RELATED APPLICATIONS

This application claims the benefit of United States provisional patent application 63/284,739 entitled "METHOD, APPARATUS AND SYSTEM FOR MONITORING A PLURALITY OF BARRELS CONTAINING A BEVERAGE", filed on December 1, 2021 and incorporated herein by reference in its entirety.

BACKGROUND

1. Field

This disclosure relates generally to containers for beverage production and more particularly to monitoring barrels containing a beverage.

2. Description of Related Art

Containers used in fermented beverage production such as winemaking have a significant effect on the properties of the beverage. Wooden barrels in particular impart certain qualities to the beverage during fermentation and aging of the beverage. Winemakers use sulfite additives to control oxidation and microbial growth in ageing wine. The concentration of sulfites in an ageing wine contained in a barrel is typically measured by sampling a small subset of barrels, or by mixing the wine together before sampling. There remains a need for improved methods for managing a barrel inventory used in beverage production.

SUMMARY

In accordance with one disclosed aspect there is provided a method performed by a host controller apparatus for monitoring a plurality of barrels containing a beverage. The method involves receiving an upload of sample data from a controller for each barrel of the plurality of barrels, the sample data including at least: a barrel identification identifying the barrel, and a free sulfur dioxide concentration (FSO2) value indicative of a concentration of free sulfur dioxide within a beverage contained in the barrel. The method also involves accumulating the sample data for the plurality of barrels to generate a barrel dataset, and performing a chemistry data analysis on the barrel dataset to generate a condition output, the condition output including information for performing a management function on at least one of the beverage, specific barrels in the plurality of barrels, or the plurality of barrels. In accordance with another disclosed aspect there is provided a host controller apparatus for monitoring a plurality of barrels containing a beverage. The apparatus includes a host processor operably configured to receive an upload of sample data from a controller for each barrel of the plurality of barrels, the sample data including at least a barrel identification identifying the barrel, and a free sulfur dioxide concentration (FSO2) value indicative of a concentration of free sulfur dioxide within a beverage contained in the barrel. The host processor is further operably configured to accumulate the sample data for the plurality of barrels to generate a barrel dataset, and perform a chemistry data analysis on the barrel dataset to generate a condition output, the condition output including information for performing a management function on at least one of the beverage, specific barrels in the plurality of barrels, or the plurality of barrels.

In accordance with another disclosed aspect there is provided a method for monitoring a plurality of barrels containing a beverage. The method involves for each barrel of the plurality of barrels, receiving sample data at a controller, the sample data including at least a barrel identification identifying the barrel, and a free sulfur dioxide concentration (FSO2) value indicative of a concentration of free sulfur dioxide within a beverage contained in the barrel. The method also involves uploading the sample data for the barrel from the controller to a host processor, and accumulating the sample data for the plurality of barrels at the host processor to generate a barrel dataset. The method further involves performing a chemistry data analysis on the barrel dataset to generate a condition output, and based on the condition output, performing a management function on at least one of the beverage, specific barrels in the plurality of barrels, or the plurality of barrels.

Receiving the sample data may involve causing the controller to initiate a sampling process to draw a fluid sample from the barrel and to deliver the fluid sample to a FSO2 sensor operable to generate the FSO2 value. The fluid sample may comprise a beverage sample.

Causing the controller to initiate the sampling process may involve causing the controller to communicate with the FSO2 sensor to initiate the sampling process.

Receiving the sample data may involve causing the controller to read a barrel identification associated with the barrel. The method may involve storing sample data on the controller for at least a portion of the plurality of barrels and uploading the sample data may involve uploading the stored sample data to the host processor.

Receiving sample data may further involve receiving, for each barrel in the plurality of barrels, at least one of a pH value indicative of the acidity of the beverage contained in the barrel, a volatile acidity indicative of an amount of acetic acid in the beverage, a temperature value indicative of the temperature of the beverage contained in the barrel, a color value indicative of a spectral reflectance, transmittance, or absorbance of the beverage, a turbidity value indicative of the clarity of the beverage, and a date and time associated with the sample data. Receiving sample data may, additionally or alternatively, further involve receiving, for each barrel in the plurality of barrels, at least one of a total SO2 concentration, an ethanol concentration, a 4- ethylphenol (4-EP) concentration and a 4-ethylguaiacol (4-EG) concentration.

The method may involve receiving SO2 dosage data representing an additional quantity of SO2 delivered to the barrel, the SO2 dosage data further including a time at which additional quantity of SO2 was delivered to the barrel and performing the chemistry data analysis may involve accounting for the additional quantity of SO2 delivered to the barrel.

Uploading may involve one of uploading the sample data from the controller to the host processor over a wired connection, uploading the sample data from the controller to the host processor over a wireless connection, or uploading the sample data from the controller via an internet connection to a remote database accessible by the host processor.

Receiving sample data at a controller may involve receiving the sample data at a device configured to initiate collection of the sample data and to wirelessly receive the sample data.

Accumulating the sample data may involve receiving the sample data at the host processor and writing the sample data to a database that is operably configured to store the barrel dataset.

Performing the chemistry data analysis may involve processing the barrel dataset through a chemistry data analyzer implemented using a set of computer codes that direct the host processor to perform the chemistry data analysis on the barrel dataset to generate the condition output. Performing the management function may involve generating an alert in response to generation of a condition output indicative of an adverse condition.

The method may involve receiving sensory data established by a human operator based on a sensory perception of the beverage in at least one barrel in the plurality of barrels, associating the sensory data with the beverage in the at least one barrel, and performing the chemistry data analysis may involve performing the chemistry data analysis based at least in part on the sensory data.

Receiving sensory data may involve causing a networked device to guide the human operator through a structured sensory analysis process to receive operator input of the sensory data, and receiving the sensory data may involve uploading the operator input sensory data from the networked device to the host processor.

The human operator may involve a plurality of human operators and the method may further involve processing the sensory data from each of the plurality of human operators to combine the sensory data for the at least one barrel.

Performing the chemistry data analysis may involve determining a metabolite concentration of a metabolite indicative of a beverage condition, the metabolite concentration based on the FSO2 values in the barrel dataset.

The method may involve receiving at least one environmental parameter associated with environmental conditions within a location housing the plurality of barrels and performing the chemistry data analysis may involve performing the chemistry data analysis to account for the at least one environmental parameter.

In accordance with another disclosed aspect there is provided a system for monitoring a plurality of barrels containing a beverage. The system includes a controller operably configured to receive sample data for each barrel of the plurality of barrels, the sample data including at least a barrel identification identifying the barrel, and a free sulfur dioxide concentration (FSO2) value indicative of a concentration of free sulfur dioxide within a beverage contained in the barrel. The system also includes a host controller operably configured to receive sample data for the barrel uploaded from the controller to the host processor, and accumulate the sample data for the plurality of barrels at the host processor to generate a barrel dataset. The host processor is further operably configured to perform a chemistry data analysis on the barrel dataset to generate a condition output, the condition output being indicative of a management function to be performed on at least one of the beverage, specific barrels in the plurality of barrels, or the plurality of barrels.

The system may include a FSO2 sensor operable to generate at least the FSO2 value of the sample data.

The controller may be embedded within the FSO2 sensor for generating and uploading the sample data. Other aspects and features will become apparent to those ordinarily skilled in the art upon review of the following description of specific disclosed embodiments in conjunction with the accompanying figures.

BRIEF DESCRIPTION OF THE DRAWINGS

In drawings which illustrate disclosed embodiments,

Figure 1 is a block diagram of a system for monitoring a plurality of barrels containing a beverage in accordance with a first disclosed embodiment;

Figure 2A is a view of one of the barrels in the plurality of barrels in the system of Fig. 1;

Figure 2B is a block diagram of a networked device 200for implementing a controller of the system shown in Fig. 1;

Figure 3 is a block diagram of a processor circuit implementation of a host controller shown in Fig. 1;

Figure 4A is a flowchart depicting blocks of code for directing the networked device of Fig. 2B and processor circuit of Fig. 3 to generate and upload a barrel dataset;

Figure 4B is a screenshot of a series of screens displayed by the controller of Fig. 2B;

Figure 4C is a representation of a set of relational database tables used in the database of the host processor shown in Fig. 3;

Figure 5A is a process flowchart depicting blocks of code for directing the networked device to receive sensory evaluation inputs;

Figure 5B is a screenshot of a series of screens displayed by the controller for implementing the process shown in Fig. 5A;

Figure 6A is a graphical depiction of free SO2 concentration and sulfite additions over time;

Figure 6B is a table of discrete free SO2 concentration and sulfite addition values over time; Figure 6C is a table showing a correspondence between continuous functions and discrete functions used for implementing a chemical analyzer on the host processor shown in Fig. 3;

Figure 7 is a graphical depiction for results for two example barrels Barrel 1 and Barrel 2 generated by the chemical analyzer on the host processor shown in Fig. 3; and

Figure 8 is a graphical depiction of a correlation analysis of predicted and measured volatile acidity.

DETAILED DESCRIPTION

Referring to Fig. 1, a system for monitoring a plurality of barrels 100 containing a beverage is shown generally at 102. The system 102 includes a controller 104 that is operably configured to receive sample data 106 for each barrel 108 of the plurality of barrels 100. The barrel 108 includes a barrel identifier 110 that uniquely identifies the barrel 108. The barrel identifier 110 may be a barcode, Quick Response (QR) code, a radio frequency id (RFID), a serial number, or other identifier. The sample data 106 includes at least the barrel identification 110 identifying the barrel 108 and also includes a free sulfur dioxide concentration (herein abbreviated as FSO2) value indicative of a concentration of free sulfur dioxide within a beverage contained in the barrel 108. The barrel identification 110 may include an identifier of the barrel 108. For example, the barrel identification 110 may include a barrel number. Alternatively, the barrel identification 110 may comprise another identifier that may be used to identify the barrel 108. For example, the barrel identification 110 may comprise a network address of a device (e.g. a sensor) that monitors that specific barrel. Since the device monitors the specific barrel, the network address of the device may be used to identify the barrel 108.

The system also includes a host processor 112, which is in communication with the controller 104 for receiving an upload of sample data for the barrel from the controller. The communication with the controller may be via a wired or wireless connection or may be established through a common network such as the internet connecting between the controller and host processor.

The host processor 112 also includes a data store for accumulating the sample data for the plurality of barrels 100 to generate a barrel dataset 114. The host processor 112 further includes a chemistry data analyzer 116 that performs an algorithmic analysis on the barrel dataset 114 to generate a condition output 118. In one embodiment the chemistry data analyzer 116 performs mathematical and or statistical operations on the data in accordance with a chemistry model to generate the condition output 118. The condition output 118 is indicative of a condition of the beverage and/or any of the plurality of barrels 100 in which the beverage is being contained. The condition output 118 is used as an input to a management function 120 to be performed on the beverage, specific barrels in the plurality of barrels 100, or the plurality of barrels. The management function 120 may involve any of a variety of tasks, processes, or generation of reports related to the beverage or any of the plurality of barrels 100 in which the beverage is contained.

In one embodiment where the beverage is a fermented beverage such as wine, the barrel 108 may be a wooden barrel commonly used in the winemaking industry. Oak barrels are commonly used in wineries for both fermentation and aging of wine.

Referring to Fig. 2A, in one embodiment the controller 104 may be implemented using a networked device 200. A block diagram of the networked device 200 is shown in Fig. 2B. The networked device 200 includes a microprocessor 202, a display 204, and an input device 206 for receiving user input. In some embodiments the input device 206 may be provided as touch screen on the display 204. The networked device 200 also includes a memory 210, which may be implemented using random access memory, non-volatile flash memory, a hard drive or combination of these and other memory types. The memory includes a location 212 for storing program codes that direct the microprocessor 202 to implement operating system functions. In one embodiment the operating system may be a smartphone operating system such as an Android™ based operating system, an iOS based operating system, or any other operating system. The memory 210 also includes a location 214 for storing barrel scanning application program codes, a location 216 for storing measurement hardware control program codes, and a location 218 for storing sensory evaluation application program codes. The memory 210 further includes a location 220 for storing sample data associated with the barrel scanning application and a location 222 for storing sensory data associated with the sensory evaluation application.

The processor circuit of the networked device 200 further includes a RF baseband radio 230 and antenna 232 for connecting to a mobile telecommunications network. The RF baseband radio 230 may be configured to provide data communications using any of a variety of communications standards including 2G, 3G, 4G, and/or 5G or any other communications standards. The networked device 200 also includes a wireless radio 234 and antenna 236 for connecting to local networks such as an IEEE 804.11 Wi-Fi local network. The wireless radio 234 may also provide for connections via other wireless links or protocols, such as Bluetooth, Wi-Fi Direct, or near-field communication. The processor circuit of the networked device 200 also includes a video/image processor 238 and a camera 240. The video/image processor 238 receives and processes image and/or video signals from the camera 240. The display 204, input device 206, memory 210, RF baseband radio 230, wireless radio 234, and video/image processor 238 are all in communication with the microprocessor 202.

Referring back to Fig. 2A, in this embodiment the barrel identifier 110 may be read by using the camera 240 to capture an image of the barrel 108. In this embodiment the networked device 200 is shown as a smartphone operated by a user 250. The RF baseband radio 230 and wireless radio 234 of the networked device 200 provide several options for uploading sample data. The networked device 200 may use the wireless radio 234 to connect to a local wireless network access point 252 for uploading either directly or via a wide area network 254 (such as the internet) to the host processor 112. Alternatively, the RF baseband radio 230 may be used to connect via a cellular data network to the wide area network 254 for uploading to the host processor 112.

In one embodiment FSO2 values may be generated by a FSO2 sensor 256, which is connected via a sample line 258 and a dip tube 260 to draw a sample from the barrel 108. In this embodiment, when drawing the sample from the barrel 108, a bung 262 that seals an opening 264 is removed, which presents a minor contamination risk while sampling. In other embodiments the sample may be drawn through a closure such as described in commonly owned PCT patent publication WO/2020/227822 entitled "System, Closure, and Interconnect for Managing a Beverage in a Bulk Liquid Container", which significantly reduces the likelihood of contamination while performing the sampling. PCT patent publication WO/2020/227822 is hereby incorporated by reference in its entirety. In some embodiments the FSO2 sensor 256 may be configured to receive wired or wireless communications from the networked device 200 for initiating the sampling process and for communicating the FSO2 values back to the networked device 200. The measurement hardware control program codes 216 in the memory 210 of the networked device 200 when executed by the microprocessor 202 may configure the device for communications with and/or control of the FSO2 sensor 256. In other embodiments the controller 104 may be embedded within the FSO2 sensor 256 and may communicate directly with the host processor 112 rather than via a networked device 200. In some embodiments, each barrel 108 may be provided with a respective FSO2 sensor 256. A respective controller 104 may be embedded within each FSO2 sensor 256, for example. In one embodiment the host processor 112 may be implemented using the processor circuit shown in Fig. 3. The processor circuit of the host processor 112 includes a microprocessor 300, a program memory 302, a variable memory 304, and an input output port (I/O) 306, all of which are in communication with the microprocessor 300. Program codes for directing the microprocessor 300 to carry out various functions are stored in the program memory 302, which may be implemented as a random access memory (RAM), and/or a solid state or hard disk drive, or a combination thereof. The code may be written in any suitable program language, such as JavaScript, Python, Java, C, C++, C#, and/or assembly code, for example.

The program memory 302 includes a block of program codes 320 for directing the microprocessor 300 to perform operating system functions. The program memory 302 also includes a block of codes 322 for directing the microprocessor 300 to provide functions for interfacing with the controller 104 to receive the sample data. In this embodiment the program memory 302 also includes a block of codes 324 for directing the microprocessor 300 to provide database functions. The program memory 302 also includes a block of codes 326 that directs the microprocessor 300 to perform the chemistry data analysis on the barrel dataset 114.

In the embodiment shown in Figure 2, the processor circuit of the controller 104 further includes a database mass storage unit 340 providing data storage for storing sample data. The I/O 306 also includes a storage interface 362 for interfacing with the database 340. In one embodiment the database mass storage unit 340 may be implemented using one or more hard drive units, solid state drives, or other persistent storage medium. In some embodiments the database mass storage unit 340 may provide storage for implementing a locally hosted database for accumulating and storing the sample data in an organized format to provide the barrel dataset 114. In other embodiments the I/O 306 may further include a communications interface 360 for connecting to the wide area network 262 and the barrel dataset 114 may be stored in a networked database 362 accessible via the wide area network.

Although the processor circuit of the host processor 112 is shown in Figure 3 as having conventional computer architecture, the processor may be implemented using shared configurable computer system resources such as those that may be provided by companies such as Microsoft, Google, or Amazon and other cloud computing resource providers. As such, the processor circuit of the host processor 112 shown may represent a virtual machine, possibly implemented using multiple processors and other resources to provide the necessary functionality. One advantage of using a shared computing resource is that the resource becomes dynamically scalable and additional processing power or storage may be allocated as required. As such, the microprocessor 300, program memory 302, variable memory 304, and I/O 306, may be parts of a virtual machine hosted on a shared and/or distributed computing resource. In other embodiments the host processor upload functions may be provided by a local networked host processor circuit, while database functions are provided by the remotely located networked database 362.

Referring to Figure 4A, a flowchart depicting blocks of code for directing the processor circuit of the networked device 200 (shown in Fig. 2B) to perform a barrel sampling process is shown generally at 400. The blocks generally represent codes that may be read from the location 214 of the memory 210 for directing the microprocessor 202 to perform the barrel sampling process. The actual code to implement each block may be written in any suitable program language, such as C, C++, C#, Java, and/or assembly code, for example.

The barrel sampling process 400 begins at block 402, when the user 250 initiates execution of the barrel scanning application codes stored in the location 214 of the memory 210. The application causes a series of screens to be displayed on the display 204 of the networked device 200, examples of which are shown as screenshots in Fig. 4B. Block 404 then directs the microprocessor 202 to read the barrel identifier 110. Referring to Fig. 4B, for reading the barrel identifier 110 in this embodiment the microprocessor 202 causes the networked device 200 to be configured to display an image captured by the camera 240 as an initial screen 430. The user 250 is guided by a set of displayed indicia 432 to move center the barrel identifier in the screen to facilitate capture of the barrel identifier. When the barrel identifier 110 has been captured, block 404 directs the microprocessor 202 to access functions for decoding the QR code and to display a second screen 434, which shows the decoded barrel identifier "BW5591". The barrel identifier BW5591 is also stored in the sample data location 220 of the memory 210.

The barrel sampling process 400 then continues at block 406, which directs the microprocessor 202 to determine whether the user has initiated the barrel sampling process by pressing a "measure" button 436 (Fig. 4B) on the screen 434. Generally, the user 250 will first scan the barrel and then sanitize the bung 262 and opening 264, before inserting the dip tube 260. Once completed the user 250 may initiate the sampling process by pressing the "measure" button 436 on the screen 434. If at block 406, the sample process has been initiated the microprocessor 202 is direct to block 408. Block 408 directs the microprocessor 202 to cause the wireless radio 234, or other output communication interface, to transmit a signal to the FSO2 sensor 256 to cause a beverage sample to be withdrawn from the barrel 108 and delivered via the dip tube 260 and sample line 258 to the FSO2 sensor 256. In one embodiment the FSO2 sensor my be implemented using the apparatus disclosed in commonly owned US provisional patent 63/270706 filed on 22 Oct 2021 and entitled "Apparatus and System for Measuring SO2 Concentration in a Beverage", which is incorporated herein by reference in its entirety. The disclosed apparatus automates the drawing of the sample and the measurement of FSO2 values. In other embodiments the sample may be drawn manually and measured using an analyzer relying on one of many measurement technologies available or by means of other analytical chemistry techniques such as optical absorption, titration, and electrochemical analysis. As an example, the Sentia™ Analyzer available from Universal Biosensors of Australia measures free SO ₂ using a test strip. In some embodiments, the sample might not comprise a sample of the beverage. For example, the sample may comprise a gas sample. There may be gas above the beverage, also referred to as the head gas, which may be at an equilibrium with the SO2 concentration of the beverage. A sample of this gas may be indicative of the FSO2 of the beverage. Therefore, in some embodiments, a gas sample may be drawn. The gas sample may be delivered to the FSO2 sensor for measurement in order to determine the FSO2 value of the beverage. In general, a fluid sample may be drawn and measured using an analyzer (e.g. the FSO2 sensor), in which the fluid may be a gas or a liquid (e.g. a beverage).

The process 400 continues at block 410, when the FSO2 sensor 256 completes the FSO2 measurement and transmits the FSO2 value to the networked device 200. Block 410 directs the microprocessor the 202 to receive the FSO2 value and to store the value in the sample data location 220 of the memory 210. In this embodiment the result is displayed in an updated version 434' of the screen 434 with the measured FSO2 value for the barrel BW5591displayed in parts per million (ppm) at 438. In this embodiment a date and time of the FSO2 measurement is also associated with the barrel identifier 110 and stored in the sample data location 220 of the memory 210.

Block 412 then directs the microprocessor 202 to determine whether the user has initiated sampling for a further barrel. If the user presses a button 440 on the screen 434', the microprocessor 202 is directed back to block 404 and blocks 404 - 410 of the process 400 are repeated for the next barrel. If at block 412, the user has not initiated sampling for another barrel, the process continues at block 414. Block 414 directs the microprocessor 202 to determine where the user has pressed an upload button 442 displayed on the screen 434'. If at block 414 no user selection at either block 412 or 414 has been received, then block 414 directs the microprocessor 202 back to block 412 to continue to await user input. In other embodiments block 414 may be omitted in favor of an automated upload by the controller 200 that directs the microprocessor 202 to block 416 without further intervention by the user. For example, if the networked device 200 detects a sufficiently powerful radio signal of a wireless network, the upload event may be initiated automatically over the network at block 416.

For the screenshot examples shown in Fig. 4B, the user is assumed to have pressed the button 440 and scanned and measured the FSO2 value for another barrel resulting in an updated screen 434" being displayed along with a second FSO2 value measurement for a barrel "BW5555". In this example, if at block 414, the upload button 442 is pressed the microprocessor 202 is directed to block 416. Block 416 directs the microprocessor 202 to initiate an upload of the sample data values for the barrel BW5591 and BW 5555. On this embodiment the samples data for the barrels is transmitted wirelessly. However, in other embodiments the sample data may remain stored in the sample data location 220 of the memory 210 until the user has completed scanning for all or part of a plurality of barrels 100 in a specific location. Subsequently the user may make a wired connection to the host processor for uploading the stored sample data. The barrel sampling process 400 thus permits the user to gather data samples for a plurality of barrels 100 before uploading the sample data to the host processor.

Referring to Fig. 4A, a flowchart depicting blocks of code for directing the processor circuit of the host processor 112 to generate the barrel dataset 114 is shown generally at 450. The process 450 is initiated in response to the sample data upload at block 414 from the networked device 200. Block 452 directs the microprocessor 300 of host processor 112 to determine whether upload has been initiated by a networked device 200. If an upload is initiated, block 452 directs the microprocessor 300 to store the uploaded sample data in the variable memory 304, which acts as a temporary storage location while the uploaded sample data is being processed.

In this embodiment the uploaded sample data is accumulated by separating and writing the sample data for each scanned barrel to the database 340 that provides organized storage for the barrel dataset 114. The database function block of codes 324 in program memory 302 includes codes executable by the microprocessor 300 for accessing, reading from, writing to, and performing queries on the database 340. Referring to Fig. 4C, an example of a set of relational database tables that may be configured in the database 340 is shown generally at 470. The relational database structure 470 is well suited for the barrel management system since it is focused on the barrel as a physical asset rather than the beverage currently contained in the barrel. The database structure embodiment shown in Fig. 4C facilitates tracking the barrel over time as it is used to contain multiple different beverages. In one embodiment, when new barrels are added to the barrel inventory at a winery, the barrels may be assigned a barrel identifier and the barrel identifier 110 applied to the barrel. At this time, the barrel may also be scanned as described above for the process 400, and a record created in the database 340. At the same time, or at some later time, various barrel properties such as the oak type (French, American, or other), the cooper that made the barrel, the year of barrel cooperage, and other properties such as toast level, seasoning etc. may be entered into a record stored in the database 340. In describing the remainder of the process 450, it will be assumed that any barrel that is being used to contain a beverage will already have a record created in the database 340 that is associated with a barrel identifier 110. The relational database structure 470 includes a table 472 that is used as a template for storing the barrel properties. The barrel identifier 110 (Barrel_ID in the table) is designated as the primary key for this table. The database 340 also stores a "Measurement Events" table 474 for storing sample data, a "Sensory Events" table 476 for storing sensory evaluation information, a "Process Events" table 478 for storing data associated with beverage processing such as sulfite additions, and a "Barrel Fills" table 480 for storing data associated with a batch of beverage currently contained in the barrel. The tables 474 to 480 are all arranged under the Barrel_ID as a primary key. Finally, a "Batches" table 482 is used to store details of the beverage that makes up a batch. In the case of a wine, properties such as the acidity (pH), varietals, etc. would be of relevance. The table 482 has a primary key of "Batch_ID".

The process 450 continues at block 454, which directs the microprocessor 300 to read the first barrel identifier for the first barrel in the uploaded sample data stored in the variable memory 304. Block 446 then directs the microprocessor 300 to write a new record to the "Measurement Events" table 474. For the example of the barrel 108, the Barrel _ID would be written as "BW5591", and the FSO2 concentration as "13.36". The date and time of the measurement would be written to the Datetime field in the record. As shown in Fig. 4C, additional sample data such as a pH value indicative of the acidity of the beverage, a volatile acidity indicative of an amount of acetic acid in the beverage, a temperature value indicative of the temperature of the beverage, a color value indicative of a spectral reflectance, transmittance, or absorbance of the beverage, and/or a turbidity value indicative of the clarity of the beverage may be written to the new record. The additional sample data may additionally or alternatively include a total SO2 concentration (e.g. including both free SO2 and bound SO2), an ethanol concentration, a 4-ethylphenol (4-EP) concentration and/or a 4-ethylguaiacol (4-EG) concentration. These values may be obtained via various measurement methods in a similar manner to the FSO2 measurement described above. Block 458 then directs the microprocessor 300 to determine whether further barrels remain to be processed in the uploaded sample data. If at block 458 a further barrel (in this case the barrel BW5555) remains to be processed, the microprocessor 300 is directed to block 464 and then block 456 and 458 are repeated. If at block 458 no further barrels remain to be processed in the uploaded sample data, the process 450 ends at block 460.

Records may be created for other tables in the relational database structure 470 in a similar manner to that described above for the "Measurement Events" table 474. For example, the "Process Events" table 480 is used to store information related to barrel processing events such as sulfite additions. Sulfites may be added to a wine beverage contained in the plurality of barrels 100 at various times through the fermentation and aging process. The quantity of sulfite is usually carefully measured and recorded. Other steps such as whether the lees were stirred when the sulphite addition was made may also be recorded. The above referenced PCT patent publication WO/2020/227822 discloses automated sulfite dosing apparatus and methods that are capable of delivering a measured quantity of sulfite to a barrel and recording the quantity and time of the dosing. For any barrel in a plurality of barrels 100, the "Process Events" table 480 thus facilitates the creation and storing of a record for each process event. Other process events such as topping (filling the barrel to reduce the headspace) may be similarly recorded in the database 340.

The "Barrel Fills" table 482 provides storage for details of the beverage actually contained in a barrel. When initially filling each barrel in a plurality of barrels 100, the filling date and time may be recorded along with a Batch_ID of the fill beverage. The "Barrel Fills" table 482 need only identify the beverage contents in the barrel via the Batch_ID, with the details of the fill batch being provided by records created under the "Batches" table 484.

The database 340 when organized in accordance with the relational database structure 470 shown in Fig. 4C facilitates the extraction of specific data from the barrel dataset 114 by a database query. As an example, FSO2 values may be extracted along with their associated Datetime values for some or all of the barrels that store a particular beverage batch over a time period. Over the same time period, additions of sulfite to the barrels may be extracted from the barrel dataset 114, together these data provide a time map of the level of sulfites in each of the barrels. The barrel management function 120 may then be performed based on this information. For example, barrels that appear to be consuming sulphite at a higher rate may be flagged for removal from the winery barrel inventory or may be flagged for a use in which sulfite concentration is less critical.

The "Sensory Events" table 478 is included to provide organized storage for recording results of user tasting and other sensory experiences related to the beverage contained in a barrel. Referring to Fig. 5A, in one embodiment the sensory information is captured from one or more human beverage tasters via the sensory evaluation application implemented by running the codes in location 218 of the memory 210 on a networked device 200 such as shown in Fig. 2A and 2B.

Referring to Figure 5A, a flowchart depicting blocks of code for directing the processor circuit of the networked device 200 (shown in Fig. 2B) to perform a sensory evaluation process is shown generally at 500. The blocks generally represent codes that may be read from the location 218 of the memory 210 for directing the microprocessor 202 to perform the sensory evaluation process. The process 500 begins at block 502, which directs the microprocessor 202 to run the program codes in location 218 of memory 210 to launch the sensory evaluation application on the networked device 200. A screenshot of an initial screen displayed is shown in Fig. 5B at 520. The initial screen 520 is configured to display an image captured by the camera 240 for capturing the barrel identifier 110 of the barrel 108, as described above in connection with Fig. 4B. Block 504 then directs the microprocessor 202 to read the barrel identifier 110 and to decode the QR code to determine the Barrel_ID.

Block 506 then directs the microprocessor 202 to display an input screen, an example of which is shown at 522 in Fig. 5B. The screen 522 displays the Barrel_ID "BW5591" in a field 524 and the date and time in a field 526. The remaining input fields on the screen are included to guide the taster through the sensory evaluation in an organized manner and to receive inputs associated with the beverage from the barrel 108 that is being evaluated. The screen 522 is designed to be scrolled through additional input fields associated with the evaluation. The taster's sensory evaluation inputs at the various fields are written to the sensory data location 222 of memory 210. Block 508 directs the microprocessor 202 to determine whether the taster has initiated an upload of the sensory data to the host processor 112. If at block 508, the upload has been initiated the microprocessor 202 is directed to block 510. Block 510 directs the microprocessor 202 to upload the data via the RF baseband radio 230 or wireless radio 234 to the host processor 112. If at block 508, the upload has not yet been initiated the microprocessor 202 is directed back to block 506 to continue receiving taster inputs. When the host processor 112 receives a sensory data upload, the microprocessor 300 writes the sensory data to a new "Sensory Events" record in the database 340, which is referenced by the Barrel _ID primary key. Sensory data uploads from a plurality of tasters may this be received and stored as part of the barrel dataset 114 in the database 340. The process 500 for receiving sensory evaluation data and the storing in the database 340 provides for a systematic and structured collection of sensory evaluation data.

Referring back to Fig. 3, the host processor 112 is configured to implement the chemistry data analyzer 116 by causing the processor circuit to execute the block of codes 326 to perform the chemistry data analysis on the barrel dataset 114. The codes 326 configure the chemistry data analyzer 116 to provide functions for generating the condition output based on the barrel dataset 114. The microprocessor 300 of the host processor 112 executes the block of codes 324 to provide database functions for extracting data from the barrel dataset 114 stored in the database 340. The extracted data provides an input to the chemistry data analyzer 116. The barrel dataset 114 includes data related to chemical measurements, physical properties, process events, barrel properties, barrel history, and sensory evaluation data. The data in its raw unprocessed form would be unlikely to reveal any human recognisable patterns that would provide an insight related to the condition of the barrels or the beverage contained therein. The chemistry data analyzer 116 is configured to processes the data to extract information that has utility for performing various management functions 120 on the barrels and/or beverage contained in the barrels.

One embodiment of a chemical data analyzer 116 that may be configured by the block of codes 326 to perform a chemistry data analysis is described below in the context of sulphite concentrations in a wine beverage contained in the plurality of barrels 100. SO2 has biocidal and antioxidant properties when it is in a free form, that is not chemically bound to other compounds. Sulfites exist in wine in several equilibrium states. Free SO2 refers collectively to SC -F O (molecular SO2), HSOa” (the bisulfite ion), or SOa ^2- (the sulfite ion). The following chemical equilibrium formulas may be written to describe reactions between the various forms of free SO2: Eqn l Eqn 2 Research has demonstrated that it is the molecular form that sulfites interrupt microbial processes and impact oxidation. Molecular SO2 interrupts microbial processes by aggressively binding to other molecules and enzymes and rendering them inactive. Once the molecular SO2 is bound to a molecule or enzyme, it is no longer available for further binding and is therefore no longer "free". As such free SO2 is consumed as it provides protection to the wine and there is thus a coupled relationship between the free SO2 concentration history of a wine and the state of dissolved oxygen and microbial activity. There is a known relationship between free SO2 and microbial activity, which is the reason that sulfites have been used in winemaking for thousands of years.

To model the consumption of free SO2 in the deactivation of microbial activity, a simple reaction mechanism is proposed represented in the following chemical formula: Eqn 3 where A is a population of active microbes, FSO2 is free sulfur dioxide, BSO2 is bound sulfur dioxide, and D is deactivated microbes. The coefficients/, a, b, and d represent the stoichiometric ratios of each reagent and product respectively.

Using mass action kinetics, an expression may be written for the change of FSO2 concentration over time: Eqn 4

Equation 4 relates the change of concentration of free SO2 in time to the concentrations of free and bound SO2 and the active and deactivated populations of microbes. Although the reaction presented as Equation 3 was written as an equilibrium, for the temperatures of concern when cellar ageing wine of (typically 1°C < T < 25°C), it may be assumed that the reverse reaction of bound SO2 back to free SO2 has a negligible rate (/.e., kpso2 » k _r ). Equation 4 may thus be simplified as follows: Eqn 5

Since direct measurement of the population of a given microbe requires specialized microbiological techniques, it may be more convenient to instead monitor the metabolite of an active microbe species of interest as being indicative of its reproduction rate. In the case of acetobacter (an acetic acid bacteria) the metabolites of interest would be volatile acids such as acetic acid. In the case of brettanomyces (a non-spore forming yeast) the metabolites of interest may be 4-Ethylphenol (4-EP) and 4-Ethylguaiacol (4-EG). In winemaking, spoilage by acetobacter is very common and inhibited growth of acetobacter is generally considered a good indicator of the overall health and oxidation state of an ageing wine.

The following simple chemical reaction formula may be written for the metabolite of volatile acidity (VA) produced by active acetobacter:

A -> VA + A, Eqn 5 where the active acetobacter population A produces the metabolite VA in addition to more copies of A. This allows the following relation to be written between the active population of acetobacter and the change of concentration of metabolite: where k _Vfl is a proportionality constant that relates the production rate of metabolite for a microbial population. Combining Equation 6 with Equation 4, the relation between the metabolite history and free SO2 history may be written as follows:

If it is assumed that first order kinetics apply for both species (i.e. f = 1 and a = 1) then a simplified and rearranged equation may be written: yielding a differential equation relating microbe metabolites and free SO ₂ time history. The coefficients k _V A and k _Fso2 typically have some dependence on temperature, which can be modelled in the form: where f(T) may be of the Arrhenius form for the temperature dependence of reaction rates or another more suitable relationship for biochemical processes in the case of k _VA .

Equation 8 may be rewritten to yield [VA](t) as follows: Equation 11 thus provides a mathematical representation of a physical model for the concentration of metabolite VA in terms of an integrated FSO2 history up to the present time t.

Referring to Fig. 6A a graphical depiction of an example of a free SO2 concentration timeseries [FS02](t) for a subject barrel in the plurality of barrels 100 is shown at 600. Concentration values C are plotted on the y- axis at various times. Free SO2 concentration C in the subject barrel will vary over time as free SO2 is bound to molecules and enzymes and thus consumed. In a typical winery workflow, a free SO2 measurement may be performed on a barrel followed by a possible addition of sulfites, typically in the form of an aqueous solution or powder. The free SO2 concentration measurement for the current time t, is denoted C,. In this example, a sulfite addition Si-i had been made at a time t/-i due to the concentration CM being at a low level at the previous time t _w . A graphical depiction 602 of a sulfite addition SM is shown represented by a bar 604. The C and S values are collected as discrete measurement values at specific times t.

Referring to Fig. 6B, a table 610 is shown that includes columns 612 representing data values of C, S,and t, 612 for the subject barrel that may be extracted from the barrel dataset 114. Records in the database 340 based on the "Measurement Events" table 474 provides the values of C, while records of based on the "Process Events" table 480 provides the sulfite addition data S. The final column 614 ([VA]) is the estimated metabolite concentration at each timestep and may be calculated based on Equation 11, which for the data in table 610 would need to be converted into a form that facilitates the use of the discrete data values 612.

Referring to Fig. 6C, in this embodiment the continuous function in Equation 11 may be discretized as shown in the table 620 where V, is the volatile acidity VA at time t/. Substituting discretized forms of the expressions from the table 620 into Equation 11 and rearranging gives the following:

Eqn 12

Solving Equation 12 for the subject barrel at each measurement time step t, provides an estimate of the integrated microbial forcing level. In this case, the forcing level is represented in terms of the acetic acid metabolite of acetobacter, but a mathematically equivalent solution could be formulated for other spoilage metrics by changing the coefficients k _Vfl and k _FS02 . In the formulation provided as Equation 12, two rate coefficients, k _V A and k _F so2 are used to capture the kinetics of metabolite formation and FSO2 consumption respectively. In the first order kinetics formulation in accordance with one embodiment, the rate coefficients may be condensed into a single coefficient k _AF representing the quotient of the two coefficients. The combined coefficient k _AF is only applicable for this formulation and captures the combined relationship of acetobacter and free SO2. Several simplifying assumptions can be made to estimate the value of k _AF : k _AF */ T i.e. the value of the constant does not change with temperature; and k _AF is constant for all barrels in the plurality of barrels 100 that are filled with the same wine.

An optimization algorithm may be applied that attempts to minimize the error between predicted VA concentration, Vi, _s and experimentally measured VA concentrations V _e (ti). In one embodiment, a Monte Carlo simulation may be employed to locate an error minima and an approximate value for k _AF , which is used for the examples presented below.

Example

The model embodied by Equation 12 was applied using a dataset of free SO2 measurements and records of sulfite additions for a winery in British Columbia, Canada. All of the barrels were sampled for initial volatile acidity (VA) concentration at the beginning of the trial, and a subset were sampled for final VA concentrations to compare to the modelled VA concentration. Results for two example barrels Barrel 1 and Barrel 2 are shown in Fig. 7 at 700 and 702 respectively. For Barrel 1 the free SO2 concentration is represented on the upper graph at 704 and was recorded approximately once per month through a 6 month measurement period. The sulfite addition history over the same time period is represented by the bars 706. The free SO2 concentration 704 and the sulphite additions 706 provide the barrel dataset for the analysis.

The estimated VA concentration 708 was calculated using Equation 12 for each time point and is shown in the lower pane of the graph 700. The VA trajectory is an example of a barrel that is at average or reduced risk of aerobic bacteria spoilage, as both the modelled and measured VA concentration suggest a relatively low population of acetobacter. Measured VA value 710 (the "+" sign) shows good agreement with the predicted VA value 708 for October.

Conversely, Barrel 2 shown in the graph 702 has a VA concentration trajectory 712 that reveals an elevated risk of spoilage as the VA concentration predicted using Equation 12 suggests a relatively strong population of acetobacter. The measured VA concentration 714 confirms the predicted concentration 714 for October and also the relatively strong population of acetobacter. For both Barrel 1 and Barrel 2 the measured VA concentration shows good agreement with the VA concentration predicted using Equation 12.

As disclosed above, Equation 8 was derived based on an assumption of first order reaction kinetics. In another embodiment, where the governing elementary reactions for this phenomena are not known, the reaction order for each species that yields the best agreement with experimental results may be chosen. As an example, for the case of second order (a = 2) with respect to VA and first order with respect to SO ₂ (f = 1), Equation 7 may be written as:

Eqn 13

For the case with second order reactions with respect to both VA and FSO214 the expression can instead be written as:

Eqn 14 where the term in [FSO ₂ ] in the denominator is raised to the power of 2. For the limited dataset considered herein, it would appear that the second order kinetics with respect to VA provide a more physically relevant fit to experimental data, although the exact fitting and optimization of this model could be further improved with some potential strategies discussed below. Referring to Fig. 8, a correlation analysis of predicted and measured VA for the first order kinetics formulation (Equation 8) and second order kinetics with respect to VA (Equation 14) is shown at 800. Even in the simplified form, a meaningful correlation between FSO ₂ history and microbial activity is revealed.

For the embodiments described above, simplifying assumptions have been made to aid in providing a tractable solution without relying on extensive supporting inputs. Even in this simplified form, Figure 8 reveals a meaningful correlation between FSO ₂ history and microbial activity. Additional information can be injected into the mass-action kinetics framework to increase the fidelity of the model, therefore increasing the confidence and utility of the outputs.

Through collection of larger datasets, the coefficients k _V A and k _F so2 in Equation 12 may be parameterized to functions of other variables that are known to impact the microbial growth, dissolved oxygen levels, and free SO2 behavior in wine barrels. One example that was mentioned previously would be the incorporation of temperature data to capture the change of chemical and microbial kinetics with changing temperatures in the wine matrix. Measurement of both cellar interior air temperature and sample wine temperature could provide useful information that would increase model fidelity. Additionally, the relative humidity of the cellar environment is known to affect the evaporation rate of wine through oak vessels, which would also likely have an impact on dissolved oxygen levels.

Other examples of additional information that can be incorporated into the parameterized coefficients are the wine varietal, other wine chemistry details measured at a tank level (e.g. residual sugar, malic acid, alcohol level, total acidity, pH) and barrel data such as cooper, year of cooperage, number of fills, toast, seasoning, and oak type.

The chemistry data analyzer 116 embodied by Equation 12 thus identifies barrels in the plurality of barrels 100 that have outlying metabolite concentrations and may be configured to produce the condition output 118 to identify the suspect barrels to the winemaker. The winemaker may thus perform barrel management functions such as drawing a sample from the identified barrels for further laboratory analysis. The laboratory analysis results may be recorded and fed back into the model to make improvements for future use at the individual winery as well as across other facilities.

The management function 120 shown in Fig. 1 may involve any of a variety of management functions related to the plurality of barrels 100, individual barrels within the 100, or the beverage in the barrels, several examples of which follow below.

The coupled relationship between the free SO2 concentration history of a wine and microbial activity allows the system 102 to generate a condition output 118 that identifies the barrels in the plurality of barrels 100 that are experiencing increased microbial activity. The winemaker is thus provided with information that allows microbial issues to be immediately addressed. For example, the winemaker may choose to add a larger than usual dose of added SO2 to kill off unwanted microbes before the metabolites produced degrade the quality of the wine. Further, by identifying individual barrels with microbial problems, winemakers may exclude these barrels from routine homogenization events wherein all barrels in a batch are mixed. This step may limit the spread of undesirable microbes and reduce the chances of entire batches being degraded or spoiled.

When discrete barrel-by-barrel SO2 data is collected over multiple vintages, patterns emerge. Certain barrels that have a higher frequency of microbial ‘flags’ may be characterized as performing poorly. Once identified, poorly performing barrels can then be retired or reassigned to lower value products.

One of the crucial functions of an oak barrel is to micro-oxygenate the aging wine. This process aids the polymerization of tannins into larger molecules which then are precipitated out of the solution, thus reducing astringency on the wine drinker's palate and improving the quality of the wine. However, if the wine is exposed to too much oxygen, oxidative faults and browning may develop. On the other hand if too little oxygen is introduced, reductive faults may develop. Winemakers would thus be advantaged by managing oxygen exposure precisely in order to ensure that wine quality is upheld. It is thus of great value to a winemaker to be made aware of each aging barrel's oxygen transfer rate (OTR). There is generally significant variability between barrel OTR within a plurality of barrels 100. The OTR can also change over time as the barrel itself ages, accumulates tartaric crystallization, and otherwise is affected by its use. Conventional winemaking methods are limited in their ability to gather and interpret OTR information on individual vessels and barrel-by-barrel OTR is not typically available for winemakers.

The system 102 allows for the approximation of each barrel's OTR through the measurement of free SO2. Since free SO2 scavenges oxygen, and is consumed in doing so, there is a coupled relationship between the free SO2 concentration history of a wine and the state of dissolved oxygen. Over time, collecting and analyzing this data systematically can be used to infer the OTR of individual oak barrels and present this information to winemakers in an efficient manner.

One of the most important tasks performed by winemakers is combining aging vessels into finished 'blends. Before bottling and selling the wine, winemakers will evaluate each vessel on a host of sensory characteristics and mix combinations based on their professional judgement. Each blend becomes its own wine label, or SKU. The system 102 allows for the structured collection of sensory data and relates it to individual barrels. With this data easily manipulated using the host processor 112 to access and query the database 340, the winemakers can sort their data on a variety of parameters, simulate the characteristics of finished blends, and project volumes of wine available for each finished blend.

Once the blending process is completed the winemaker can use the system to perform sensory analysis on finished blend. With enough data collected on both the inputs (individual barrels used as blending ingredients) and the finished blends, the system will be able to more accurately predict the sensory outcomes before the blending occurs. This will allow winemakers a greater ability to experiment with different possible combinations of their available blending ingredients, and ultimately develop the optimal portfolio of blends to support their commercial goal. Wine sales and marketing data, and wine consumer preference data on can be incorporated into the system in order to suggest blending combinations that have a higher potential to be well received by the market or command a higher price. The free SO2 data collected is similarly related to each individual barrel. With a sufficient quantity of both sensory and free SO2 data, the system allows for the ability to use free SO2 data gathered automatically during the aging process to predict various sensory outcome. Thus, the winemaker can be informed early in the process which barrels of wine will be of high quality and which of low quality once aging is complete. Factoring in metadata relating to each barrel (oak species, toast level, manufacturer, age, etc.) as well as data on the cellar processes and environmental conditions allows the system to make suggestions to increase the likelihood of achieving specific sensory, or wine quality outcomes.

The management function 120 may also involve aspects of facility and asset management related to the plurality of barrels 100. One application of the system described above would be to suggest factors (barrel purchases, processes, cellar environmental conditions, etc.) to the winemaker that tend to reduce the aging time required for the wine to reach maturity. Reducing the aging time would allow wineries to bring their products to market faster, resulting in a competitive advantage.

Important factors for a winery purchasing a used barrel includes the barrel's history of OTR, sensory outcomes, and microbial incident. Regarding the latter, it is possible for barrels to harbor harmful microbes so that they infect multiple vintages of win. Having this data rigorously tracked could help winemakers receive a premium price for used barrels that have a history of performing well in these regard. It could also help purchasers avoid contaminated barrels or barrels with a history of poor performance. These data would help sellers of used barrels to command better pricing compared to their competitors as they can offer a product that carries less risk.

Another application is to assist a winery with optimizing their oak purchasing decision. As oak barrels rank highly on a winery's list of expenses, any opportunity to increase the winemaker's ability to purchase the optimal barrels would be highly valued by the industry. Various oak types, ages, toast levels and cooperage (manufacturing) will impact free SO2 consumption rates as well as sensory outcomes and wine quality. The system described can recommend wineries to purchase the barrels that are best suited to their commercial goals.

Additionally, the management function 120 of the system 102 may recommend, based on both an individual winery's data and aggregate data from multiple wineries, which barrels should have their useful life extended. For example, if a barrel was set to be retired after five years, the system could suggest that it be used for seven, which would have an immediate economic benefit for the winery. Alternatively, if a barrel is performing poorly or has a high rate of microbial flags it should be retired earlier than its projected useful life, resulting in possible quality degradations being avoided.

In the above disclosure the chemical analysis has been described based on a knowledge of the physical processes at play, which facilitates the generation of a useful chemical analysis model without a need for large quantities of data. Alternatively, statistical models may be implemented in conjunction with, or in place of, the physical models, where there is a sufficiently large dataset. In an embodiment where the winemaker has a large inventory of barrels, over time a sufficiently large set of sample data may be accumulated to implement the chemistry data analyzer 116 using machine learning. For example, the chemistry data analyzer 116 may be implemented using a neural network that may be trained using the large set of sample data that may be annotated based on actual measurement results and observations.

While specific embodiments have been described and illustrated, such embodiments should be considered illustrative only and not as limiting the disclosed embodiments as construed in accordance with the accompanying claims.