Roasting system, application program, and roaster

Abstract

To provide a technique which can perform suitable roasting according to a roasting object.SOLUTION: A roasting system includes: an attribute reception unit which receives attribute information of a roasting object; a roasting state reception unit which receives targeted roasting state information; a roasting profile calculation unit which calculates a targeted roasting profile which represents targeted roasting temperature variation in roasting the roasting object according to the attribute information and the roasting state information; a heater which heats the roasting object; a roasting temperature sensor which measures a roasting temperature; and a roasting control unit which controls the heater in accordance with the targeted roasting profile while referring to the roasting temperature measured by the roasting temperature sensor.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to a roasting system, an application program, and a roasting machine.

Background Art

[0002] Patent Document 1 discloses a coffee bean roaster. This roaster is devised so that the thin skin of coffee beans called chaff is less likely to scatter.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Even when using the above-described roaster, it is necessary to manually adjust the heat when roasting. However, beginners do not know how much heat and for how long to heat, so it has been difficult to roast. Further, it has been even more difficult to roast to a preferred flavor. Therefore, a technique capable of performing appropriate roasting according to the roasting target has been desired.

Means for Solving the Problems

[0005] According to a first aspect of the present disclosure, a roasting system for roasting a roasting target is provided. This roasting system includes an attribute reception unit that receives attribute information of the roasting target, a roasting state reception unit that receives target roasting state information, a roasting profile calculation unit that calculates a target roasting profile indicating a target roasting temperature change when roasting the roasting target according to the attribute information and the roasting state information, a heater that heats the roasting target, a roasting temperature sensor that measures the roasting temperature, and a roasting control unit that controls the heater according to the target roasting profile while referring to the roasting temperature measured by the roasting temperature sensor.

[0006] According to a second aspect of the present disclosure, an application program implemented on a roasting machine for roasting a roasting target and a mobile terminal communicably connected to a cloud computer is provided. This application program causes a processor of the mobile terminal to execute a process of receiving attribute information of the roasting target, a process of receiving target roasting state information, a process of transmitting the attribute information and the roasting state information from the mobile terminal to the cloud computer, causing a roasting profile calculation unit of the cloud computer to calculate a target roasting profile indicating a target roasting temperature change when roasting the roasting target according to the attribute information and the roasting state information, a process of obtaining the target roasting profile from the cloud computer, and a process of transmitting the target roasting profile to the roasting machine and causing the roasting machine to perform roasting according to the target roasting profile.

[0007] According to a third aspect of the present disclosure, there is provided a roasting machine communicably connected to a mobile terminal and a cloud computer, for roasting a roasting target. The mobile terminal includes an attribute receiving unit that receives attribute information of the roasting target, and a roasting state receiving unit that receives target roasting state information. The cloud computer includes a roasting profile calculation unit that calculates a target roasting profile indicating a target roasting temperature change when roasting the roasting target according to the attribute information and the roasting state information. The roasting machine includes a roasting profile acquisition unit that acquires the target roasting profile from the cloud computer via the mobile terminal, a heater that heats the roasting target, a roasting temperature sensor that measures the roasting temperature, and a roasting control unit that controls the heater according to the target roasting profile while referring to the roasting temperature measured by the roasting temperature sensor.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Embodiments for Carrying Out the Invention

[0009] FIG. 1 is a block diagram showing the configuration of a roasting system 10 in one embodiment. The roasting system 10 includes a user terminal 100, a cloud computer 200, and a roasting machine 300. The user terminal 100 is communicably connected to the cloud computer 200 and the roasting machine 300 via a wired or wireless line. The user terminal 100 is, for example, a smartphone or a personal computer. The user terminal 100 is also referred to as a "portable terminal."

[0010] The user terminal 100 includes a manual input unit 110, an attribute sensor 120, an attribute reception unit 130, a roasting state reception unit 140, a roasting data acquisition unit 150, and a roasting data storage unit 160. The functions of each part of the user terminal 100 may be realized by one or more processors executing an application program or a computer program, or may be realized by a hardware circuit.

[0011] In this embodiment, green coffee beans are the object to be roasted. However, various plant fruits, seeds, leaves, etc. may also be the objects to be roasted. For example, the present disclosure can also be applied when roasting barley seeds to make barley tea or when roasting tea leaves to make roasted green tea.

[0012] The attribute reception unit 130 receives attribute information BA indicating the attributes of the object to be roasted from at least one of the manual input unit 110 and the attribute sensor 120. The attribute information BA may include the coffee bean variety of the green coffee beans. The coffee bean variety is input by the user using the manual input unit 110. The coffee bean variety is, for example, the variety of coffee beans such as Brazil, Colombia, Guatemala, Mocha, Kilimanjaro, Blue Mountain, etc.

[0013] The attribute information BA may include the attribute values measured by the attribute sensor 120. Such attribute values may include the size and moisture content of the green coffee beans. For example, if an infrared sensor is used as the attribute sensor 120, various component amounts such as the moisture content of the green coffee beans, the amount of chlorogenic acid, and the amount of protein can be measured. Also, if a scale is used as the attribute sensor 120, the weight of the green coffee beans can be measured. Furthermore, if a camera is used as the attribute sensor 120, the size of the green coffee beans can be measured. The size of the green coffee beans is expressed in a unit called "screen size". "Screen size" corresponds to 1 / 64 inch. For example, "screen size 20" is 1 / 64 × 20 inches, which corresponds to 8 mm. Usually, the maximum value of the coffee beans shipped is "screen size 20".

[0014] The roasting state receiving unit 140 receives the roasting state information RC from the manual input unit 110. The roasting state information RC includes, for example, the roasting depth and the beverage taste data. The roasting depth is, for example, an index value representing four levels of roasting depth: light roast, medium roast, dark roast, and very dark roast, or eight levels of roasting depth obtained by further dividing each of these levels into two. The beverage taste data is data indicating the taste of coffee brewed using the roasted coffee beans. At this time, it is preferable to use a fully automatic coffee maker to keep the method of brewing coffee constant.

[0015] FIG. 2 is an explanatory diagram showing an example of the beverage taste data TD. This beverage taste data TD includes index values for four items: sweetness, sourness, bitterness, and strength. In this example, the individual index values are integer values from 0 to 5. The beverage taste data TD may include an index value indicating the aroma of coffee. These beverage taste data TD can be obtained by sensory evaluation.

[0016] The roasting data acquisition unit 150 acquires the attribute information BA received by the attribute reception unit 130 and the roasting state information RC received by the roasting state reception unit 140, and supplies them as roasting data to the roasting data storage unit 160. The roasting data storage unit 160 stores the roasting data supplied from the roasting data acquisition unit 150 in the memory within the user terminal 100.

[0017] The cloud computer 200 shown in FIG. 1 includes a learning data acquisition unit 210, a learning data storage unit 220, and a roasting profile calculation unit 230. The functions of each part of the cloud computer 200 may be realized by one or more processors on the cloud executing a computer program, or may be realized by a hardware circuit.

[0018] The learning data acquisition unit 210 acquires the learning data used for the learning of the roasting profile calculation unit 230 from the roasting data storage unit 160 and the manual input unit 110 of the user terminal 100, and stores it in the learning data storage unit 220. The learning data includes the attribute information BA of the roasting target, the roasting state information RC, and the roasting profile RP described later. The attribute information BA and the roasting state information RC are supplied from the roasting data storage unit 160. The roasting profile RP is input by the manual input unit 110 and supplied to the learning data acquisition unit 210 via the roasting data storage unit 160. The roasting profile calculation unit 230 executes learning using this learning data.

[0019] FIG. 3 is an explanatory diagram showing the functions of the roasting profile calculation unit 230. The roasting profile calculation unit 230 has a machine learning model for calculating a target roasting profile RPt. This machine learning model LM is configured as a multi-layer perceptron having an input layer IL, an intermediate layer HL, and an output layer OL, and functions as a regression analysis model. In this example, the input layer IL has eight input nodes. Roasting state information RC including four index values of the beverage taste data TD and an index value of the roasting intensity is input to five of the eight input nodes. Attribute information BA including the bean variety, the size of the raw beans, and the moisture content of the raw beans is input to the other three input nodes. The output layer OL has n output nodes, and is configured to output n target roasting temperatures T1 to Tn from these output nodes. These target roasting temperatures T1 to Tn are temperatures for n roasting times, and represent the target roasting profile RPt. n is an integer of 2 or more, and is set to a value of about 100 to 200, for example. When the target roasting profile RPt shows the roasting temperature change for 30 minutes, if the target roasting temperatures T1 to Tn are values at 10-second intervals, then n = 180.

[0020] FIG. 4 is an explanatory diagram showing an example of the target roasting profile RPt. The horizontal axis is the roasting time j, and the vertical axis is the target roasting temperature Tj. j is represented by an integer value from 1 to n. In a typical example, the upper limit value of the roasting temperature Tj is 230°C. The reason why the roasting temperature once decreases in the target roasting profile RPt of FIG. 4 is that the roasting machine 300 is preheated, and the roasting temperature once drops when the raw beans are put in.

[0021] Before inputting the input data into the machine learning model LM shown in FIG. 3, preprocessing may be performed on the input data. In this preprocessing, each input data is normalized to a real number value in the range of 0.0 to 1.0. For example, the beverage taste data TD normalizes the levels from 0 to 5 to real number values in the range of 0.0 to 1.0. The size of the green coffee beans is also normalized to a real number value in the range of 0.0 to 1.0 by dividing the actually measured size by the maximum size of the green coffee beans, which is 8 mm. Also, the moisture content of the green coffee beans is normalized to a real number value in the range of 0.0 to 1.0 by the value measured with a moisture meter.

[0022] The target roasting temperature Tj output from the machine learning model LM takes a real number value in the range of 0.0 to 1.0. The value of each target roasting temperature Tj is converted to a temperature in the range of 0 to 230 °C, for example, by multiplying by 230.

[0023] As the input data input to the machine learning model LM, some of the items of the roasting state information RC and the attribute information BA shown in FIG. 3 may be omitted. Also, other items other than the items shown in FIG. 3 may be input.

[0024] The roasting profile calculation unit 230 may use a configuration other than the machine learning model LM. For example, it may calculate the target roasting profile RPt using a calculation formula, a function, or a look-up table.

[0025] As shown in FIG. 1, the roasting machine 300 includes a roasting profile acquisition unit 310 that acquires the target roasting profile RPt, a roasting control unit 320, a roasting temperature sensor 330 that measures the roasting temperature inside the roasting machine 300, and a heater 340 that heats the roasting target. The roasting control unit 320 performs roasting by controlling the heater 340 according to the target roasting profile RPt while referring to the roasting temperature measured by the roasting temperature sensor 330. The roasting temperature sensor 330 is, for example, a contact thermometer installed at the position of the coffee beans inside the roasting machine 300. Alternatively, an infrared thermometer that measures the temperature of the coffee beans inside the roasting machine 300 non-contact may be used.

[0026] Note that it is possible to arbitrarily change which device among the user terminal 100, the cloud computer 200, and the roasting machine 300 implements each part shown in FIG. 1. For example, the function of the roasting profile calculation unit 230 may be implemented in the user terminal 100. Also, all the functions of each part of the user terminal 100 may be implemented in the roasting machine 300.

[0027] FIG. 5 is a flowchart showing the procedure of the learning process of the roasting profile calculation unit 230. In step S11, the learning data acquisition unit 210 receives the input of the attribute information BA. In step S12, the learning data acquisition unit 210 receives the input of the roasting depth. In step S13, the learning data acquisition unit 210 receives the input of the roasting profile RP. The data input in steps S11 to S13 are data for constituting the learning data.

[0028] In step S14, roasting of coffee beans is performed using the roasting machine 300. At this time, the roasting control unit 320 controls the heater 340 according to the roasting profile RP input in step S13. In step S15, the learning data acquisition unit 210 receives the input of the beverage taste data TD. That is, for coffee brewed using the roasted coffee beans, the input of the beverage taste data TD obtained by a sensory test is received. In step S16, the learning data acquisition unit 210 creates learning data using the data obtained in steps S11 to S15.

[0029] In step S17, the learning data acquisition unit 210 determines whether a sufficient amount of learning data has been collected for the same roasting target. If a sufficient amount of learning data has not been collected, the process returns to step S12, at least one of the roasting depth and the roasting profile RP is changed, and the processes in steps S14 to S16 are executed again. On the other hand, if a sufficient amount of learning data has been collected, the process proceeds to step S18.

[0030] In step S18, the roasting profile calculation unit 230 performs learning of the machine learning model LM using the collected learning data. In this learning, an appropriate loss function and optimization algorithm are selected to optimize the parameters of the machine learning model LM. As the loss function, for example, L1 norm or L2 norm can be used. As the optimization algorithm, for example, ADAM can be used. Note that the learning data may be periodically additionally collected to remake the learning data with the latest data. Also, the machine learning model LM may be validated to evaluate the prediction performance of the learned machine learning model LM. In this way, it can be verified whether the roasting profile can be appropriately predicted for unknown data.

[0031] FIG. 6 is a flowchart showing the procedure of the roasting process. In step S21, the roasting profile calculation unit 230 receives the attribute information BA and the roasting state information RC as roasting conditions. When the roasting state information RC includes the beverage taste data TD as in the example of FIG. 3 described above, a value indicating the user's preferred taste is specified.

[0032] In step S22, the roasting profile calculation unit 230 performs preprocessing of the roasting conditions as necessary. As described above, this preprocessing is a process of normalizing the input data of the machine learning model LM to real values in the range of 0 to 1.

[0033] In step S23, the roasting profile calculation unit 230 calculates a target roasting profile RPt according to the roasting conditions. The target roasting profile RPt is data indicating the target roasting temperature change when roasting the green coffee beans to be roasted. In the example of FIG. 1, the target roasting profile RPt is transferred from the cloud computer 200 to the roasting machine 300 via the user terminal 100. However, the target roasting profile RPt may be directly transferred from the cloud computer 200 to the roasting machine 300.

[0034] In step S24, the roasting machine 300 performs roasting according to the target roasting profile RPt. That is, the roasting control unit 320 controls the heater 340 according to the target roasting profile RPt while referring to the roasting temperature measured by the roasting temperature sensor 330. As a result, the green coffee beans to be roasted can be roasted to a desirable roasted state.

[0035] As described above, in this embodiment, an appropriate target roasting profile RPt can be calculated according to the attribute information of the roasting target and the roasting state information, and appropriate roasting can be easily performed according to the target roasting profile RPt.

[0036] · Other forms: The present disclosure is not limited to the above-described embodiments, and can be implemented in various forms without departing from the gist thereof. For example, the present disclosure can also be implemented by the following aspects. The technical features in the above embodiments corresponding to the technical features in each of the following aspects can be appropriately replaced or combined in order to solve part or all of the problems of the present disclosure, or to achieve part or all of the effects of the present disclosure. Further, if the technical feature is not described as essential in this specification, it can be appropriately deleted.

[0037] (1) According to the first aspect of the present disclosure, a roasting system for roasting a roasting target is provided. This roasting system includes an attribute reception unit that receives the attribute information of the roasting target, a roasting state reception unit that receives the target roasting state information, a roasting profile calculation unit that calculates a target roasting profile indicating a target roasting temperature change when roasting the roasting target according to the attribute information and the roasting state information, a heater that heats the roasting target, a roasting temperature sensor that measures the roasting temperature, and a roasting control unit that controls the heater according to the target roasting profile while referring to the roasting temperature measured by the roasting temperature sensor. According to this roasting system, an appropriate target roasting profile can be calculated according to the attribute information and roasting state information of the object to be roasted, so that an appropriate roasting can be easily performed according to the target roasting profile.

[0038] (2) In the above roasting system, the attribute information may include the size and moisture content of the object to be roasted. According to this roasting system, appropriate roasting can be performed according to the size and moisture content of the object to be roasted.

[0039] (3) In the above roasting system, the roasting state information may include a roasting depth indicating the depth of roasting. According to this roasting system, the depth of roasting such as light roasting, medium roasting, dark roasting, and extra-dark roasting can be adjusted.

[0040] (4) In the above roasting system, the roasting state information may include taste data indicating the taste preference regarding the beverage extracted using the object to be roasted. According to this roasting system, roasting can be performed according to the taste preference regarding the beverage.

[0041] (5) According to the second aspect of the present disclosure, there is provided an application program implemented on a roasting machine that roasts an object to be roasted and a mobile terminal communicably connected to a cloud computer. This application program includes a process of receiving the attribute information of the object to be roasted, a process of receiving the target roasting state information, a process of transmitting the attribute information and the roasting state information from the mobile terminal to the cloud computer, and causing a roasting profile calculation unit of the cloud computer to calculate a target roasting profile indicating a target roasting temperature change when roasting the object to be roasted according to the attribute information and the roasting state information, a process of acquiring the target roasting profile from the cloud computer, and a process of transmitting the target roasting profile to the roasting machine and causing the roasting machine to perform roasting according to the target roasting profile, and causing the processor of the mobile terminal to execute these processes.

[0042] (6) According to a third aspect of the present disclosure, there is provided a roasting machine communicably connected to a mobile terminal and a cloud computer for roasting a roasting target. The mobile terminal includes an attribute reception unit that receives attribute information of the roasting target, and a roasting state reception unit that receives target roasting state information. The cloud computer includes a roasting profile calculation unit that calculates a target roasting profile indicating a target roasting temperature change when roasting the roasting target according to the attribute information and the roasting state information. The roasting machine includes a roasting profile acquisition unit that acquires the target roasting profile from the cloud computer via the mobile terminal, a heater that heats the roasting target, a roasting temperature sensor that measures the roasting temperature, and a roasting control unit that controls the heater according to the target roasting profile while referring to the roasting temperature measured by the roasting temperature sensor.

[0043] The present disclosure can also be implemented in various other forms other than the above. For example, it can be implemented in the form of a computer program for realizing the functions of the roasting system, a non-transitory storage medium recording the computer program, and the like.

Description of Reference Numerals

[0044] 10…Roasting system, 100…User terminal, 110…Manual input unit, 120…Attribute sensor, 130…Attribute reception unit, 140…Roasting state reception unit, 150…Roasting data acquisition unit, 160…Roasting data storage unit, 200…Cloud computer, 210…Learning data acquisition unit, 220…Learning data storage unit, 230…Roasting profile calculation unit, 300…Roasting machine, 310…Roasting profile acquisition unit, 320…Roasting control unit, 330…Roasting temperature sensor, 340…Heater

Claims

1. A roasting system for roasting a roasting target, an attribute receiving unit that receives attribute information of the roasting target, a roasting state receiving unit that receives target roasting state information, a roasting profile calculation unit that calculates a target roasting profile indicating a target roasting temperature change when roasting the roasting target according to the attribute information and the roasting state information, a heater that heats the roasting target, a roasting temperature sensor that measures the roasting temperature, a roasting control unit that controls the heater according to the target roasting profile while referring to the roasting temperature measured by the roasting temperature sensor, A roasting system comprising:

2. The roasting system according to claim 1, wherein the attribute information includes the size and moisture content of the roasting target. A roasting system.

3. The roasting system according to claim 1, wherein the roasting state information includes a roasting depth indicating the depth of roasting. A roasting system.

4. The roasting system according to claim 3, wherein the roasting state information includes taste data indicating taste preferences regarding a beverage extracted using the roasting target. A roasting system.

5. An application program implemented on a roasting machine for roasting a roasting target and a mobile terminal communicably connected to a cloud computer, a process of receiving attribute information of the roasting target, a process of receiving target roasting state information, Transmit the attribute information and the roasting state information from the mobile terminal to the cloud computer, and cause the roasting profile calculation unit of the cloud computer to calculate a target roasting profile indicating the target roasting temperature change when roasting the roasting target according to the attribute information and the roasting state information, and obtain the target roasting profile from the cloud computer; Transmit the target roasting profile to the roasting machine and cause the roasting machine to perform roasting according to the target roasting profile; An application program to be executed by the processor of the mobile terminal.

6. A roasting machine that is communicably connected to a mobile terminal and a cloud computer and roasts a roasting target, The mobile terminal includes an attribute reception unit that receives attribute information of the roasting target and a roasting state reception unit that receives target roasting state information. The cloud computer includes a roasting profile calculation unit that calculates a target roasting profile indicating the target roasting temperature change when roasting the roasting target according to the attribute information and the roasting state information. The roasting machine A roasting profile acquisition unit that acquires the target roasting profile from the cloud computer via the mobile terminal; A heater that heats the roasting target; A roasting temperature sensor that measures the roasting temperature; A roasting control unit that controls the heater according to the target roasting profile while referring to the roasting temperature measured by the roasting temperature sensor. A roasting machine comprising the above.

Images