System and method for monitoring can container usage data

Abstract

We provide a can container usage data monitoring system and method that precisely predicts demand by time of day based on sealing can consumption data by store and calculates the optimal automatic delivery time and route by considering logistics conditions. A can container usage data monitoring system according to one aspect of the present invention comprises a memory configured to store commands and a processor configured to execute said commands to: receive a motor rotation signal from an encoder linked to a motor of a can seamer for sealing cans, analyze said motor rotation signal to detect a rotation cycle corresponding to a can sealing operation, count the number of can sealings for each detected rotation cycle and record it in said memory, and transmit accumulated can usage data according to the counting of the number of can sealings to a manager server.

Description

Technology Field

[0001] The present invention relates to a can container usage data monitoring system and method that implements an efficient supply system by analyzing operation data of sealing equipment in a store in real time to automatically measure the usage of components, predicting the time of inventory depletion based on this, and generating automatic ordering and logistics schedules in a timely manner. Background Technology

[0002] Various packaging technologies are being introduced in the food and beverage industry to enhance product hygiene and shelf life. Among these, sealing technology for liquid foods in cans is widely used by many establishments because it is effective in preventing spoilage and guaranteeing consistent quality to consumers. Particularly in franchises and independent shops that prepare customized beverages such as juices, cocktails, and mixed drinks, the practice of using small can sealing machines to seal drinks immediately upon ordering is becoming increasingly common. These sealing machines serve not only to maintain the freshness of food and beverage products but also to instill confidence in customers that the products were "prepared immediately on-site."

[0003] However, as the adoption of can sealing machines increases, so does the complexity of operations and supply chain management based on them. Each store must receive consumables, such as cans and lids used during machine operation, on a regular basis; failure to secure the appropriate quantity of inventory at the right time poses a risk of disrupting customer service or even suspending operations. Conversely, stockpiling unnecessarily large quantities of consumables leads to inefficient space utilization and increased costs. Particularly for franchise businesses operating dozens to hundreds of branches, forecasting demand for sealing cans and coordinating delivery timing are directly linked to operational efficiency.

[0004] Against this backdrop, some companies have collected usage pattern data over specific periods to implement simple stock shortage alerts or manual ordering systems; however, these systems still have the following limitations. First, most systems operate based on static criteria (e.g., daily average usage), failing to reflect the impact of hourly fluctuations in demand, specific events, or weather factors. For instance, even though customer traffic plummets on rainy evenings and demand surges during lunchtime on sunny, hot days, the failure to account for these factors leads to recurring issues of misdeliveries or underdeliveries. Second, applying uniform ordering standards despite varying consumption patterns based on store characteristics, location, and customer preferences results in some stores accumulating excess inventory while others experience operational disruptions due to early stockouts. Third, simple, repetitive logistics assignments that do not comprehensively consider delivery personnel and vehicle schedules, as well as store storage conditions, hinder logistics efficiency and lead to increased overall costs. In particular, existing systems capable of only reactive manual processing rather than proactive logistical responses for stores at risk of stock shortages are not suitable for the modern logistics environment that demands real-time performance and predictability.

[0005] Furthermore, while current technologies are limited to simply suggesting order timing or determining optimal inventory levels, solutions capable of automating proactive logistics timing decisions by integrating and analyzing complex factors—such as store-specific sealing machine usage logs, weather / day / time trends, delivery availability times, and store storage capacity—are currently lacking. In other words, there is virtually no technology that goes beyond merely storing data and issuing alerts to provide dynamic automatic scheduling that calculates logistics based on data and even considers the possibility of merged deliveries with other stores.

[0006] Therefore, there is a demand for an integrated solution that collects sealing machine data by store in real time to precisely analyze consumption by time period, automatically derives the optimal timing and route for logistics based on this, and supports customized operational settings tailored to the specific situation of each store based on SaaS. The problem to be solved

[0007] The present invention was devised to solve the aforementioned problems and aims to provide a can container usage data monitoring system and method that precisely predicts demand by time period based on sealing can consumption data by store and calculates the optimal automatic delivery time and route by considering logistics conditions.

[0008] However, the technical problems that the present invention aims to solve are not limited to those described above, and other unmentioned problems will be clearly understood by those skilled in the art from the description of the invention below. means of solving the problem

[0009] A can container usage data monitoring system according to one aspect of the present invention comprises a memory configured to store commands and a processor configured to execute said commands to: receive a motor rotation signal from an encoder linked to a motor of a can seamer for sealing cans, analyze said motor rotation signal to detect a rotation cycle corresponding to a can sealing operation, count the number of can sealings for each detected rotation cycle and record it in said memory, and transmit accumulated can usage data according to the counting of the number of can sealings to a manager server.

[0010] Preferably, the processor may be configured to determine whether there is a match by comparing the number of can sealings with the actual can inventory quantity, and if the cumulative error between the actual can inventory value and the number of can sealings counted through rotation cycle detection exceeds a predetermined standard, send a notification to the administrator server, store information on the consumption time and quantity of cans in the memory, and set a delivery trigger for the store by predicting the delivery cycle and quantity based on the stored data.

[0011] Preferably, the processor may be configured to assign a unique identifier (ID) to each of the multiple sealing heads of a can seamer, and to independently detect the rotation cycle of each sealing head based on the unique identifier to record how many sealing operations each sealing head has performed.

[0012] Preferably, the processor may be configured to transmit data on the number of can sealings and can usage collected from can seamers installed in multiple stores to a manager server, predict the time when cans will be depleted and when cans need to be refilled for each store based on the time-based can usage data of multiple stores stored in the manager server, generate a can ordering schedule for each store in accordance with the predicted time when cans need to be refilled, and establish a plan for delivering cans to stores corresponding to the can ordering schedule.

[0013] Preferably, the processor may be configured to calculate a delivery route by integrating can ordering schedules for multiple stores within a specific region to sequentially supply cans to nearby stores in a single delivery, calculate a risk score for each store by considering the delivery distance to each store, the can inventory depletion curve of each store, and logistics availability, and reorder the delivery order to each store according to the calculated risk score for each store.

[0014] Preferably, the processor may be configured to record can usage data by separating the consumption rate into component units, including the can body and lid of the can, independently analyze the inventory level and depletion rate for each component unit, and independently set a priority supply trigger for the corresponding component when the consumption rate of a specific component exceeds a preset threshold.

[0015] Preferably, the processor may be configured to calculate a Pressure Index using location information of each store and recent can usage data according to the formula (average speed of can seamer sealing × deviation from recent can consumption) ÷ current remaining quantity × delivery distance coefficient, calculate and sort the calculated Pressure Index for multiple stores, and then set priority supply triggers for the top N stores.

[0016] Preferably, the processor may be configured to analyze the usage pattern of a can seamer installed in a store to collect time-based operation history, cluster can consumption cycles and peak sections based on the number of can sealings by the can seamer in each store by reflecting can consumption trends by store, and predict whether pre-delivery of cans is required on a specific day and time based on time-based demand information derived from the clustered can consumption cycles and peak sections, and to determine whether to pre-delivery cans corresponding to a specific time on the next day or the next week.

[0017] Preferably, the processor may be configured to identify a Recycle Time Window within an existing can delivery plan to evaluate whether additional deliveries can be merged into the existing can delivery plan, collect can consumption and cap consumption rates from can usage data to determine the conditions for replenishment delivery, and when the conditions for replenishment delivery occur, examine the possibility of merging replenishment deliveries into the existing can delivery plan by considering the time, distance, and inventory leeway of the existing delivery route, and perform a simulation on delivery routes of multiple stores to calculate which delivery route is optimal for merging each replenishment delivery.

[0018] A method for monitoring can container usage data according to one aspect of the present invention comprises the steps of: receiving a motor rotation signal from an encoder linked to a motor of a can seamer for sealing cans; analyzing the motor rotation signal to detect a rotation cycle corresponding to a can sealing operation; counting and recording the number of can sealing operations for each detected rotation cycle; and transmitting accumulated can usage data based on the counting of can sealing operations to an administrator server. Effects of the invention

[0019] According to an embodiment of the present invention, by measuring the number of can sealing cycles in real time based on the motor rotation signal of a can seamer and automatically counting them, errors that occurred in conventional manual inspection or estimation methods can be significantly reduced. In particular, as the processor counts the number of can sealing cycles and transmits the accumulated usage data to a cloud server, managers or the headquarters system can monitor the consumption status of each store in real time at any time, and enable data-based automatic ordering and logistics planning. This dramatically improves management efficiency and data accuracy, and allows for the prevention of unexpected inventory shortages in advance.

[0020] In addition, various other additional effects may be achieved by various embodiments of the present invention. These various effects of the present invention are described in detail in each embodiment, or the description of effects that are easily understood by those skilled in the art is omitted. Brief explanation of the drawing

[0021] The following drawings attached to this specification illustrate preferred embodiments of the present invention and serve to further enhance understanding of the technical concept of the present invention together with the detailed description of the invention provided below; therefore, the present invention should not be interpreted as being limited only to the matters described in such drawings. FIG. 1 is a diagram illustrating components for performing a can container usage data monitoring method according to an embodiment of the present invention. FIG. 2 is a drawing for explaining the elements constituting a system for performing a can container usage data monitoring method according to an embodiment of the present invention. Figure 3 is a diagram showing an example of a method for monitoring can container usage data implemented by the system of Figure 2. FIGS. 4 to 8 are drawings illustrating other examples of a method for monitoring can container usage data implemented by the system of FIG. 2. FIG. 9 is a flowchart illustrating a method for monitoring can container usage data according to an embodiment of the present invention. Specific details for implementing the invention

[0022] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings. Prior to this, terms and words used in this specification and claims should not be interpreted as being limited to their ordinary or dictionary meanings, and should be interpreted in a meaning and concept consistent with the technical spirit of the present invention, based on the principle that the inventor can appropriately define the concept of the terms to best describe his invention.

[0023] Therefore, the embodiments described in this specification and the configurations illustrated in the drawings are merely the most preferred embodiments of the present invention and do not represent all of the technical ideas of the present invention; thus, it should be understood that various equivalents and modifications that can replace them may exist at the time of filing this application.

[0024] FIG. 1 is a diagram illustrating components for performing a can container usage data monitoring method according to an embodiment of the present invention, FIG. 2 is a diagram illustrating elements constituting a system (100) for performing a can container usage data monitoring method according to an embodiment of the present invention, and FIG. 3 is a diagram showing an example of a can container usage data monitoring method implemented by the system (100) of FIG. 2. In the embodiment of the present invention, the can seamer described below is denoted by the reference numeral 'C'.

[0025] Referring to FIGS. 1 to 3, each component for performing a can container usage data monitoring method according to one embodiment of the present invention may include a system (100) and an administrator server (200) for performing a can container usage data monitoring method.

[0026] Specifically, the system (100) can perform a method for monitoring can container usage data by exchanging data with a manager server (200) via wired or wireless communication. In one embodiment, the system (100) for performing the method for monitoring can container usage data may be a server device or a computer device. Also, the manager server (200) may be a server device or a computer device used by a manager.

[0027] A method for monitoring can container usage data can be implemented in the form of a computer program or a mobile application. For example, processes for the operation of the can container usage data monitoring method can be performed by the administrator server (200) running the computer program and the mobile app, and the system (100) running the computer program.

[0028] In addition, the system (100) that performs the method of monitoring can container usage data includes an AI model and can perform the method of monitoring can container usage data through the learning of the AI ​​model.

[0029] The above AI model may be trained based on supervised learning methods, but is not limited thereto, and may also be trained based on unsupervised learning or semi-persistent learning methods.

[0030] In addition, an AI model may consist of one or more neural network layers, and each neural network may include one or more weights. An AI model can perform learning or inference by performing operations between one or more weights and input data.

[0031] Referring to FIG. 2, a system (100) for performing a can container usage data monitoring method may include a memory (110) and a processor (120). However, it is not limited thereto, and other general-purpose components may be further included in the system (100) for performing a can container usage data monitoring method.

[0032] Memory (110) may be configured to store instructions of a computer program or application that implements a method for monitoring can container usage data, and a processor (120) may execute the program or application by executing the instructions stored in memory (110). For example, memory (110) may be implemented as non-volatile memory such as ROM, PROM, EPROM, EEPROM, flash memory, PRAM, MRAM, RRAM, FRAM, etc., or volatile memory such as DRAM, SRAM, SDRAM, PRAM, RRAM, FeRAM, etc., and may be implemented in the form of HDD, SSD, SD, Micro-SD, etc., or a combination thereof. The processor (120) may be implemented as an array of logic gates, a microprocessor, a CPU, a GPU, an AP, or a combination thereof.

[0033] Specifically, the processor (120) can receive a motor rotation signal from an encoder connected to the motor of a can seamer for sealing cans. In this process, the encoder detects the rotation angle of the motor in real time through a high-precision sensor, and the processor (120) collects the signal using a high-speed sampling method to remove noise and measure the accurate number of rotations. This method has the advantage of minimizing the accumulation of errors even in a repetitive operation environment compared to proximity sensors or simple time-based estimation.

[0034] Additionally, the processor (120) can analyze the motor rotation signal to detect a rotation cycle corresponding to the operation of sealing a can. During the detection process, the processor (120) uses the periodic characteristics of the rotation pattern to exclude irregular rotation or idle rotation and applies a filtering algorithm to identify only normal sealing cycles. This prevents incorrect counting and ensures data reliability.

[0035] Additionally, the processor (120) can count the number of can sealing cycles detected for each rotation cycle and record them in the memory (110).

[0036] Additionally, the processor (120) may be configured to transmit accumulated can usage data to the administrator server (200) based on the counting of can sealing counts. The transmitted data is aggregated in a cloud-based integrated management platform, and the server side can analyze usage patterns by time period to predict consumption, automatically place orders, and schedule deliveries. This minimizes logistics costs caused by stock shortages or urgent orders and enables supply chain optimization that reflects consumption characteristics by store.

[0037] According to the present invention, by measuring the number of can sealing cycles in real time based on the motor rotation signal of a can seamer and automatically counting them, errors that occurred in conventional manual inspection or estimation methods can be significantly reduced. In particular, since the processor (120) counts the number of can sealing cycles and transmits the accumulated usage data to a cloud server, the manager or headquarters system can monitor the consumption status of each store in real time at any time, and data-based automatic ordering and logistics planning can be established. This dramatically improves management efficiency and data accuracy, and prevents unexpected inventory shortages in advance.

[0038] In one embodiment, the processor (120) can determine whether there is a match by comparing the number of can sealings with the actual amount of can inventory.

[0039] Additionally, the processor (120) can send a notification to the administrator server (200) if the cumulative error between the actual can inventory value and the number of cat sealing counts counted through rotation cycle detection exceeds a predetermined standard.

[0040] And, the processor (120) can be configured to store information on the consumption time and quantity of the can in memory (110), and to predict the delivery cycle and quantity based on the stored data to set a delivery trigger for the store.

[0041] At this time, the processor (120) can strengthen the judgment logic by dynamically assigning weights to multiple factors to increase prediction accuracy and notification efficiency.

[0042] Specifically, the processor (120) may define the target factors for weighting as the consistency between the number of can sealings and the actual inventory quantity (first factor), the cumulative error rate (second factor), and the future depletion prediction curve based on the timing and pattern of consumption (third factor). Subsequently, if the combined value of the first factor and the second factor appears above a predetermined threshold, the processor (120) may assign a relatively high weight to the notification transmission condition. This is because it is important to detect potential system errors early by comparing with the actual inventory value rather than a simple sealing count. On the other hand, if a rapid increase in consumption is observed at a specific time in recent sealing patterns, or if a strong repeating pattern of specific days and times appears in past data, the processor (120) may increase the weight for the third factor so that the prediction algorithm prioritizes the corresponding factor in determining the order trigger. This is because delivery prediction and notifications must be carried out more aggressively for stores where the risk of depletion within a short period is high, even if the same error rate exists.

[0043] Additionally, the processor (120) can learn the frequency of inventory discrepancies that occurred in the same store in the past and the degree of correction of actual inventory differences after notification, and adjust the weights for each factor according to the characteristics of a specific store. For example, the processor (120) can improve automation efficiency by relatively increasing the weight for the second factor for stores where error rates frequently occurred or manager intervention was frequent in the past history, and conversely, increasing the prediction-based ordering weight of the third factor for stores with low error rates and stable patterns. In this way, the processor (120) can perform more precise inventory anomaly detection, order timing prediction, and delivery plan optimization by dynamically adjusting weights according to store-specific consumption characteristics and historical data, rather than simply determining conditions based on fixed thresholds.

[0044] FIGS. 4 to 8 are drawings illustrating other examples of a method for monitoring can container usage data implemented by the system (100) of FIG. 2.

[0045] Referring to FIG. 4, the processor (120) can assign a unique identifier (ID) to each of the multiple sealing heads of the can seamer.

[0046] Additionally, the processor (120) may be configured to independently detect the rotation cycle of each sealing head based on a unique identifier and record how many sealing operations each sealing head has performed.

[0047] At this time, the processor (120) can assign weights to multiple factors to evaluate the quality of usage data per sealing head and the impact on future supply plans, and optimize data interpretation and decision-making based on this.

[0048] Specifically, the processor (120) may set the key factors for weight calculation as: first factor: cumulative usage frequency per head, second factor: rate of abnormal rotation cycle occurrence, third factor: variability in usage patterns, and fourth factor: contribution to the consumption forecast of the head. Regarding the first factor, the processor (120) may assign a higher weight in the data analysis and forecasting algorithm because the higher the usage frequency of a specific head, the greater the likelihood that the records of that head represent the consumption trend within the entire system. For example, if head number 1 among three sealing heads in the same store performs more than 60% of the total sealing operations, the reliability of the demand forecast generated based on the data of that head is determined to be relatively high, so the system may use this as key reference data.

[0049] On the other hand, the second factor, the rotation cycle anomaly rate, refers to the frequency at which abnormal conditions, such as failure to reach the target angle, excessive torque fluctuations, and slip phenomena, are repeatedly detected compared to a normal sealing pattern, and data from heads with high such anomaly rates can reduce prediction reliability. Therefore, the processor (120) can adjust the impact of a head on the order trigger calculation by applying a warning weight to a head with a high anomaly rate or by lowering the data reliability score. This is because, in situations where a specific head is not operating normally, there is a high possibility that the number of sealing cycles will accumulate inaccurately, which can lead to a greater discrepancy between the actual inventory depletion amount and the data.

[0050] Additionally, the third factor, usage pattern variability, indicates whether each head maintains a constant usage pattern, and heads with high variability are likely to compromise the stability of the prediction algorithm. For example, if a specific head is used intensively during a specific period and then switches to a long-term non-use state, there is a risk of overestimation or underestimation when predicting future demand based on the head data. In such cases, the processor (120) can be designed to dynamically adjust weights according to variability to make the stable data source more reliable.

[0051] Finally, the fourth factor, the contribution to consumption prediction, can be calculated by analyzing how highly the usage pattern of a specific head correlates with the overall store sealing speed based on historical data. At this time, the processor (120) can reduce the error of the prediction model by setting a high weight for heads with high correlation, as they act as important variables in store-level prediction, and assigning a relatively low weight to heads with low correlation. In this way, the processor (120) calculates weights by synthesizing the operation history, anomaly signals, variability indicators, and correlation analysis results for each head, rather than relying on simple fixed rule-based analysis. This strengthens the adaptability of inventory management and automatic ordering logic, minimizes data bias in real-time operations, and thereby maximizes delivery efficiency and prediction accuracy.

[0052] Referring to FIG. 5, the processor (120) can transmit data on the number of can sealings and can usage collected from can seamers installed in multiple stores to the manager server (200).

[0053] Additionally, the processor (120) can predict the time when cans will run out and when cans need to be refilled for each store based on the time-based can usage data of multiple stores stored in the manager server (200).

[0054] Additionally, the processor (120) may be configured to generate a can ordering schedule for each store in accordance with the predicted time when can replenishment is needed, and to establish a can delivery plan to the store corresponding to the can ordering schedule. At this time, the processor (120) can maximize overall logistics efficiency and inventory stability by not simply generating orders based on usage, but dynamically assigning conditional weights to multiple factors and calculating ordering priorities and delivery plans based on said weights.

[0055] Specifically, the processor (120) may define the main factor group used for weight calculation as: first factor: speed of depletion per store, second factor: variability of recent consumption patterns, third factor: distance from logistics hubs, fourth factor: possibility of conflict with existing logistics schedules, and fifth factor: deviation in component depletion by SKU. Here, SKU (Stock Keeping Unit) is a minimum unit identifier that distinguishes individual product units, and each detailed component constituting a can, such as the can body, lid, and sealing material, can be managed as a separate SKU.

[0056] The first factor, the depletion rate, is calculated based on the number of times the sealing machine has operated in each store, for example, the depletion rate indicator = number of sealing operations in the last 24 hours ÷ standard inventory quantity. The processor (120) may assign a higher weight to stores with a high depletion rate because they have a greater risk of inventory shortage even with the same remaining inventory. For example, the processor (120) may be configured to apply a weight of 1.5 times to stores with a depletion rate that is 30% or more faster than the average.

[0057] The second factor, consumption pattern volatility, represents the stability of demand by store and can be determined based on the standard deviation of the consumption amount by day and time of day over the past four weeks. Since the higher the standard deviation, the greater the prediction uncertainty, the processor (120) can induce early ordering by applying additional weights. For example, the processor (120) defines the volatility index as = σ (weekly consumption amount) / average consumption amount, and can apply a weight of 1.2 to 1.4 times if it exceeds a certain threshold.

[0058] Since the third factor, the distance from the logistics hub, directly affects the delivery lead time, early ordering is required as the distance from the logistics center increases. For example, the processor (120) can set the weight to 1.3 times for stores located more than 100 km away and 1.6 times for stores located more than 200 km away. This method enables preemptive measures for outlying stores where emergency response is difficult.

[0059] The fourth factor, the possibility of a schedule conflict, is an indicator that evaluates the possibility of merging with an existing delivery route. If the possibility of merging is high, a weight can be increased to improve logistics efficiency. For example, the processor (120) may prioritize automatic merging by assigning a weight of 1.2 times when the difference in time and distance with a store delivery scheduled on the same route is below a certain threshold (e.g., within 15 minutes, within 5 km).

[0060] Finally, the fifth factor, the consumption deviation by SKU, reflects the difference in consumption rates for each component, such as can bodies and lids. For example, if the processor (120) has a lid consumption rate of 80% and a body can consumption rate of 60%, there is a need to prioritize ordering only the lids; therefore, a weight of 2 times can be applied to the lid SKU to enable selective ordering at the component level. This method can minimize the risk of the entire production being halted due to a shortage of specific components.

[0061] The processor (120) applies this weighting method to calculate the order priority score for each store in the form of Priority Score = (depletion rate index × α) + (volatility index × β) + (distance coefficient × γ) + (schedule merging coefficient × δ) + (SKU deviation coefficient × θ), and can automatically execute an order trigger based on this score. Through this, the present invention utilizes a dynamic adaptive algorithm instead of a static standard in order generation and delivery planning, thereby lowering the prediction failure rate, increasing logistics efficiency, and minimizing unnecessary urgent orders.

[0062] Referring to FIG. 6, the processor (120) can calculate a delivery route by integrating the can ordering schedules for multiple stores within a specific area to sequentially supply cans to nearby stores in a single delivery.

[0063] Additionally, the processor (120) can calculate a risk score for each store by considering the delivery distance to each store, the can inventory depletion curve of each store, and logistics availability.

[0064] And, the processor (120) can be configured to rearrange the delivery order to each store according to the calculated store-specific risk score. . At this time, the processor (120) can be designed to enable more realistic and adaptive optimization by applying a dynamic weight adjustment mechanism that assigns different weights to each factor depending on the situation, rather than simply using the risk score as a static value.

[0065] Specifically, the processor (120) can calculate the risk score R as follows: R = w_D × f(D) + w_U × g(U) + w_A × h(A)

[0066] Here, f(D) is a distance function that reflects delivery distance and traffic congestion, g(U) is an urgency function that considers the time when can stock is depleted by store and consumption deviation, and h(A) is a function that reflects the availability of logistics resources such as vehicle and driver schedules. The processor (120) may not set the weights w_D, w_U, and w_A applied to these functions as fixed values, but may dynamically adjust them according to specific situational conditions. For example, if a highway congestion occurs during a specific time period or if a decrease in driving speed is expected due to bad weather such as rain or snow, the processor (120) may increase the importance of the distance factor by increasing w_D by more than 1.5 times compared to the existing value. This is because if traffic conditions worsen, travel time increases even for the same distance, and the efficiency of the entire delivery route may decrease rapidly. On the other hand, if the can inventory depletion curve of a specific store drops sharply and the expected depletion time approaches within 6 hours, the processor (120) can increase w_U to more than double the existing value to reflect the store at risk of inventory depletion as the top of the route. This method is a reasonable approach that reflects the time element in determining urgency and can function as a key mechanism to prevent supply chain disruption in advance. Additionally, if the number of dispatchable vehicles within the logistics center is limited or the driver schedule is saturated, the processor (120) can increase w_A to reflect the constraints on logistics availability. For example, if the processor (120) cannot deploy additional vehicles during a specific time period or delivery personnel are limited, a strategy to maximize logistics resource efficiency is essential, so the feasibility of route planning can be secured by strengthening the weight of w_A.

[0067] Furthermore, the processor (120) may apply a non-linear adjustment coefficient to both w_D and w_U when complex conditions occur simultaneously, for example, when traffic congestion and the risk of stock depletion increase at the same time. In this case, the adjustment function may be designed in the form of, for example, α = exp(β·C), so that when the risk C (congestion severity, stock urgency) exceeds a certain threshold, the weight increases exponentially. Since this dynamic adjustment reflects actual supply chain risks more sensitively than simple linear weight adjustment, real-world adaptability can be greatly improved.

[0068] Additionally, the processor (120) can calculate the supply priority by normalizing the risk score by the consumption rate and the delivery lead time. Specifically, the processor (120) may apply the following formula: Supply Priority = (Risk Score × Consumption Rate) / Delivery Lead Time

[0069] This formula is designed to prioritize stores with faster consumption rates and longer delivery lead times, even if they have the same risk score. For example, even if two stores maintain the same inventory levels, if Store A consumes 50 cans per hour and Store B consumes 10 cans per hour, Store A faces a much higher risk of running out of stock, so its supply priority may be raised. Additionally, since stores with long delivery lead times are more likely to experience replenishment delays after route adjustments, including this factor in the denominator allows for a more precise reflection of urgency. This approach can contribute to enhancing supply chain stability and minimizing revenue losses caused by inventory disruptions, compared to simple risk-based prioritization.

[0070] The processor (120) can perform such weight adjustment and supply priority calculation processes based on real-time simulation. For example, the processor (120) can integrate GPS traffic data, weather forecasts, store POS (Point of Sale) data, and sealing count information collected from a can seamer to predict changes in consumption patterns by store, and by virtually executing various route scenarios, compare and evaluate the estimated total time required, fuel costs, and the need for urgent delivery for each route. In particular, when adjusting the order quantity by SKU unit, the processor (120) can include the supply risk of a specific SKU (Stock Keeping Unit) as an additional factor, and apply additional penalty weights to stores where the consumption rate of a specific SKU is abnormally high. For example, the processor (120) can apply an SKU-specific coefficient in the risk score calculation because in stores where 500ml can SKUs are rapidly consumed due to promotions, the depletion of the stock of that SKU has a significant impact on overall sales. Through this, the processor (120) can establish a fine supply strategy at the SKU level.

[0071] Furthermore, the processor (120) may combine a machine learning-based prediction model with the above weight adjustment process so that a model learned from past delivery history and real-time operational data can automatically set initial values ​​for weights for each factor. For example, the processor (120) may operate by pre-allocating the weight of the distance factor for days and times when traffic congestion is predicted to be high, using a model that has learned delivery delay patterns in a specific region over the past three months. This enables dynamic weight optimization that is much more sophisticated than manual fixed rules, and consequently minimizes delivery delay rates and reduces the frequency of stock shortages. This design can increase the reliability of supply chain management and improve the customer experience.

[0072] Consequently, this embodiment has technical features that, rather than simply statically evaluating store-specific risks, it proactively manages supply chain risks and maximizes the efficiency of logistics resources through dynamic weighting of multiple factors and real-time priority realignment. The processor (120) can achieve two goals simultaneously: reducing operating costs and improving service levels by calculating the optimal delivery route and order by reflecting the complex interaction of various conditions and variables.

[0073] Referring to FIG. 7, the processor (120) can record can usage data by separating the consumption rate into component units including the can body and lid of the can.

[0074] Additionally, the processor (120) can independently analyze the inventory level and depletion rate for each component unit.

[0075] And, the processor (120) may be configured to independently set a priority supply trigger for a specific component when the consumption rate of that component exceeds a preset threshold.

[0076] Additionally, the processor (120) may not evaluate all components with equal importance during the trigger setting process, but may dynamically assign conditional weights by reflecting the characteristics and supply risks of each component. Specifically, even if the consumption rate of the can body and the lid is the same, the processor (120) may assign a higher weight to the lid in environments where the manufacturing lead time for the lid is relatively long and it is difficult to secure specific metal materials. This is because there is a high probability that a bottleneck will occur in which the entire can sealing process is halted if a shortage of lids occurs, and this risk can have a greater impact on operational efficiency than the simple consumption rate.

[0077] Additionally, the processor (120) can adjust the weights by comprehensively evaluating external factors such as the degree of supplier diversification for each component, the possibility of urgent procurement, and the time required for inventory replenishment. For example, if the can body can be replaced and supplied from multiple suppliers within a short period, but the lid of a specific specification can only be procured from a single certified supplier, the processor (120) may set the weight of the lid relatively higher even under the same consumption rate conditions. This is because procurement constraints are highly likely to lead to unexpected production disruptions in actual supply chain risks, so the processor (120) can assign weights by reflecting these external constraint factors.

[0078] Furthermore, the processor (120) can analyze real-time equipment operation data generated at the production site, such as the operation pattern of a can seamer and motor rotation signal, and if it detects a situation where the usage speed of a specific component increases abnormally, it can further increase the weight for that component. This is because early intervention is required to secure a stable supply, as a sudden increase in consumption tends to continue for a certain period in the future.

[0079] In this way, the processor (120) can optimize an algorithm that sets priority supply triggers by comprehensively evaluating the consumption rate, lead time, procurement difficulty, supplier diversification, and real-time usage patterns of each component to assign differential weights. Through this, the risk of production stoppage due to inventory shortages can be prevented in advance, and the effect of maximizing supply chain efficiency can be achieved. Consequently, unlike the existing method of setting thresholds based on fixed criteria, the present invention enables more sophisticated priority supply judgments through a dynamic weighting method based on multidimensional factors.

[0080] Referring to FIG. 8, the processor (120) can calculate the pressure index using the formula (average speed of can seamer sealing × deviation from recent can consumption) ÷ current remaining amount × delivery distance coefficient by using location information of each store and recent can usage data.

[0081] Additionally, the processor (120) may be configured to calculate and sort the calculated consumption pressure index for a number of stores, and then set priority supply triggers for the top N stores.

[0082] At this time, the processor (120) may not assign equal importance to each element, but may dynamically adjust the weight of each factor by considering the characteristics of the store, geographical conditions, and supply chain constraints. For example, if the store is located on the outskirts of the city and urgent delivery is difficult, the processor (120) may assign a higher weight to the delivery distance coefficient even if it has the same consumption pressure index. This is because the risk is high for stores where urgent resupply is physically impossible, so delivery distance and accessibility can serve as important judgment factors in determining the actual supply priority.

[0083] Additionally, the processor (120) may assign a high weight to the rate of change in consumption patterns when the recent consumption deviation at a specific store increases rapidly. This is because stores with large recent consumption deviations have high prediction uncertainty and supply plans are not stabilized, so it is necessary to prevent sales suspension due to inventory depletion through preemptive measures. Therefore, the processor (120) can design an algorithm such that even if they have the same consumption pressure index, stores with high consumption pattern volatility are more likely to be included as priority supply targets.

[0084] Additionally, the processor (120) may assign differential weights based on the sales contribution of each store or the diversity of SKUs (Stock Keeping Units). For example, the processor (120) may apply a relatively high weight to core stores with a high sales proportion to minimize the risk of sales loss, even if they have the same consumption pressure index. Conversely, the processor (120) may assign a low weight to small stores that handle only a single SKU, considering logistics efficiency rather than urgency. This is a reasonable method for setting strategic priorities to maximize the overall profit of the company, given the limited supply chain resources.

[0085] Furthermore, the processor (120) can readjust the weighting for the delivery distance coefficient by additionally reflecting external data such as real-time traffic conditions and weather conditions. For example, even if two stores are located at the same distance, if there is a high possibility of delivery delay due to traffic congestion or heavy rain on the access route to a specific store, the processor (120) can reduce the risk of stockout due to logistics delays by increasing the delivery distance weighting for that store.

[0086] In this way, the processor (120) does not stop at simply calculating the consumption pressure index, but analyzes the influence of each factor in real time and dynamically assigns weights, thereby calculating a priority supply order optimized for the situation of each store. Through this, the stability and operational efficiency of the supply chain can be maximized, and ultimately, the present invention can significantly improve prediction accuracy and supply reliability compared to existing methods through a weight adjustment mechanism based on real-time conditions rather than static criteria.

[0087] In one embodiment, the processor (120) can collect time-based operation history by analyzing the usage pattern of the can seamer provided in the store.

[0088] Additionally, the processor (120) can cluster can consumption periodicity and peak intervals based on the number of can sealing cycles of can seamers per store by time period, reflecting can consumption trends per store.

[0089] And, the processor (120) may be configured to predict whether pre-delivery of cans is required on a specific day and time based on time-time demand information derived from clustered can consumption periodicity and peak section information, and to determine whether pre-delivery of cans corresponding to a specific time-time of the next day or next week is required.

[0090] In one embodiment, the processor (120) can collect operation history by time period by analyzing the usage pattern of the can seamer installed in the store. Specifically, the processor (120) collects sealing operation event logs from the seamer device for each store in real time and, based on this, can calculate the number of can sealings by time period, the average work interval, and the operation rate for each device in detail. Through this analysis, the processor (120) can accurately reflect the seamer usage characteristics that vary from store to store, and can detect intensive can usage patterns occurring on specific days or at specific times with high precision.

[0091] Additionally, the processor (120) can learn can consumption patterns in a multidimensional way by combining can usage data accumulated over a certain period and delivery history data to reflect can consumption trends by store. For example, the processor (120) can determine that demand variability may differ depending on external variable conditions, even among stores with the same number of sealings, by performing a multi-variable-based analysis combined with external environmental factors (store location, changes in nearby pedestrian traffic, temperature, etc.) as well as the number of can sealings by day of the week and time of day. Through this, the processor (120) can build a sophisticated consumption trend model that reflects store-specific characteristics beyond simple average usage analysis.

[0092] Furthermore, the processor (120) may be configured to dynamically adjust weights for each factor according to data conditions to increase the reliability of demand forecasting by time period by integrating sealing machine usage patterns and can consumption trends by store. For example, if a repeated sharp increase in can consumption during weekday evening hours is observed in the past sealing machine usage logs of a specific store, the processor (120) may assign a relatively higher weight to the sealing data by time period of that store. On the other hand, if the analysis of long-term consumption trends reveals that seasonal factors (e.g., a surge in beverage consumption during the summer) have a decisive influence on the demand pattern of that store, the processor (120) may be designed to assign a higher weight to the long-term consumption trend factors. In this way, the processor (120) can avoid rigid prediction models centered on a single factor and establish an adaptive weight-based prediction system that reflects consumption characteristics by store and time period.

[0093] Additionally, the processor (120) may utilize a clustering technique to identify the periodicity and peak intervals of can consumption. Specifically, the processor (120) may set the number of can sealings per time period as an input variable and derive periodicity by clustering time periods that exhibit the same or similar consumption patterns. For example, if a consumption peak occurs repeatedly in the afternoon of every Thursday at a specific store, the processor (120) may classify that time period as a separate high-demand segment and automatically increase the priority of pre-logistics response for that segment. On the other hand, for a store with irregular equipment usage patterns and large consumption deviations, the processor (120) may ensure prediction reliability by adjusting weights by prioritizing long-term accumulated consumption data over sealing logs in the clustering analysis.

[0094] In this way, the processor (120) according to the present invention can implement a demand forecasting model that finely reflects the characteristics of each store by performing an integrated analysis by linking store-specific shimmer usage patterns, can consumption trends, and periodic information by time period, and dynamically adjusting the weights assigned to each factor according to data conditions. Through this, the processor (120) can more accurately determine the need for pre-delivery of cans on specific days and times and can stably provide base data for delivery schedule optimization.

[0095] In one embodiment, the processor (120) can precisely analyze can consumption patterns based on the time-based operation history of can sealers for each store and derive time-based demand clusters. In particular, the processor (120) can subdivide the consumption patterns for each store into time units by utilizing sealing count logs and can usage data accumulated over a certain period, and can automatically identify periodicity and peak sections based on this. In this clustering process, operating hours for each store, characteristics of nearby commercial areas, and patterns of specific days of the week or holidays may be reflected, and the processor (120) can quantify the demand intensity for each cluster by comprehensively considering these various factors.

[0096] Additionally, the processor (120) can dynamically adjust weights according to the demand intensity of each cluster. For example, if a specific store shows a peak consumption period during lunch or dinner, the processor (120) can maximize future prediction accuracy by assigning a high weight to the demand cluster for that time period. Conversely, if the processor (120) has high variability in past data or irregular patterns on a specific day or time period, the weight for that period can be set relatively low. This adaptive weighting strategy can reflect changes in demand in the real environment more precisely than a simple prediction model that relies on fixed criteria, and the processor (120) can increase the stability of can consumption prediction by time period through this.

[0097] The processor (120) can also set weights in a multi-layered manner by simultaneously considering long-term trends and short-term volatility of consumption patterns. For example, if a specific store shows a sudden increase in event-driven sales over the past two weeks, the processor (120) can prioritize giving high weights to peak periods appearing in short-term data to reflect the immediate need for pre-delivery. On the other hand, for time periods showing consistently stable consumption in monthly or quarterly aggregated data, the processor (120) can set high weights based on long-term trends so that the prediction model is not excessively affected by short-term volatility. In this way, the processor (120) can perform a multi-layered weighting strategy that simultaneously considers short-term events and long-term patterns.

[0098] Meanwhile, the processor (120) can reinforce the reliability of the demand forecasting model by considering external environmental factors. For example, the temperature of a specific region, weather changes, and events such as local festivals can have a direct impact on can consumption patterns, and the processor (120) can analyze these environmental factors and automatically adjust the demand cluster weights for the corresponding period. In particular, if short-term volatility due to external factors is expected, the processor (120) can set the sensitivity of the forecasting model high to more quickly reflect the need for advance delivery during that period. Conversely, in periods where external factors have relatively little influence, the processor (120) can maintain the stability of the forecasting model by prioritizing the application of internal factor weights based on historical data.

[0099] Based on such prediction results, the processor (120) can calculate priorities for pre-delivery decision-making. Specifically, by combining clustered time-time demand information and weighted consumption pattern analysis results, the likelihood of a shortage of can stock at a specific store or time period can be quantitatively evaluated. Based on these evaluation values, the processor (120) scores the need for pre-delivery for each store and can automatically set pre-supply triggers for stores with high priority.

[0100] Additionally, the processor (120) can also consider logistics resource constraints during the pre-delivery decision-making process. For example, the availability of delivery vehicles, transit times per stop, and inventory levels in logistics warehouses can act as major factors limiting the feasibility of pre-delivery execution. By reflecting these constraints in real time, the processor (120) can derive an optimal delivery scenario with a high probability of actual execution, going beyond simply setting priorities based on demand forecasting. This allows the system to maximize logistics efficiency while preventing sales losses caused by a shortage of can inventory at each store.

[0101] Consequently, the processor (120) can construct a pre-delivery decision model by integratively considering four axes: clustering-based peak demand analysis, adaptive weighting, environmental factor correction, and logistics constraint reflection. Unlike static models that rely simply on past data, this configuration can reflect real-time environmental changes and the interaction between complex factors, thereby maximizing both prediction accuracy and effectiveness simultaneously.

[0102] The processor (120) learns can consumption patterns based on time-of-day clustering by store and can dynamically assign weights according to the demand characteristics of each time period. For example, store A may show a tendency for can consumption to be concentrated during the lunch peak time, while store B may show a tendency for consumption to surge on weekend evenings. In this case, the processor (120) can reflect the consumption patterns specialized for each store in the prediction model by assigning a relatively high weight to the lunch time cluster of store A and a higher weight to the weekend evening cluster of store B. This dynamic weighting method can dramatically improve prediction accuracy compared to a simple average-based model, as it can consider different operational characteristics and consumer behaviors for each store.

[0103] Additionally, the processor (120) may set different weights based on the reliability of historical data. For example, if the processor (120) finds that the hourly operating patterns and consumption of can seamers in the data of a specific store over the past 12 months are very consistent, the processor (120) may give high weight to the historical data when predicting future demand for that store. Conversely, if there is a significant difference between the historical data and the current pattern due to recent store remodeling or changes in operating hours, the processor (120) may adjust the weight of the historical data to be lowered and reflect real-time sensor data and recent consumption deviations relatively more. Through this, the processor (120) can minimize discrepancies between historical data and real-time data and configure a prediction model that operates adaptively even in a changing environment.

[0104] Even in the pre-delivery decision-making stage, the processor (120) can perform weight-based optimization considering multiple factors. For example, when a time of expected surge in demand arrives at a specific store, the processor (120) can determine the supply priority by simultaneously considering the store's inventory level, consumption trends, distance from the logistics center, and the delivery schedules of nearby stores. In this process, the processor (120) may assign a high weight to stores close to the inventory threshold to minimize the risk of inventory shortage, and may also apply a high weight to stores with large consumption variability for a conservative response due to the high uncertainty in demand forecasting. On the other hand, the processor (120) can optimize overall logistics efficiency by assigning a relatively low priority to stores with sufficient inventory or low consumption variability.

[0105] Furthermore, the processor (120) can simulate delivery routes for multiple stores within the same area to automatically determine which store will trigger a pre-delivery. For example, if three stores within a specific area are expected to experience a demand peak at the same time, the processor (120) can automatically calculate the most efficient delivery route by comprehensively analyzing the consumption patterns and inventory levels of each store, as well as existing delivery routes and logistics vehicle availability. Through this, logistics costs can be minimized by integrating replenishment orders from multiple stores into a single vehicle route or by merging additional deliveries using a recycle time window.

[0106] As a result, the processor (120) can simultaneously ensure accuracy and efficiency in pre-delivery decision-making by adjusting the weights for various factors on a situational basis, going beyond simple demand forecasting. Since this configuration can reflect rapidly changing store-specific consumption patterns, regional logistics conditions, and real-time sensor data variability, it can contribute to minimizing forecasting errors and implementing a stable inventory management system.

[0107] In one embodiment, the processor (120) can identify a Recycle Time Window within an existing can delivery plan and evaluate whether additional deliveries can be merged into the existing can delivery plan.

[0108] Here, the Recycle Time Window refers to the time window during which the same delivery vehicle can revisit a store on a specific route within a certain period. In other words, it refers to the buffer time that occurs before a vehicle departs a store and moves to the next route under an existing delivery schedule, or a temporal and spatial window that allows for additional visits to other stores within a certain range along the route. By utilizing this Recycle Time Window, the existing delivery network can be efficiently optimized without setting up new routes or deploying separate vehicles for additional deliveries.

[0109] For example, if it is determined that urgent replenishment delivery is necessary because can consumption exceeding expectations occurs at a specific store, the processor (120) comprehensively analyzes the location, consumption volume, and remaining inventory level of the store and then compares the route and schedule of the existing delivery vehicle. In this process, vehicles with a sufficiently secured recycle time window are selected as priority candidates, and it can be evaluated whether merging replenishment delivery into the existing route is highly effective for overall transportation efficiency and cost-effectiveness.

[0110] Additionally, the processor (120) can determine the need for replenishment delivery by collecting the can consumption amount and the lid consumption rate from the can usage data.

[0111] And, when a condition requiring a replacement delivery arises, the processor (120) can consider the possibility of merging the replacement delivery into the existing can delivery plan by taking into account the time, distance, and stock availability of the existing delivery route.

[0112] Additionally, the processor (120) may be configured to perform a simulation on the delivery routes of multiple stores to calculate which delivery route is optimal for merging each replenishment delivery.

[0113] In one embodiment, the processor (120) can determine whether replenishment delivery is necessary by collecting real-time can usage data and lid consumption rates from can sealers and various IoT sensors installed in the store. To this end, the processor (120) dynamically analyzes recent trends in can consumption by store, remaining inventory levels, and estimated depletion times, and considers that conditions requiring replenishment delivery have occurred when specific conditions are satisfied. For example, if the hourly can consumption of the sealing machine at a specific store increases rapidly, causing the estimated depletion time to be brought forward faster than the existing plan, the processor (120) can set this as a candidate for replenishment delivery.

[0114] Additionally, the processor (120) does not determine whether to replenish delivery based solely on an increase in consumption, but also considers the connection with the Recycle Time Window. That is, even if a specific store satisfies the requirements for replenishment delivery, the processor (120) first evaluates the possibility of merging it into an existing route only if the store is located within the Recycle Time Window of the delivery vehicle currently in operation. In this case, if the Recycle Time Window is insufficient, the processor (120) can simulate a separate emergency delivery route or a plan to link it to future delivery.

[0115] The weighting method plays a key role in this decision-making process. The processor (120) compares multiple factors to evaluate the need for replenishment delivery and dynamically assigns weights to each factor to determine whether to merge them. For example, if there is a high correlation between the consumption pattern by store and the lid consumption rate, the processor (120) may assign a higher weight to the lid consumption rate factor. This is because there is a possibility that the inventory depletion cycles of cans and lids may appear unbalanced, and it is intended to reflect the risk that the depletion rate of a specific component could halt the entire production line.

[0116] In addition, distance factors along the delivery route and the remaining load capacity of the existing vehicle are also utilized as key elements in the weight evaluation. For example, if there are two stores that satisfy the same replenishment delivery requirements, the processor (120) assigns a relatively higher weight to the store that is closer to the vehicle's current route to increase the priority of merging. Conversely, the processor (120) may limit the possibility of merging by assigning a lower weight if the distance from the route is far or the remaining load capacity is insufficient.

[0117] In this way, the processor (120) can utilize sensor-based real-time data to preemptively determine the need for replenishment delivery and apply an adaptive weighting strategy that takes into account the recycle time window to optimize the possibility of merging.

[0118] In one embodiment, the processor (120) can perform a real-time simulation on the delivery routes of multiple stores to calculate an optimal strategy for merging replenishment delivery requests into existing delivery routes. To this end, the processor (120) comprehensively analyzes various factors such as the estimated arrival time, delivery interval, loading limit per vehicle, remaining inventory amount, and can usage rate of each store for each route included in the existing delivery plan.

[0119] For example, if a specific store experiences a rapid surge in can consumption over a short period and requires replenishment delivery, the processor (120) first checks whether the store falls within the Recycle Time Window range of the existing delivery vehicle. If the store falls within the Recycle Time Window, the processor (120) may first consider merging the replenishment delivery into the vehicle's route. On the other hand, for stores outside the window, the processor (120) may create a separate emergency delivery route or re-evaluate the possibility of merging with another vehicle.

[0120] The AI-based weighted optimization model plays a key role in this simulation process. The processor (120) quantifies variables such as store-specific consumption patterns, inventory depletion curves, delivery distances, and vehicle loading capacity, assigns weights to each, and calculates the overall merging probability as a predicted score. For example, even if two stores simultaneously satisfy the replenishment delivery requirement under the same conditions, the processor (120) assigns a relatively higher weight to the store with a smaller difference in distance from the vehicle's current route to increase the merging priority. This is evaluated as a reasonable choice that can reduce logistics costs by minimizing the increase in driving distance caused by additional deliveries.

[0121] In addition, the processor (120) can incorporate component-unit inventory data, such as the individual consumption rates of can bodies and lids, into the merge optimization, rather than simply considering distance and time factors. For example, if a specific store has an abnormally high lid consumption rate and a high risk of production process interruption, the processor (120) can evaluate the need for replenishment delivery for that store as higher than for other stores and preemptively raise the merge priority for lid supply. This has the advantage of more precisely reflecting the actual operational stability of the store compared to the existing method of determining mergers based on simple can consumption.

[0122] In this way, the processor (120) can apply a dynamic algorithm that optimally merges replenishment deliveries based on real-time data while maintaining the existing delivery plan. Through this, an intelligent can delivery management system can be implemented that maximizes logistics efficiency, ensures the timeliness of replenishment deliveries, and minimizes operating costs.

[0123] FIG. 9 is a flowchart illustrating a method for monitoring can container usage data according to an embodiment of the present invention.

[0124] Referring to FIG. 9, a method for monitoring can container usage data according to one embodiment of the present invention may include the following steps S1 to S4. However, it is not limited thereto, and other general steps may be further included in the method for monitoring can container usage data.

[0125] The above-described can container usage data monitoring method may consist of steps processed in a time-series manner in a system (100) that performs the above-described can container usage data monitoring method. Therefore, even if details are omitted below, the description of the system (100) that performs the can container usage data monitoring method above may be equally applicable to the can container usage data monitoring method described later.

[0126] In the above S1 step, the processor (120) can receive a motor rotation signal from an encoder linked to the motor of a can seamer for sealing a can.

[0127] In the above S2 step, the processor (120) can analyze the motor rotation signal to detect a rotation cycle corresponding to the operation of sealing the can.

[0128] In the above S3 step, the processor (120) can count and record the number of can sealing cycles for each detected rotation cycle.

[0129] In the above S4 step, the processor (120) can transmit accumulated can usage data to the administrator server (200) based on the counting of the number of can sealings.

[0130] Meanwhile, the can container usage data monitoring method performed by the aforementioned system (100) may be implemented in the form of a computer program stored on a computer-readable storage medium. That is, the computer program may include instructions for implementing the can container usage data monitoring method, and the instructions of the computer program may be stored on a computer-readable storage medium. The computer program may include a mobile application.

[0131] For example, computer-readable storage media may include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute computer program instructions, such as ROM, RAM, and flash memory. Computer program instructions may include machine code generated by a compiler and high-level language code that can be executed by a computer using an interpreter, etc.

[0132] As described above, although the present invention has been explained by limited embodiments and drawings, the present invention is not limited thereto, and it is obvious that various modifications and variations are possible within the scope of the technical spirit of the present invention and the equivalent scope of the claims described below by those skilled in the art to which the present invention belongs. Explanation of the symbols

[0133] 100 : System 110 : Memory 120 : Processor 200 : Administrator Server

Claims

Claim 1 Memory configured to store instructions; And by executing the above commands: a processor configured to receive a motor rotation signal from an encoder linked to a motor of a can seamer for sealing cans, analyze the motor rotation signal to detect a rotation cycle corresponding to the operation of sealing cans, count the number of can sealings for each detected rotation cycle and record it in the memory, and transmit accumulated can usage data according to the counting of can sealings to a manager server; the processor is configured to transmit the number of can sealings and can usage data collected from can seamers equipped in multiple stores to a manager server, predict the future time when cans are depleted and the time when cans need to be refilled for each store based on the time-based can usage data of multiple stores stored in the manager server, generate a can ordering schedule for each store according to the predicted time when cans need to be refilled, and establish a can delivery plan to the store corresponding to the can ordering schedule, calculate a delivery route to sequentially supply cans to nearby stores by integrating can ordering schedules for multiple stores within a specific area, and the delivery distance to each store, the can inventory depletion of each store A can container usage data monitoring system configured to calculate a risk score for each store by considering curves and logistics availability, and to reorder the delivery order to each store according to the calculated risk score for each store, wherein the risk score for each store is determined by the following mathematical formula: [Mathematical Formula] R = w_D × f(D) + w_U × g(U) + w_A × h(A) (where R is the risk score for each store, f(D) is a distance function reflecting delivery distance and traffic congestion, g(U) is an urgency function considering the time of can inventory depletion and consumption deviation for each store, and h(A) is a function reflecting the availability of logistics resources such as vehicle and driver schedules). The processor is configured not to set the weights w_D, w_U, and w_A applied to the above mathematical formula as fixed values, but to dynamically adjust them according to specific situational conditions. Claim 2 A can container usage data monitoring system according to claim 1, wherein the processor determines whether there is a match by comparing the number of can sealings with the actual can inventory quantity, transmits a notification to the administrator server when the cumulative error between the actual can inventory value and the number of can sealings counted through rotation cycle detection exceeds a predetermined standard, stores information on the consumption time and quantity of cans in the memory, and sets a delivery trigger for the store by predicting the delivery cycle and consumption quantity based on the stored data. Claim 3 A can container usage data monitoring system according to claim 1, wherein the processor is configured to assign a unique identifier (ID) to each of a plurality of sealing heads of a can seamer, and to independently detect the rotation cycle of each sealing head based on the unique identifier to record how many sealing operations a sealing head has performed. Claim 4 delete Claim 5 delete Claim 6 A can container usage data monitoring system according to claim 1, wherein the processor is configured to record can usage data by separating the consumption rate into component units including the can body and lid of the can, independently analyze the inventory level and depletion rate for each component unit, and independently set a priority supply trigger for the corresponding component when the consumption rate of a specific component exceeds a preset threshold. Claim 7 A method for monitoring can container usage data performed by a can container usage data monitoring system, comprising: receiving a motor rotation signal from an encoder linked to a motor of a can seamer for sealing a can; analyzing the motor rotation signal to detect a rotation cycle corresponding to a can sealing operation; and counting and recording the number of can sealing operations for each detected rotation cycle. The method includes the step of transmitting accumulated can usage data to an administrator server based on the counting of can sealing counts. The can container usage data monitoring system is configured to transmit can sealing counts and can usage data collected from can sealers installed in multiple stores to an administrator server, predict the future can depletion time and the time when can replenishment is needed for each store based on the time-based can usage data of multiple stores stored in the administrator server, generate a can ordering schedule for each store according to the predicted time when can replenishment is needed, and establish a can delivery plan to the store corresponding to the can ordering schedule. It is configured to calculate a delivery route to sequentially supply cans to nearby stores in a single delivery by integrating can ordering schedules for multiple stores within a specific area, calculate a risk score for each store considering the delivery distance to each store, the can inventory depletion curve of each store, and logistics availability, and rearrange the delivery order to each store according to the calculated risk score for each store. The risk score for each store is determined by the following mathematical formula: [Mathematical Formula] R = w_D × f(D) + w_U × g(U) + w_A × h(A)(where R is the risk score per store, f(D) is a distance function reflecting delivery distance and traffic congestion, g(U) is an urgency function considering the time of can inventory depletion and consumption deviation per store, and h(A) is a function reflecting the availability of logistics resources such as vehicle and driver schedules.A can container usage data monitoring method configured such that the above can container usage data monitoring system does not set the weights w_D, w_U, and w_A applied to the above mathematical formula as fixed values, but dynamically adjusts them according to specific situational conditions.

Images