality of trained models obtained by learning to determine the degree of doneness or control baking based on an image of the outer periphery of the dough of the Baumkuchen that is stacked on a rotating roll and baked, and artisan data associated with each of the plurality of trained models, the artisan data indicating the artisans who contributed to the generation of the teacher data used in training each trained model; The Baumkuchen baking machine includes an oven, a dough container, a rotating roll that is movable between a baking position of the oven and the dough container, and a camera that photographs a part of the outer circumferential surface of the Baumkuchen dough layered on the rotating roll, a step of receiving, by the computer, an instruction for automatic control using the trained model from an operator; a step in which the computer receives a designation of a craftsman from the operator; a step in which the computer determines the degree of doneness or baking control based on an image of the outer surface of the Baumkuchen dough being baked at the baking position of the oven taken by the camera, using a trained model provided by the server and associated with craftsman data indicating the craftsman specified by the operator, and automatically controls the baking of the dough for each layer of the Baumkuchen using the determination result.

Images