Background management system and method of intelligent coffee station

Abstract

The invention provides a background management system and method for an intelligent coffee station, and the method comprises the steps: collecting data in each intelligent coffee station in real time, so as to calculate and analyze raw material depletion time, equipment maintenance urgency, a sales trend report and user preference data; raw material depletion time, equipment maintenance urgency, a sales trend report and user preference data are input into a preset background management system to generate an operation decision scheme, and the operation decision scheme is analyzed through the steps of real-time data acquisition, inventory consumption rate analysis, equipment health assessment, sales trend identification, user preference mining and the like. The background management system can carry out dynamic, accurate and automatic operation decision making on replenishment, maintenance and sales strategies of the intelligent coffee station, and the problems that replenishment is not timely, equipment faults are not timely processed and the sales strategies are difficult to adjust when the number of equipment is large, the distribution area is wide or user requirements change frequently in the prior art are solved.

Description

Technical Field

[0001] This application relates to the field of data processing, and in particular to a back-end management system and method for smart coffee stations. Background Technology

[0002] With the development of unmanned retail, smart beverage preparation terminals, and self-service equipment, smart coffee stations are gradually becoming more common in public office areas, transportation hubs, and commercial complexes. Smart coffee stations typically use automated components to quickly prepare coffee and other beverages. Users can place orders and make payments through touch interfaces or mobile devices, thus achieving unmanned and standardized beverage supply services.

[0003] In existing technologies, the back-end management of smart coffee stations is typically implemented through a remote monitoring system. This system mainly obtains basic operating information and simple inventory status of the equipment through periodic reporting or manual input. Then, operators make manual judgments and scheduling based on this information. By viewing the sales reports, equipment operation logs, and static information on the remaining raw materials of each device in the back-end, replenishment plans and maintenance plans can be formulated based on experience.

[0004] Although the daily management of smart coffee stations can be accomplished through remote monitoring and manual dispatch, when there are a large number of devices, a wide distribution area, or frequent changes in user needs, manual judgment and fixed-cycle maintenance methods often cannot keep up with dynamic changes in a timely manner. This can lead to problems such as untimely replenishment, untimely handling of equipment failures, and difficulty in adjusting sales strategies, resulting in reduced operational efficiency and a decline in user experience. Summary of the Invention

[0005] The technical problem to be solved by the present invention is that when the number of devices is large, the distribution area is wide, or the user demand changes frequently, there are problems such as untimely replenishment, untimely handling of equipment failures, and difficulty in adjusting sales strategies. Therefore, the present invention provides a back-end management system and method for smart coffee stations.

[0006] In view of this, a first aspect of the present invention provides a back-end management method for a smart coffee station, comprising: real-time collection of sales record data, equipment status data, raw material inventory data, and user behavior data in each smart coffee station; calculation of the current raw material consumption rate based on the sales record data and the raw material inventory data, and prediction of the raw material depletion time based on the current raw material consumption rate; analysis of the health score of each smart coffee station based on the equipment status data, and identification of the urgency of equipment maintenance based on the health score of each smart coffee station; analysis of sales trend reports and user preference data at different time periods and locations through the sales record data and the user behavior data; and inputting the raw material depletion time, the urgency of equipment maintenance, the sales trend reports, and the user preference data into a preset back-end management system to generate an operational decision plan.

[0007] Preferably, the step of real-time collection of sales record data, equipment status data, raw material inventory data, and user behavior data in each smart coffee station includes: collecting data from the main controller of the smart coffee station equipment, which automatically records the transaction timestamp, product code, product name, sales quantity, transaction amount, and payment method each time a user completes a transaction, through a data acquisition module preset in the smart coffee station equipment, to construct sales record data; reading the real-time values ​​of the weight sensor or liquid level sensor connected to the raw material silo in the smart coffee station equipment to obtain the raw material inventory data of each raw material container; reading the real-time monitoring values ​​of the temperature sensor, pressure sensor, flow sensor, and motor speed sensor distributed on each key component of the smart coffee station equipment's production system to obtain equipment operating parameters, and receiving the equipment fault codes and runtime records generated by the main controller to construct equipment status data; and recording the user's product browsing time, menu click path, and customized option selection on the interactive interface to construct user behavior data.

[0008] Preferably, the step of calculating the current raw material consumption rate based on the sales record data and the raw material inventory data, and predicting the raw material depletion time based on the current raw material consumption rate, includes: extracting the sales quantity of each product within a preset time window from the sales record data, and querying the corresponding raw material ratio relationship for each product according to a preset product formula database; multiplying the sales quantity and the raw material ratio relationship to obtain the theoretical consumption amount of each raw material within the preset time window, and dividing the theoretical consumption amount by the duration of the preset time window to obtain the theoretical consumption rate of each raw material; extracting the initial inventory amount and the final inventory amount for the same preset time window from the raw material inventory data, and obtaining the actual consumption rate of each raw material by calculating the difference between the initial inventory amount and the final inventory amount and dividing it by the duration of the preset time window; comparing and analyzing the theoretical consumption rate and the actual consumption rate to obtain the current raw material consumption rate, and dividing the current remaining inventory amount in the raw material inventory data by the current raw material consumption rate to obtain the raw material depletion time of each raw material.

[0009] Preferably, the step of analyzing the health score of each smart coffee station based on the equipment status data and identifying the urgency of equipment maintenance based on the health score of each smart coffee station includes: extracting the equipment operating parameters and equipment fault codes of each smart coffee station from the equipment status data; comparing the temperature sensor values, pressure sensor values, flow sensor values, and motor speed sensor values ​​in the equipment operating parameters with the corresponding standard operating ranges to obtain the health score of each operating parameter; calculating the fault deduction value based on the historical cumulative number of fault codes and the severity level of the fault type, and subtracting the fault deduction value from the health score to obtain the total equipment health score; and comparing the total equipment health score with a preset equipment maintenance urgency threshold to identify the urgency of equipment maintenance.

[0010] Preferably, the step of analyzing sales trend reports and user preference data for different time periods and locations using the sales record data and user behavior data includes: dividing the sales record data into multiple time periods according to transaction timestamps, statistically analyzing the total sales revenue, number of sales orders, and sales percentage of each product category for each smart coffee station within each time period, identifying the upward, downward, or stable trend of sales revenue based on the time period sales data to obtain the time period sales trend; grouping the sales record data according to the geographical location information of the smart coffee stations to obtain different location types, statistically analyzing the regional sales data of smart coffee stations of different location types, and identifying high-sales and low-sales areas based on the regional sales data to obtain the location sales trend; generating a sales trend report based on the time period sales trend and the location sales trend; extracting product browsing time and menu click paths from the user behavior data, and analyzing user preference data based on the product browsing time and menu click paths.

[0011] Preferably, the step of extracting product browsing time and menu click path from the user behavior data, and analyzing user preference data based on the product browsing time and menu click path, includes: statistically analyzing the cumulative browsing time of each product and the number of clicks on the menu click path within the product browsing time; identifying products with a cumulative browsing time greater than or equal to a preset time and a number of clicks greater than or equal to a preset number of clicks as high-attention products; identifying products with a cumulative browsing time greater than or equal to a preset time and a number of clicks less than a preset number of clicks, or products with a cumulative browsing time less than a preset time and a number of clicks greater than or equal to a preset number of clicks as hesitant products; extracting customized option selection records from the user behavior data; statistically analyzing the frequency distribution of user selections of sugar content, temperature, and milk volume based on the customized options; identifying the parameter combinations with the highest frequency in the selection frequency distribution as mainstream preferences; identifying parameter combinations with an upward trend in frequency in the selection frequency distribution as emerging preferences; and generating user preference data based on the high-attention products, the hesitant products, the mainstream preferences, and the emerging preferences.

[0012] Preferably, the step of inputting the raw material depletion time, the equipment maintenance urgency, the sales trend report, and user preference data into a preset backend management system to generate an operational decision plan includes: filtering out a list of smart coffee stations whose raw material depletion time is less than a preset replenishment threshold; obtaining the geographical location information of each smart coffee station in the list; and planning the shortest replenishment route based on the geographical location information using a shortest path algorithm; adding smart coffee stations with high equipment maintenance urgency to a preset maintenance task queue; obtaining the current location and current work status of maintenance personnel; and assigning corresponding maintenance tasks to maintenance personnel based on the current location and current work status using a task allocation algorithm; identifying the time periods and locations where sales in the sales trend report show a downward trend to generate corresponding promotional strategies; generating corresponding menu optimization suggestions based on high-attention products, hesitant products, mainstream preferences, and emerging preferences in the user preference data; and generating an operational decision plan using the replenishment route, the maintenance tasks, the promotional strategies, and the menu optimization suggestions.

[0013] A second aspect of this invention provides a back-end management system for smart coffee stations, comprising: a data acquisition module for real-time acquisition of sales record data, equipment status data, raw material inventory data, and user behavior data from each smart coffee station; a calculation module for calculating the current raw material consumption rate based on the sales record data and the raw material inventory data, and predicting the raw material depletion time based on the current raw material consumption rate; an identification module for analyzing the health score of each smart coffee station based on the equipment status data, and identifying the urgency of equipment maintenance based on the health score of each smart coffee station; an analysis module for analyzing sales trend reports and user preference data for different time periods and locations through the sales record data and the user behavior data; and a generation module for inputting the raw material depletion time, the equipment maintenance urgency, the sales trend reports, and the user preference data into a preset back-end management system to generate an operational decision plan.

[0014] The technical solution of this invention has the following advantages: Through real-time data collection, inventory consumption rate analysis, equipment health assessment, sales trend identification, and user preference mining, it can make dynamic, accurate, and automated operational decisions on replenishment, maintenance, and sales strategies for smart coffee stations. It does not rely on human experience judgment but is based on continuous analysis of real-time data, thereby significantly improving the accuracy and timeliness of decisions, reducing the risk of supply disruptions due to replenishment delays, and minimizing equipment downtime losses due to maintenance delays. Furthermore, the automatically generated sales trend reports and user preference data can support rapid adjustments to product structure and marketing strategies, enabling smart coffee stations to better suit the consumption characteristics of different locations and time periods, improving sales conversion rates and user satisfaction. This solves the problems in existing technologies where, when there are many devices, a wide distribution area, or frequent changes in user demand, there are issues such as untimely replenishment, untimely handling of equipment malfunctions, and difficulty in adjusting sales strategies. Attached Figure Description

[0015] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating the backend management method of a smart coffee station provided in an embodiment of the present invention; Figure 2 This is a schematic block diagram illustrating the structure of the back-end management system for a smart coffee station provided in an embodiment of the present invention.

[0017] Figure label: 10. Back-end management system for smart coffee stations; 11. Data acquisition module; 12. Calculation module; 13. Recognition module; 14. Analysis module; 15. Generation module. Detailed Implementation

[0018] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] In the description of this invention, it should be noted that the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0020] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0021] Example 1: like Figure 1 As shown in the example, this application also provides a back-end management method for smart coffee stations, which can be used for operational monitoring, replenishment planning, maintenance scheduling, and sales strategy optimization of multiple smart coffee stations distributed in different locations. This back-end management method includes: Step S1: Collect sales records, equipment status data, raw material inventory data, and user behavior data in real time at each smart coffee station.

[0022] The main controller, installed inside the smart coffee station equipment, collects and processes various types of data in real time. This includes sales record data such as transaction timestamp, product name, product code, sales quantity, transaction amount, and payment method; equipment status data such as operating parameters and fault codes; raw material inventory data such as the real-time remaining quantity in each raw material container; and user behavior data such as user browsing and selection behavior on the interactive interface. The main controller records all of this data in real time via an internal data bus and uploads it to the backend management system at fixed intervals.

[0023] Once a user completes a latte order, the main controller writes the order's transaction timestamp, product code, corresponding coffee recipe, and temperature sensor, flow sensor, and motor speed sensor values ​​during the brewing process into the local operation log, and uploads it to the backend management system within 5 seconds.

[0024] Step S2: Calculate the current raw material consumption rate based on sales record data and raw material inventory data, and predict the raw material depletion time based on the current raw material consumption rate.

[0025] By statistically analyzing the sales volume of each product within a specified time window in the sales records, and combining this with the raw material ratios for each product in the product formula database, the theoretical consumption of each raw material is obtained. Then, the difference between the initial inventory and the current inventory is extracted from the raw material inventory data to calculate the actual consumption. Based on the difference between the theoretical consumption and the actual consumption, the current raw material consumption rate is determined and the time when the raw materials are exhausted is predicted.

[0026] If 20 cups of Americano are sold within the last 30 minutes, and each cup of Americano requires 15 grams of coffee beans, then the theoretical consumption of coffee beans is 300 grams. If the original inventory is 2000 grams and the current inventory is 1730 grams, then the actual consumption is 270 grams. If the deviation between theoretical consumption and actual consumption is less than 10%, then the weighted average is taken, for example, (300 grams + 270 grams) / 2 = 285 grams. The hourly consumption rate is obtained based on the length of the time window, and then the current inventory is divided by the consumption rate to obtain the estimated time when the raw materials will be exhausted.

[0027] Step S3: Analyze the health score of each smart coffee station based on the equipment status data, and identify the urgency of equipment maintenance based on the health score of each smart coffee station.

[0028] The values ​​of temperature sensor, pressure sensor, flow sensor and motor speed sensor are compared with their respective standard operating ranges, and full score and deduction strategies are set according to the comparison results; then, the fault deduction value is calculated according to the fault type and frequency of occurrence of the equipment fault code, and finally the total health score of the equipment is obtained. The total score is compared with the preset maintenance threshold to identify the urgency of maintenance.

[0029] If the standard operating range of the flow sensor is 200-260 mL / min, but the actual measured flow rate is 170 mL / min, the set health score will be directly deducted from the corresponding item of the sensor. If the equipment experiences three consecutive abnormal failures of the heating module, additional health scores will be deducted according to the severity of the failure, so that the equipment health score drops below the maintenance warning threshold, and the system will determine that the equipment has a high maintenance urgency.

[0030] Step S4: Analyze sales trend reports and user preference data for different time periods and locations using sales record data and user behavior data.

[0031] Sales record data is grouped and statistically analyzed according to time period and geographical location to identify sales trends and regional differences; at the same time, user preference data is analyzed by statistically analyzing the duration of product browsing and the frequency of selection of customization options in the browsing interface.

[0032] During weekday lunch hours, sales of milk-based coffee products in commercial areas are higher than at other times, while in campus areas, cold brew products are more popular in the afternoon. Based on this, the system generates sales trend reports and user preference data.

[0033] Step S5: Input the raw material depletion time, equipment maintenance urgency, sales trend report, and user preference data into the preset back-end management system to generate an operational decision plan.

[0034] The above analysis conclusions are input into the decision generation module in the back-end management system. The decision generation module outputs replenishment plans, maintenance schedules, promotional strategies, and menu optimization strategies based on replenishment priority, maintenance urgency, sales trend changes, and user preference characteristics.

[0035] When a sales trend report for a certain device's region shows a decline in sales, and user preference data indicates that cold brew beverages are the most popular products, the system can generate a promotional strategy of "offering a limited-time discount on cold brew beverages" and automatically push it to the operations team.

[0036] In this embodiment, real-time data collection of sales records, equipment status data, raw material inventory data, and user behavior data from each smart coffee station provides dynamic information on sales, operation, inventory, and user interaction. Subsequently, the current raw material consumption rate is calculated based on the sales record data and raw material inventory data, and the raw material depletion time is predicted based on this rate, allowing for advance assessment of replenishment needs. Furthermore, a health score is analyzed for each smart coffee station based on the equipment status data, and the urgency of equipment maintenance is identified based on the health score to guide the priority of maintenance task scheduling. Simultaneously, sales trend reports and user preference data for different time periods and locations are analyzed using sales record data and user behavior data to reflect changes in market demand. Finally, the raw material depletion time, equipment maintenance urgency, sales trend reports, and user preference data are input into a pre-defined backend management system, which automatically generates an operational decision-making plan for unified management and dynamic adjustment of replenishment plans, maintenance arrangements, promotional strategies, and menu optimization. This addresses the problems in existing technologies where, with a large number of devices, a wide distribution area, or frequent changes in user demand, there are issues such as untimely replenishment, untimely handling of equipment malfunctions, and difficulty in adjusting sales strategies.

[0037] By updating raw material depletion times in real time, order failures caused by raw material shortages can be effectively reduced; by identifying the urgency of equipment maintenance based on health scores, sudden equipment failures can be reduced and the stable operation rate of equipment can be improved; by analyzing sales trend reports and user preference data, product strategies and promotional activities can be quickly adjusted, thereby increasing sales and user satisfaction; overall, this method can significantly improve the operational efficiency, resource scheduling rationality, and service experience quality of smart coffee stations.

[0038] Example 2: Furthermore, in step S1, the following embodiment is preferred: The data acquisition module pre-installed in the smart coffee station device collects data from the main controller of the smart coffee station device, which automatically records the transaction timestamp, product code, product name, sales quantity, transaction amount, and payment method each time a user completes a transaction, in order to build sales record data.

[0039] The data acquisition module establishes a data interaction connection with the main controller of the smart coffee station equipment through an internal bus. After the user completes payment, the main controller stores the order data in the local cache. The data acquisition module polls the data in the cache periodically, with a fixed polling period of 1 second. During the polling, it reads the newly added order data and stores it in the local sales record storage table. At the same time, it adds a data acquisition time tag to each record and sends the sales record data to the backend management server in JSON format through the backend communication module to ensure that the data transmission format is consistent and facilitates subsequent data cleaning and analysis.

[0040] For example, when a user purchases a "Latte" at 10:25:36, the main controller records the transaction timestamp as "2024-04-12 10:25:36", the product code as "LT001", the product name as "Latte", the sales quantity as "1", the transaction amount as "18.00", and the payment method as "mobile payment". The data acquisition module writes the above data into the local sales record table and uploads it to the backend.

[0041] The system reads real-time values ​​from the weight or level sensors in the raw material silos connected to the smart coffee station equipment to obtain raw material inventory data for each raw material container.

[0042] Each raw material container corresponds to an independent weight sensor or liquid level sensor. The data acquisition module reads the real-time measurement value of the sensor at a fixed acquisition cycle of 5 seconds, and performs unit unification processing on the real-time measurement value (for example, converting the weight to grams (g) and the liquid level to milliliters (mL)). Combined with the capacity parameter of the raw material container, the remaining inventory of raw materials is calculated, and finally a raw material inventory data table is constructed.

[0043] If the weight sensor reading for the coffee bean container is 680g and the original maximum capacity is 1000g, then the current inventory is recorded as 680g; if the liquid level sensor for the milk tank shows a remaining capacity of 1200mL, then the inventory is recorded as 1200mL.

[0044] The system reads real-time monitoring values ​​from temperature sensors, pressure sensors, flow sensors, and motor speed sensors distributed on key components of the smart coffee station equipment to obtain equipment operating parameters, and receives equipment fault codes and runtime records generated by the main controller to construct equipment status data.

[0045] The data acquisition module reads parameters from the manufacturing system via the CAN bus and samples the equipment operating parameters at a fixed interval of 3 seconds. It converts temperature sensor data into Celsius (°C), pressure sensor data into kilopascals (kPa), flow sensor data into milliliters per second (mL / s), and motor speed sensor data into revolutions per minute (rpm). It also records the running time and fault code byte sequence and stores them as an equipment status data record table.

[0046] For example, the temperature sensor reading is 87℃, the pressure sensor reading is 220kPa, the flow sensor reading is 11mL / s, the motor speed sensor reading is 1200rpm, the current cumulative running time is 358 hours, and the most recent fault code is "E03" (indicating that the extraction pressure deviates from the standard value).

[0047] Record the duration of product browsing, menu click paths, and customized option selections on the user interface to build user behavior data.

[0048] The interactive interface has an embedded event collection script that triggers and records click events and dwell time events of different UI components. It also records the user's dwell time on the product page in milliseconds (ms), encodes the menu click path as a click sequence, and records customized options (such as sugar content, temperature, milk volume) as parameter selection values ​​and stores them in a local user behavior data table.

[0049] If a user stays on the "Americano" page for 4.2 seconds and the click sequence is "Home → Coffee → Americano → Add Ice → Confirm", the customized parameters are recorded as "Sugar content: Sugar-free, Temperature: Cold, Milk content: 0".

[0050] Furthermore, in step S2, the following embodiment is preferred: Extract the sales quantity of each product within a preset time window from the sales record data, and query the corresponding raw material ratio relationship of each product according to the preset product formula database.

[0051] The back-end management system groups sales record data by time window (e.g., 30 minutes or 1 hour), calculates the cumulative sales quantity of each product in each group, and reads the raw material usage parameters corresponding to each product from the product formula database to form a product-raw material ratio mapping table for subsequent calculation of raw material consumption.

[0052] If 25 cups of "latte" are sold within 1 hour, the product recipe database records that each cup of latte consumes 12g of coffee beans, 150mL of milk, and 5mL of syrup.

[0053] The theoretical consumption of each raw material within a preset time window is obtained by multiplying the sales volume with the raw material ratio. The theoretical consumption is then divided by the duration of the preset time window to obtain the theoretical consumption rate of each raw material.

[0054] The system performs product calculations for each commodity and summarizes the results by raw material dimension. Finally, it divides the result by the time window length (e.g., 3600 seconds) to obtain a quantified value of the theoretical consumption rate.

[0055] If the theoretical consumption of coffee beans in a latte is 25 × 12g = 300g, then the theoretical consumption rate is 300g / hour.

[0056] Extract the initial and final inventory levels for the same preset time window from the raw material inventory data. Calculate the difference between the initial and final inventory levels and divide it by the duration of the preset time window to obtain the actual consumption rate of various raw materials.

[0057] The initial and final inventory values ​​are read from the timestamp index of the raw material inventory database, and the actual consumption rate is obtained by the difference / duration method.

[0058] If the initial inventory is 900g and the final inventory is 610g, then the actual consumption is 290g, and the actual consumption rate is 290g / hour.

[0059] The theoretical consumption rate is compared with the actual consumption rate to obtain the current raw material consumption rate. When the deviation between the two exceeds a preset threshold, the actual consumption rate is used as the current raw material consumption rate. When the deviation does not exceed the preset threshold, the weighted average of the theoretical and actual consumption rates is used as the current raw material consumption rate. The current remaining inventory in the raw material inventory data is divided by the current raw material consumption rate to obtain the raw material depletion time of various raw materials. The predicted depletion time is then corrected by time series. The time series correction is based on the periodic sales fluctuation pattern identified by historical sales records of the same period, and the raw material consumption rate for future periods is dynamically adjusted based on the periodic sales fluctuation pattern to obtain the corrected raw material depletion time.

[0060] The deviation threshold is set to a fixed value (e.g., 10%), and the weighted average weight can be set to 0.4 for the theoretical rate and 0.6 for the actual rate. The seasonal exponential correction or moving average correction method is commonly used for time series correction.

[0061] When the theoretical rate is 300g / hour and the actual rate is 290g / hour, the deviation is 3.3%, which is lower than the threshold. Then the current consumption rate is (300×0.4+290×0.6)=294g / hour. When the remaining inventory is 680g, the raw material depletion time is 680 / 294≈2.31 hours, which is adjusted to about 1.9 hours after periodic midday peak sales fluctuations.

[0062] Furthermore, in step S3, the following embodiment is preferred: Extract the equipment operating parameters and equipment fault codes of each smart coffee station from the equipment status data.

[0063] The backend management system parses the equipment status data, associates the values ​​of temperature sensor, pressure sensor, flow sensor, and motor speed sensor by equipment number, records the equipment fault codes uploaded by the main controller in chronological order as a fault event sequence, and stores the equipment runtime field as a reference parameter for equipment lifespan, forming the basic dataset for equipment operation, which is used for subsequent health score analysis and calculation.

[0064] For example, the back-end management system can parse the equipment status data to find that the temperature of the extraction module in a certain smart coffee station is 86℃, the extraction pressure is 225kPa, the flow rate is 10mL / s, the motor speed is 1180rpm, and at the same time record the fault code "E03" and the running time is 360 hours.

[0065] The values ​​of temperature sensor, pressure sensor, flow sensor, and motor speed sensor in the equipment's operating parameters are compared with the corresponding standard operating ranges to obtain the health score of each operating parameter.

[0066] The background management system reads the standard operating range of each parameter from the equipment standard operating parameter library. For example, the standard temperature range is 84℃ to 92℃, the standard pressure range is 210kPa to 245kPa, the standard flow range is 9mL / s to 14mL / s, and the standard motor speed range is 1150rpm to 1350rpm. If the parameter value is within the standard operating range, the corresponding item's health score is 100 points. If it deviates from the standard range, the health score is calculated according to the deviation ratio and the linear deduction rule.

[0067] If the actual temperature is 86℃ and within the standard range, the temperature item scores 100 points; if the actual pressure is 225kPa and within the range, the pressure item scores 100 points; if the actual flow rate is 8.5mL / s, which is lower than the standard lower limit of 9mL / s, the deviation is 0.5mL / s, and the deduction is 0.5×10=5 points, so the flow rate item scores 95 points; if the motor speed is 1180rpm and within the range, the score is 100 points.

[0068] The fault deduction value is calculated based on the historical cumulative number of equipment fault codes and the severity level of the fault type. The total health score of the equipment is obtained by subtracting the fault deduction value from the health score.

[0069] The backend management system presets severity level coefficients for different fault codes. For example, the severity level coefficient for minor faults is 1, the severity level coefficient for moderate faults is 3, and the severity level coefficient for serious faults is 6. The cumulative number of faults is multiplied by the corresponding severity level coefficient to obtain the fault deduction value. Then, the fault deduction value is deducted from the total health score of the operating parameters (average of four items) obtained in the previous step to obtain the total health score of the equipment.

[0070] If the equipment's overall health score is 98, and fault code "E03" indicates a moderate fault level with a coefficient of 3, and has occurred twice in the past week, then the fault deduction is 2 × 3 = 6 points, and the total equipment health score is 98. 6 = 92 points.

[0071] The total equipment health score is compared with a preset equipment maintenance urgency threshold to identify equipment maintenance urgency.

[0072] The device's total health score is compared with a threshold (e.g., set to 85 points). When the device's total health score is below the threshold, it is marked as needing immediate maintenance. When the total health score is greater than or equal to the threshold, it is marked as being in normal operating condition.

[0073] When the total health score of the equipment is 92 points, which is higher than 85 points, it is identified as a normal state; if the total health score of the equipment is 78 points, which is lower than 85 points, it is identified as a state with high maintenance urgency.

[0074] Furthermore, in step S4, the following embodiment is preferred: Sales record data is divided into multiple time periods according to transaction timestamps. These time periods include those divided by hour, by weekdays and rest days, and by season. The total sales revenue, number of sales orders, and sales percentage of each product category for each smart coffee station are statistically analyzed within each time period. Based on the time period sales data, the upward trend, downward trend, or stable trend of sales revenue are identified to obtain the time period sales trend.

[0075] The back-end management system aggregates sales record data in segments by timestamp, uses a moving average algorithm to smooth the rate of change of sales indicators for each period, and determines the trend type based on the positive, negative, and near-zero values ​​of the rate of change.

[0076] If the sales of a smart coffee station in a commercial area show an increasing rate of change for three consecutive periods during the weekday afternoon period from 16:00 to 19:00, it is identified as having a significant upward trend during that period.

[0077] Sales record data is grouped according to the geographical location information of smart coffee stations to obtain different location types. Regional sales data of smart coffee stations in different location types are statistically analyzed. Location types include office building areas, commercial areas, transportation hub areas, residential areas and campus areas. Based on the regional sales data, high-sales areas and low-sales areas are identified to obtain location sales trends.

[0078] The backend management system calls the geographic tag mapping table to automatically classify each smart coffee station according to its geographical location. It calculates the average sales and order volume for each type of location, and marks the top 20% of the regions in terms of sales volume as high-sales regions and the bottom 20% as low-sales regions.

[0079] If the average daily sales in an office building area of ​​a city is 2,800 yuan, and the average daily sales in a campus area is 1,500 yuan, then the office building area is identified as a high-sales area.

[0080] A sales trend report is generated based on time-of-day and location-based sales trends.

[0081] The back-end management system generates sales trend reports by presenting trend conclusions in the form of charts and text descriptions, and stores them in the back-end management data warehouse. It also supports operations personnel to view and export the reports.

[0082] The report can demonstrate the trend that "sales peak in commercial areas around lunchtime, while sales peak in transportation hubs during the morning rush hour."

[0083] Product browsing time and menu click paths are extracted from user behavior data, and user preference data is analyzed based on these data.

[0084] The system performs correlation analysis on browsing time and menu click paths, and identifies the user's level of interest through a page dwell weight model.

[0085] If a user browses the "Mocha" product page for 3.8 seconds and clicks to enter the details page, it is identified as having a clear preference intention.

[0086] The steps of extracting product browsing time and menu click path from user behavior data, and analyzing user preference data based on product browsing time and menu click path, can preferably be: The system calculates the cumulative browsing time and the number of clicks along the menu path for each product within the total browsing time. Products with a cumulative browsing time greater than or equal to a preset duration and a number of clicks greater than or equal to a preset number of clicks are identified as high-attention products.

[0087] The preset duration can be set to 3 seconds, and the preset number of clicks can be set to 2. If a user spends a total of 5 seconds browsing the "Latte" page and clicks 3 times, then "Latte" will be identified as a high-interest product.

[0088] Products whose cumulative browsing time is greater than or equal to the preset duration and whose number of clicks is less than the preset number of clicks, or whose cumulative browsing time is less than the preset duration and whose number of clicks is greater than or equal to the preset number of clicks, are identified as hesitant products.

[0089] A product that is considered "hesitant" indicates that the user has shown interest but has not yet made a final decision. For example, browsing "Matcha Latte" for 4 seconds without clicking on it is identified as a product that is considered "hesitant".

[0090] Customized option selection records are extracted from user behavior data. Based on the customized options, the frequency distribution of user selections of sugar content, temperature, and milk volume is statistically analyzed. The parameter combinations with the highest frequency in the selection frequency distribution are identified as mainstream preferences, and the parameter combinations with an upward trend in selection frequency are identified as emerging preferences.

[0091] The backend system uses frequency distribution histograms to analyze user parameter preference trends. If the combination of sugar-free + hot beverage + less milk has the highest frequency, it is identified as a mainstream preference; if the combination of iced beverage + low sugar has a growth rate of 18% in the past week, it is identified as an emerging preference.

[0092] User preference data is generated based on high-attention products, hesitant products, mainstream preferences, and emerging preferences.

[0093] The above identification results are modeled separately according to user and site dimensions, generating data outputs that can be used for product recommendation and menu optimization. For a specific regional site, user preference data shows that consumers in that region prefer low-sugar and low-dairy formula drinks.

[0094] Furthermore, in step S5, the following embodiment is preferred: A list of smart coffee stations whose raw material depletion time is less than a preset replenishment threshold is selected. The geographical location information of each smart coffee station in the list is obtained, and the shortest replenishment route is planned based on the geographical location information using the shortest path algorithm.

[0095] By periodically querying the backend database of the smart coffee station, the system retrieves the raw material depletion time data of each station, compares it with the preset replenishment threshold, and outputs the smart coffee stations that meet the conditions in the form of station number, geographical coordinates, and raw material depletion type, forming a replenishment demand list for smart coffee stations. Subsequently, the system reads the latitude and longitude coordinates of the smart coffee stations in the map coordinate system, calls the shortest path algorithm (either Dijkstra's algorithm or A* search algorithm can be used) to calculate the shortest driving path between different replenishment stations, and plans the global path based on the route structure of vehicles departing from and returning to the logistics center, to obtain the replenishment order and the shortest replenishment route.

[0096] When the replenishment threshold is set to 12 hours, the system queries the inventory management database to find that the corrected raw material depletion times for smart coffee stations A, B, and C are 8 hours, 10 hours, and 16 hours, respectively. Then, smart coffee stations A and B are added to the replenishment list, and the A* algorithm is used to plan the path from warehouse → A → B → back to warehouse, generating a replenishment route of warehouse → station B → station A → warehouse, with a minimum total distance of 18.6 kilometers.

[0097] Smart coffee stations with high equipment maintenance urgency are added to a preset maintenance task queue. The current location and current work status of maintenance personnel are obtained, and a task allocation algorithm is used to assign corresponding maintenance tasks to maintenance personnel based on their current location and current work status.

[0098] The system reads the maintenance urgency level, equipment failure type, and component type requiring replacement or repair for each site in the maintenance task queue. Simultaneously, it obtains the real-time geographical location, remaining time of the current task, and skill matching level of the maintenance personnel through the personnel management module. Subsequently, the system calls the task allocation algorithm (either a weighted nearest neighbor matching algorithm or a Hungarian algorithm) to comprehensively consider factors such as maintenance priority, distance between maintenance personnel and sites, and skill matching degree of maintenance personnel to generate a mapping relationship between maintenance personnel and maintenance tasks.

[0099] The pump body of equipment X at maintenance site is malfunctioning, and the maintenance is urgent. The corresponding skill requirement is "hydraulic repair". Among maintenance personnel P1 and P2, P1 is closer but does not have the skill, while P2 is farther away but has the skill. Therefore, the system will assign the maintenance task to P2.

[0100] Identify the time periods and locations where sales are declining in the sales trend report to generate corresponding promotional strategies, including the selection of discounted products, the setting of discount levels, and the arrangement of promotional periods.

[0101] Based on the "downward trend" label in the sales trends of time period and location, select smart coffee stations in the corresponding time period and location to implement promotional strategies. The promotional strategy parameters include the list of discounted products, the discount percentage, and the execution time range. The list of discounted products is selected from the high-attention products in the user preference data, and the discount percentage is preset according to the range of decline. For example, when the decline is greater than 20%, the discount percentage is set to 20%, and when the decline is between 10% and 20%, the discount percentage is set to 10%.

[0102] If a significant downward trend in sales is identified in office building areas between 2:00 PM and 5:00 PM on weekdays, "Americano (medium)" will be designated as a discounted item with a 10% discount, and the promotion period will be from 2:00 PM to 5:00 PM.

[0103] Based on user preference data, we can generate corresponding menu optimization suggestions for high-interest products, hesitant products, mainstream preferences, and emerging preferences.

[0104] The system selects high-interest products and mainstream preference combinations from user preference data, analyzes their current display level and recommendation priority in the menu, and adjusts them to the homepage recommendation position; for hesitant products and emerging preference combinations, the system sets "new product recommendation" or "customized recommendation" prompt labels to increase exposure; the system pushes the optimized menu content to the front-end interactive interface of the smart coffee station by refreshing the configuration file.

[0105] If the user preference data shows that the combination of "less sugar + low temperature + milk" is selected most frequently, the system will promote the product corresponding to this combination to the first level of the menu.

[0106] Generate operational decision-making solutions using replenishment routes, maintenance tasks, promotional strategies, and menu optimization suggestions.

[0107] The four types of decision-making content are output in a structured data format, including a replenishment plan table, a maintenance scheduling table, a promotion strategy configuration table, and a menu layout adjustment table, and are synchronized to the operations personnel backend and the smart coffee station main controller execution terminal via API push. The generated operational decision plan includes specific vehicle routes, maintenance personnel assignments, a list of discounted products and promotion execution times, as well as a table adjusting the order of products displayed on the homepage in the menu.

[0108] Example 2: like Figure 2 As shown in the example, this application also provides a back-end management system 10 for a smart coffee station, which mainly includes a data acquisition module 11, a calculation module 12, a recognition module 13, an analysis module 14, and a generation module 15.

[0109] The data acquisition module 11 is mainly used to collect sales record data, equipment status data, raw material inventory data and user behavior data in each smart coffee station in real time.

[0110] The calculation module 12 is mainly used to calculate the current raw material consumption rate based on sales record data and raw material inventory data, and to predict the raw material depletion time based on the current raw material consumption rate.

[0111] The identification module 13 is mainly used to analyze the health score of each smart coffee station based on the equipment status data, and to identify the urgency of equipment maintenance based on the health score of each smart coffee station.

[0112] Analysis module 14 is mainly used to analyze sales trend reports and user preference data at different times and locations through sales record data and user behavior data.

[0113] The generation module 15 is mainly used to input raw material depletion time, equipment maintenance urgency, sales trend reports, and user preference data into the preset back-end management system to generate operational decision-making plans.

[0114] In this embodiment, the data acquisition module 11 performs unified data access and storage for sales record data, equipment status data, raw material inventory data, and user behavior data. The calculation module 12 performs raw material consumption rate calculation and raw material depletion time prediction based on the sales record data and raw material inventory data. The identification module 13 completes equipment health score calculation and equipment maintenance urgency identification based on the equipment status data. The analysis module 14 generates sales trend reports and user preference data based on the sales record data and user behavior data. The generation module 15 comprehensively processes the raw material depletion time, equipment maintenance urgency, sales trend reports, and user preference data to output an operational decision-making scheme that can be directly used for replenishment planning, maintenance task arrangement, promotion strategy design, and menu optimization suggestions. This enables refined operation management of smart coffee stations, improves raw material replenishment efficiency, equipment maintenance efficiency, and product sales matching, reduces manual management costs, and increases overall operational revenue.

[0115] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the device and each module described above can be referred to the corresponding process in the aforementioned Embodiment 1, and will not be repeated here.

[0116] The above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for back office management of an intelligent coffee station, characterized in that, The method comprises the following steps: real-time collection of sales record data, equipment state data, raw material inventory data and user behavior data in each intelligent coffee station; calculating the current raw material consumption rate according to the sales record data and the raw material inventory data, and predicting the raw material depletion time based on the current raw material consumption rate; analyzing the health score of each intelligent coffee station based on the equipment state data, and identifying the equipment maintenance urgency based on the health score of each intelligent coffee station; analyzing the sales trend report and user preference data of different time periods and different locations through the sales record data and the user behavior data; inputting the raw material depletion time, the equipment maintenance urgency, the sales trend report and the user preference data into a preset background management system to generate an operation decision scheme.

2. The method of claim 1, wherein, The step of real-time collection of sales record data, equipment state data, raw material inventory data and user behavior data in each intelligent coffee station comprises the following steps: collecting the transaction timestamp, product code, product name, sales quantity, transaction amount and payment method recorded by the main controller of the intelligent coffee station device automatically when each user completes a transaction through a data collection module preset in the intelligent coffee station device to construct sales record data; reading the real-time values of the weight sensor or liquid level sensor connected to the raw material bin of the intelligent coffee station device to obtain the raw material inventory data of each raw material container; reading the real-time monitoring values of the temperature sensor, pressure sensor, flow sensor and motor speed sensor distributed on each key component of the production system of the intelligent coffee station device to obtain the equipment operation parameters, and receiving the equipment fault code and operation duration record generated by the main controller to construct the equipment state data; recording the product browsing duration, menu click path and customized option selection of the user on the interactive interface to construct the user behavior data.

3. The method of claim 1, wherein, The step of calculating the current raw material consumption rate according to the sales record data and the raw material inventory data, and predicting the raw material depletion time based on the current raw material consumption rate comprises the following steps: extracting the sales quantity of each product within a preset time window from the sales record data, and querying the raw material proportioning relationship corresponding to each product according to a preset product formula database; performing product operation on the sales quantity and the raw material proportioning relationship to obtain the theoretical consumption amount of each raw material within the preset time window, dividing the theoretical consumption amount by the duration of the preset time window to obtain the theoretical consumption rate of each raw material; extracting the starting inventory and ending inventory of the same preset time window from the raw material inventory data, calculating the difference between the starting inventory and the ending inventory and dividing the difference by the duration of the preset time window to obtain the actual consumption rate of each raw material; comparing and analyzing the theoretical consumption rate and the actual consumption rate to obtain the current raw material consumption rate, and dividing the current remaining inventory in the raw material inventory data by the current raw material consumption rate to obtain the raw material depletion time of each raw material.

4. The method of claim 1, wherein, The step of analyzing the health score of each intelligent coffee station based on the device state data and identifying the device maintenance urgency based on the health score of each intelligent coffee station comprises: extracting the device running parameters and device fault codes of each intelligent coffee station from the device state data; comparing the temperature sensor value, pressure sensor value, flow sensor value and motor speed sensor value in the device running parameters with the corresponding standard running interval to obtain the health score of each running parameter; calculating the fault deduction value according to the historical cumulative number of the device fault codes and the severity level of the fault type, and obtaining the total device health score by subtracting the fault deduction value from the health score; comparing the total device health score with the preset device maintenance urgency threshold to identify the device maintenance urgency.

5. The method of claim 2, wherein, The step of analyzing the sales trend report and user preference data of different time periods and different locations based on the sales record data and the user behavior data comprises: dividing the sales record data into multiple time periods according to the transaction time stamp, and calculating the total sales, the number of sales orders and the sales proportion of each product category in each time period to obtain the time period sales data, and identifying the upward trend, downward trend or stable trend of sales based on the time period sales data to obtain the time period sales trend; grouping the sales record data according to the geographical location information of the intelligent coffee station to obtain different location types, and calculating the regional sales data of the intelligent coffee stations in different location types based on the regional sales data to identify the high sales area and the low sales area to obtain the location sales trend; generating a sales trend report based on the time period sales trend and the location sales trend; extracting the product browsing time and menu click path from the user behavior data, and analyzing the user preference data based on the product browsing time and the menu click path.

6. The method of claim 5, wherein, The step of extracting the product browsing time and menu click path from the user behavior data, and analyzing the user preference data based on the product browsing time and the menu click path comprises: calculating the cumulative browsing time of each product in the product browsing time and the number of clicks of the menu click path, and identifying the product with a cumulative browsing time greater than or equal to a preset time and a number of clicks greater than or equal to a preset number of clicks as a high attention product; identifying the product with a cumulative browsing time greater than or equal to a preset time and a number of clicks less than a preset number of clicks or a cumulative browsing time less than a preset time and a number of clicks greater than or equal to a preset number of clicks as a hesitant product; extracting the customized option selection record from the user behavior data, and calculating the selection frequency distribution of sugar, temperature and milk volume based on the customized options, identifying the parameter combination with the highest selection frequency in the selection frequency distribution as the mainstream preference, and identifying the parameter combination with an upward trend in the selection frequency distribution as the emerging preference; generating user preference data based on the high attention product, the hesitant product, the mainstream preference and the emerging preference.

7. The method of claim 6, wherein, The step of inputting the raw material depletion time, the equipment maintenance urgency, the sales trend report, and user preference data into a preset background management system to generate an operation decision scheme includes: Screening a list of intelligent coffee stations with a raw material depletion time less than a preset replenishment threshold, obtaining geographical location information of each intelligent coffee station in the list, and planning the shortest replenishment route based on the geographical location information through a shortest path algorithm; Intelligent coffee stations with high equipment maintenance urgency are added to a preset maintenance task queue, the current location and current working state of maintenance personnel are obtained, and a task allocation algorithm is used to allocate corresponding maintenance tasks to the maintenance personnel based on the current location and the current working state; Identifying periods and locations with a declining sales trend in the sales trend report to generate corresponding promotion strategies; Based on the high-attention goods, hesitant goods, mainstream preferences, and emerging preferences in the user preference data, corresponding menu optimization suggestions are generated; Operation decision schemes are generated using the replenishment route, the maintenance task, the promotion strategy, and the menu optimization suggestion.

8. A back office management system for an intelligent coffee station, characterized in that, It includes: A collection module for collecting sales record data, equipment status data, raw material inventory data, and user behavior data in each intelligent coffee station in real time; A calculation module for calculating the current raw material consumption rate based on the sales record data and the raw material inventory data, and predicting the raw material depletion time based on the current raw material consumption rate; An identification module for analyzing the health score of each intelligent coffee station based on the equipment status data, and identifying the equipment maintenance urgency based on the health score of each intelligent coffee station; An analysis module for analyzing sales trend reports and user preference data at different times and in different locations through the sales record data and the user behavior data; A generation module for inputting the raw material depletion time, the equipment maintenance urgency, the sales trend report, and user preference data into a preset background management system to generate an operation decision scheme.

Images