CN111008857A

CN111008857A - System and method for consumption prediction of consumer packaged goods

Info

Publication number: CN111008857A
Application number: CN201910951175.3A
Authority: CN
Inventors: L.高
Original assignee: Campbell AG
Current assignee: Campbell AG
Priority date: 2018-10-08
Filing date: 2019-10-08
Publication date: 2020-04-14

Abstract

A system and method uses low cost and power efficient motion sensors integrated into (or attached to) various general purpose Consumer Packaged Goods (CPG) containers or packages to learn and identify the posture(s) of a consumer during a consumption event. Sensed input from motion sensors (e.g., accelerometers, compasses, and/or gyroscopes) is provided to a consumption prediction model/program that maps past consumption events to determine the percentage of the quantity of the good(s) remaining in the CPG. The consumption prediction model may take into account the habitual consumption rates and history of the individual consumer. Based on the input(s) from the motion sensor, the consumption prediction model may be updated continuously or at certain time intervals. The consumption prediction model can then be used to predict the remaining amount of the tracked CPG, allowing the system to further handle the replenishment inventory with the retailer's online system.

Description

System and method for consumption prediction of consumer packaged goods

RELATED APPLICATIONS

The present application claims the priority of U.S. provisional patent application No. 62/731,191 entitled "SYSTEM AND METHOD FOR making a plastic article with a consistency of performance RATE AND PREDICTING degree performance and No. 62/742,490 entitled" SYSTEMAND METHOD FOR DEPLETION PREDICTION OF CONSUMER-package group USING polyester article with a consistency of performance sensos "filed on 14.9.2018, which is hereby incorporated by reference in its entirety.

Technical Field

The present disclosure relates to systems and methods for determining consumption of a Consumer-packaged good (CPG).

Background

Consumer packaged goods ("CPG") retailers are continually innovating ways in which they interact with consumers that today have more options than ever before. New technologies and novel enhancements to the "buying experience" help retailers and brands retain consumers and drive sales. In 2015, an electronic home retailer introduced a battery-powered, Wi-Fi based device that allowed consumers to order specific products from a brand by pressing a button. It is rapidly becoming a ubiquitous and channel-wide model of retailer and brand quick-love purchases, a satisfying consumer experience, and increased brand loyalty, all of which result in a substantial increase in sales. Many consumers are excited about the experience and transaction that doubles in 2017 over the last year through this channel.

Although the aforementioned facts have proven simple and fast importance in channel-wide commerce, this type of solution has some limitations. One of the obstacles is the cognitive burden of initiating purchases. Although programmatically simple, the referenced solution requires the consumer to make decisions and take explicit action to purchase. This greatly limits its generality to ultimately support a consistently available, consumer-optimized channel between brands and consumers. For example, consumers cannot use the aforementioned types of solutions to establish a reorder relationship with brands and retailers because the system does not know when to replenish before the consumer realizes that a transaction needs to be initiated.

Existing methods, such as object and pose recognition using cameras, are convenient in principle, but present serious reliability challenges in practice. Services have been introduced that allow smart appliance vendors to build their own consumption meters. However, this approach requires each vendor to develop its own specific monitoring system for quantity evaluation, and thus it is not a system-on-key model that can be easily transformed and applied to many other CPGs and vendors.

The e-commerce department is growing rapidly. With the success of the above types of product/service offerings, more and more retailers and brands are looking to establish a channel-wide solution for business communication with their customers to improve loyalty, convenience, and sales. Today, there is a lack of a simple and reliable method for allowing consumers to automatically replenish inventory as the product is depleted. A time-based only reorder-policy will not yield the best results because the elapsed time of old goods varies for each shopping cycle and over time. Consumers want to consume fresh, newly produced goods, but they worry about their supply gaps and remember the cognitive burden of reordering on time poses a challenge.

Disclosure of Invention

The present disclosure addresses the problems of known solutions, such as supply gaps and having to remember and take action on an ongoing basis to reorder, by implementing systems and methods for estimating consumption rates of Consumer Packaged Goods (CPGs) and predicting remaining amounts of CPGs using motion sensors. The disclosed systems and methods use low cost and power efficient motion sensors that require minimal footprint on the PCB and can be easily integrated into (or attached to) various common CPG containers or packaging. The sensors are programmed to learn and recognize the gesture(s) of the consumer during a consumption event using sensory input from motion sensors, such as accelerometers, compasses, and/or gyroscopes. In accordance with the present disclosure, a mathematical consumption prediction model may be implemented that maps past consumption events to determine the percentage of the amount of the good(s) remaining. The consumption prediction model may take into account the habitual consumption rates and history of the individual consumer. Based on the sensed input(s) from the motion sensor, the consumption prediction model may be updated continuously or at certain time intervals. The model can then be used to predict the remaining amount of the tracked CPG, allowing the system to further handle the replenishment inventory with the retailer's online system. The sensors may also include environmental sensors as described in commonly owned and co-pending U.S. patent application serial No. 16/137,835, which is incorporated by reference herein in its entirety. Systems and methods according to the present disclosure may also use data from environmental sensors to estimate the expiration time of the good(s).

The disclosed systems and methods provide an effective and affordable solution for consumers that alerts consumers before their favorite items are exhausted and that enables intelligent replenishment orders to be placed with the CPG suppliers to deliver new replenishment items to the consumers without gaps in their supply.

The foregoing and other advantages of the present disclosure will become more apparent to those skilled in the art from the following detailed description of the present disclosure, which is shown and described by way of illustration. It is to be understood that the disclosed subject matter is capable of other and different embodiments and its details are capable of modifications in various respects.

Drawings

Embodiments of the apparatus, system, and method are illustrated in the accompanying drawings, which are meant to be exemplary and not limiting, wherein like reference numerals are intended to refer to the same or corresponding parts, and wherein:

FIG. 1 illustrates components of the disclosed system for monitoring consumption events in accordance with the present disclosure;

fig. 2 illustrates a process flow diagram of an exemplary embodiment of a system for monitoring CPG targets for consumption events according to the present disclosure;

FIG. 3 illustrates a process flow diagram for detecting a consumption event using another embodiment of the system according to the present disclosure;

FIG. 4 illustrates a process flow diagram for monitoring CPG targets and determining when CPG targets will require replenishment in accordance with the present disclosure;

fig. 5 illustrates a process flow diagram for monitoring the quantity and quality of CPG targets in accordance with the present disclosure.

Detailed Description

The detailed description of aspects of the disclosure set forth herein makes reference to the accompanying drawings that show, by way of illustration, various embodiments. Although these various embodiments are described in sufficient detail to enable those skilled in the art to practice the disclosure, it is to be understood that other embodiments may be realized and that logical and mechanical changes may be made without departing from the spirit and scope of the present disclosure. Accordingly, the detailed description herein is provided for purposes of illustration only and is not intended to be limiting. For example, the steps recited in any method or process description may be performed in a different order and are not necessarily limited to the order presented. Moreover, references to a single embodiment can include multiple embodiments, and references to more than one component can include a single embodiment.

FIG. 1 illustrates components of the disclosed system for monitoring consumption events. The system includes a smart tag 104 with a motion sensor configured to communicate with the server 102 to estimate and predict consumption of a CPG target (referred to as a "target" or "CPG") 105. To automate the monitoring process, the consumer may manually attach tag 104 to target 105. Alternatively, the equivalent of the tag 104 including the motion sensor may be built directly into the container or packaging of the CPG 105. Manual attachment may take different forms depending on the particular shape and design of tag 104 and the packaging of target 105. For example, the tag 104 may be clamped, taped or glued or otherwise attached to the packaging of the object 105. The attachment may be secure such that any movement of the container or packaging of object 105 will cause tag 104 to move accordingly.

The consumer may configure tag 104 with information describing a set of gestures, which are typically associated with consuming a particular type of merchandise. Additionally, parameters used in the model relating the consumption event to the remaining amount of the target 105 may be estimated in advance and initial values of the parameters configured onto the tag 104. Depending on the particular model and motion sensor type(s) used, these configurations and parameters may include, but are not limited to, the average number of consumption events in a full tracking period (session); thresholds for sensory input describing consumer gestures such as tilt, lift, drop, and pan (panning); and the average length of the household size/tracking period for the type of merchandise to be monitored. The configuration of the tag 104 may require the consumer to use a smart phone or device 100 to communicate with the tag 104, for example, using RFID programming or Bluetooth Low Energy (BLE) applications. Alternatively, the consumer may first enter the configuration via the smartphone or device 100 and save it to the server 102 for retrieval by the tag 104.

Once tag 104 is configured and securely attached to target 105, the consumer may indicate the start of the tracking period by issuing a command to tag 104; for example, by pressing a button or sending a start signal to tag 104. During the tracking period, the motion sensor(s) of tag 104 continue to operate and provide input to detect consumer gestures to detect consumption events. Gesture detection is based on pattern recognition from observed changes in orientation, angular velocity, and/or acceleration along 3 axes, or a combination thereof, of the target 105 or its packaging. Once a complete set of gestures related to a consumption event is identified, tag 104 notifies server 102, and server 102 records details about the consumption event and updates its consumption prediction model. The complete set of gestures includes one or more gestures associated with a consumption event.

Tag 104 may also include environmental sensors such as thermometers, humidity sensors, light sensors, and other sensors critical to monitoring the preservation environment of target 105. The preservation environment may be compared to a data set of preferred or optimal preservation conditions for the longest lifetime or "best at use" conditions for the target 105 to determine when a significant deviation occurs.

The server 102 maintains an operational model that keeps track of consumption events detected since the beginning of the tracking period. As more consumption events are identified and over time, the state of the model is updated. The model can be used to predict how much of the target 105 is left by considering historical data about consumption, user habits, and the state of the model in the current tracking period. If the model determines that the remaining amount is low, it can alert the consumer's smart phone or device 100 and automatically place a restock order to the supplier 103 to replenish the merchandise.

The consumer may stop the tracking period by issuing a command to tag 104, for example, by double-pressing a button on tag 104 or otherwise sending an end period signal. When this occurs, tag 104 sends a notification to server 102 and server 102 stops the corresponding model for tracking target 105. The key model parameters, such as average tracking period length and total number of consumption events, are updated according to the rule set described herein.

Communication between the tag 104 and the hub 101 uses wireless technology, such as Bluetooth Low Energy (BLE) or Zigbee. If tag 104 communicates with server 102 using Wi-Fi, hub 101 is considered a Wi-Fi access point. Other wireless technologies may similarly be used as transmission technologies without departing from this disclosure. Multiple hubs 101 may be deployed to increase reception and coverage and thereby minimize data loss during wireless communication with multiple tags 104. Each hub 101 communicates with a server 102 via LAN and Wi-Fi. Server 102 may perform the above-described consumption modeling and prediction in the cloud on behalf of the consumer for multiple targets 105 simultaneously. The modeling results, reminders and descriptive information can be sent to the consumer's smart phone or device 100 when needed. Server 102 communicates with suppliers 103 to place and manage new/existing restocking orders and deliveries.

Motion sensor

Tag 104 may be equipped with a plurality of motion sensors to detect consumption events. These sensors are typically lightweight, consume low power, and are resistant to environmental changes. The sensor may optimally have a small footprint to enable attachment to a Printed Circuit Board (PCB) or flex circuit of the tag 104 as an electronic component. Alternatively, the tag 104 may be built into the CPG packaging.

Illustrative embodiments of the present disclosure may include an accelerometer, a compass, and a gyroscope for detecting consumption events. The tag 104 may also be equipped with environmental sensors to monitor the quality (quality) of the target 105 in addition to the remaining amount. Environmental sensors are described in further detail in U.S. co-pending patent application serial No. 16/137,835, which is incorporated by reference herein in its entirety.

Fig. 2 shows a process flow diagram of an illustrative embodiment of a system and method for monitoring a CPG target 105 for consumption events according to the present disclosure. The system starts 200 a tracking period and monitors 202 the target 105 for consumption events.

An illustrative embodiment of tag 104 may include accelerometer 204 as a motion sensor. The accelerometer 204 is a device that measures acceleration. The accelerometer 204 measures linear acceleration of the target 105 and its container or package as the consumer performs at least one of a series of poses, such as lifting, tilting, moving, and lowering.

Industrial micro-electro-mechanical systems (MEMS) type accelerometers can be used because they are configured for ultra-low power and have a high sensitivity to three-axis linear acceleration output. Advanced techniques, such as FIFO buffers, allow the sensed data to be stored to limit interference by the host processor, thereby further reducing power consumption during operation. Accelerometers are typically made using small thin plastic Land Grid Arrays (LGAs) and can operate over a wide temperature range of-40 to 85 degrees celsius. Such an accelerometer thus allows monitoring of the target 105 during storage in a climate controlled environment (e.g., a food storage compartment, refrigerator, or freezer).

Another sensor used in this illustrative embodiment is a compass 206 or magnetometer. The compass 206 is a magnetic sensor that detects a change in orientation of the commodity target 105. When the target 105 moves relative to the orientation and thus may experience a consumption event, the compass 206 may be used to detect motion 202 due to the consumer.

Tag 104 may also include a gyroscope 208, either alone or in coordination with other sensors. The gyroscope measures the angular displacement/velocity of the target 105 in 3 dimensions. The gyroscope 208 is capable of digitally detecting and describing information about the gesture(s) of the consumer during a consumption event. Available MEMS gyroscopes can be implemented as small surface mount electronic components with low power consumption and high energy efficiency. They are capable of operating in extreme temperature and humidity environments.

The system and method shown in fig. 2 monitors the target 105 using a tag 104 that includes at least one motion sensor, e.g., an accelerometer 204, compass 206, or gyroscope 208. Tag 104 may periodically check whether motion of object 105 has been detected. If motion of the target 105 has not been detected, the system returns to the start point 200.

When tag 104 detects motion, it may send detected motion information to signal accumulator 212 to buffer 210 the signal detection(s) and information (e.g., information related to consumption events). The buffer 210 may wait for further signals to accumulate and thereby form a motion signal sequence aligned with a particular set of time stamps. The pattern recognition processor 216 may analyze the signal(s) and compare 214 the data to the listed indicators of the gesture. Comparing the buffered signal(s) with a preconfigured, e.g. stored, gesture library may take different forms and embodiments, e.g. using various distance measurements. If the system determines that no gesture 218 is detected, the system returns to the start 200. If the system determines that a gesture is detected 220, the system adds data to the model for processing 222. The system continues to monitor target 224 for additional poses.

Table 1 below illustrates examples of the types of sensors discussed. A reference is also provided for each sensor type, containing information about approximate cost, package size, operating temperature, and power consumption rate.

TABLE 1 motion sensor and Attribute

A combination of multiple motion sensors, such as accelerometer 204, compass 206, and gyroscope 208, may be used to provide accurate detection and to mask noise that leads to false positive/negative identifications. Alternatively, only one type of sensor may be used. The prior art MEMS technology makes it possible to allow the tag 104 to include all three types of motion sensors, thereby achieving higher identification accuracy at a low level of overall cost, combined power budget, and relatively small footprint.

Gesture recognition

An active consumption event from a consumer is typically associated with at least one position at which the consumer approaches, removes and/or returns merchandise target 105 or its container. Since tag 104 is attached to (or integrated with) target 105, the system is able to detect and recognize one pose or a set of poses for a single consumer, for each type of merchandise, each pose being characterized by linear acceleration, orientation, angular displacement/velocity, and changes in these quantities. As described above, definitions of these gestures and their corresponding sensing descriptions may be configured into tag 104 (e.g., in the form of a gesture table having characteristic accelerations, orientations, and/or angular displacements, etc. that define a particular gesture or consumption event). However, the consumer can override (override) the default set of poses by providing an individualized set of pose definitions for a particular merchandise object 105 under its profile.

At the beginning 200 of a defined time interval (e.g., every second), the system checks 202 to see if there is a significant change in orientation, acceleration, or angular velocity from a motion sensor such as accelerometer 204, compass 206, or gyroscope 208. It will be appreciated by those skilled in the art that as an alternative to the system checking to see if motion sensor activity is present, the tag may be configured to periodically transmit information from the sensor(s). If there is no change, the system returns to the origin 200. If a significant change is detected from one or more of these quantities, this may be the beginning of a consumption event, and the change is first written to buffer 210 by signal accumulator 212. The signal accumulator 212 may wait for further signals to accumulate and thus contain a sequence of motion signals at a particular set of time stamps. The buffer 210 is then reviewed 214 by the pattern recognition processor 216 to see if a defined signal pattern representing a preconfigured gesture matches.

Comparing the buffered signal(s) to the preconfigured poses may take different forms and embodiments, e.g. using various distance measurements. The measurements may preferably be based on the defined pose to be recognized, computational complexity and memory requirements. Other poses may use different measurements to determine similarity. A look-up table containing all preconfigured poses may be stored on the server 102 or the tag 104. A look-up table may be used and gestures with similarity scores that pass a defined threshold may be considered positive detections 220. If no gestures in the lookup table match the current buffer and no gestures 218 are detected, the system returns to its origin 200 and is ready to process again in the next second or time period. If a pattern is recognized, the recognized gesture is sent to a rule set to determine if the consumption event can be adequately determined 222. Sometimes, a consumption event requires more than one matching gesture rather than a series of gestures performed over time. After a positive match, the system then resets 224 the signal accumulator 212 and returns to its initial state 200, and is ready to start over again in the next time period.

Table 2 provides an example of preconfigured gestures in a consumption event, which contains corresponding signal descriptions for the relevant motion sensors. By way of example, table 2 describes gestures typically associated with the use or consumption of laundry detergent by a washing machine, however, it should be understood that this is merely an example illustrating a consumption event. In this illustration, consumer a opens the lid of the detergent bottle, lifts the bottle, and pours detergent into the drawer of the washing machine. The series of gestures performed collectively during a typical consumption event is defined in the table below and, as in this example, if the associated target 105 is identified as a laundry detergent, it may be pre-configured onto the tag 104.

TABLE 2 attitude and Association measurements

Indeed, the detergent bottle may have a screw cap or a flip cap. Sometimes it may not have a removable cover at all. The label 104 may be attached to, for example, the cap or body of the bottle (e.g., on its neck or handle) at the option of the consumer. Depending on the configuration, the version of tag 104 may be equipped with only accelerometers or with multiple types of sensors. Thus, the rule set is designed to accommodate these different situations and determine whether a gesture or series of gestures justifies a valid consumption event.

Table 3 is an example of an exemplary rule set that assumes that an accelerometer sensor participates therein and optionally a gyroscope or compass is present. Table 3 is an exemplary embodiment and is provided as an example only. More complex rule sets may be used, and different rule sets may be specified and used for different types of goods without departing from the disclosure.

TABLE 3 motion sensor and Attribute

Step (ii) of	Evaluation/declaration	True	False
				1	Pattern #2 detected?	Go to step 2	Go to step A
2	Is the tag equipped with a compass?	Go to step 3	Go to step 4
				3	Pattern #1 detected?	Go to step B	Go to step A
4	Is the tag equipped with a gyroscope?	Go to step 5	Go to step B
				5	Pattern #3 detected?	Go to step B	Go to step A
A	And (3) judging: no, no valid consumption event
				B	And (3) judging: is that a valid consumption event is detected

FIG. 3 shows a process flow diagram for detecting a consumption event as described in Table 3. At the beginning 300 of a defined time interval (e.g., every second), the system checks 302 to see if there is a significant change in orientation 306 or angular velocity 310 from a motion sensor, such as compass 304 or gyroscope 308. If no change is detected, the system returns that no consumption event has occurred 312. If a change is detected from one or more of these quantities, this may indicate the beginning of a consumption event 314.

Consumption prediction

An illustrative consumption prediction model is described herein that describes the relationship between consumption events and the remaining amount of merchandise (as a percentage). Different models may be used for different CPG targets 105. The server 102 may maintain a plurality of models and use at least one model to perform predictions about consumption of the commodity object 105 depending on the consumer's preferences and behavioral changes.

The consumption prediction characteristics for each commodity include a set of parameters and variables. The corresponding parameters and variables represent the current state of the CPG target 105. For example, the variables may include the time since the start of the tracking period, the cumulative number of consumption events detected in the current tracking period, and the elapsed time between past consumption events. The consumption prediction parameter is a coefficient that relates a variable to an output state, such as the amount of target commodity remaining as a percentage.

During the tracking period, tag 104 may continuously feed sensed information to server 102, including the elapsed time since the beginning of the tracking period, as well as detailed information about the identified consumption event, and other information, such as the timestamp of the consumption event. The information may be used to immediately update the state of the consumption prediction model running on the server 102. During the tracking period, the parameters are unchanged. After the tracking period is over, the model parameters are updated. While the predictive model is being consumed, prediction can be performed from the current state of the model by following the model and reaching the level of the predicted residual of the target 105.

Simpler models may be easier to implement and estimate their parameters, but may suffer from lower accuracy. More comprehensive models with more parameters and variables can be used to achieve greater accuracy if desired. The parameters and variables used will depend on the sensors mounted on the tag 104. The following simple model is shown as an example of a linear relationship between the number of cumulative consumption events and the remaining amount as a percentage:

R＝a+bA+e。

where R is the amount of CPG target 105 remaining in its container. R may be a percentage. A is a variable representing the number of cumulative consumption events since the beginning of the tracking period. The model parameters a and b are used to determine the residual amount (R) based on the number a of tracking periods. At the beginning of the tracking period, a is the total amount of CPG targets 105 present. In the case where the CPG target 105 starts from 100%, an example value of a is 1. The average decrease per consumption event is defined as b. The consumer may enter parameter b or may estimate parameter b based on past consumption events or information obtained from server 102. Errors and inaccuracies can be captured by e. For example, e may capture an error rate in the record that detects the consumer gesture to trigger a consumption event. A is defined as 0. ltoreq. A. ltoreq. a/b, otherwise R.ltoreq.0%.

This simple model requires estimation of the coefficients a, b. If b is assumed to be constant, it can be measured directly from a previously recorded tracking period of the same commodity object 105. Once b is known, R can be easily predicted given the state variable a.

Various techniques may be used to improve the model and reduce the error rate. For example, a and b can be considered as random variables and regression techniques applied to estimate their values and minimize their mean square error. Assuming that a and b do not change significantly over a defined period of time (e.g., over the past three month period) in which the consumer's habits and consumption behavior remain stable, the server 102 may collect historical data about the recorded (R, a) pairs from previously completed tracking periods of the same CPG target 105. The consumption prediction model may use the data set to estimate two model parameters.

Table 4 is an example of 5 tracking periods over a three month period. The (R, a) pairs recorded in these tracking periods are as follows.

TABLE 4 historical data observed for the detergent example

Tracking periods	Automatic collection (R, A)	User input (R, A)
			1	(1,0),(0,10)	Is free of
2	(1,0)	(0.5,6)
			3	(1,0),(0,9)	Is free of
4	(1,0),(0,11)	Is free of
			5	(1,0),(0,10)	Is free of

The auto-collect column of table 4 refers to samples collected during a past tracking period. In this example, when the tracking period begins, the system assumes that R is 100% and a is 0. When the consumer ends the tracking period, the system assumes that the target 105 has been completely consumed, and thus R is 0%. When the tracking period is over and the system estimates that target 105 may not be completely consumed, the system may ask the consumer to enter an estimate of the remaining amount of target 105. In the above example, the consumer provides information for tracking period #2 when the goods remain 50% of the way through 6 consumption events. After performing linear regression, the estimation model based on example data is:

R＝1.00309-0.0985A

as shown, this example finds that A is 0 when R is about 100%. Furthermore, by looking at the estimated coefficient b, each consumption event in this example is a reduction of about 10% in the total quantity of the commodity 105. The system may use this estimated model in future consumption events.

As a further example, assume that in the next detergent tracking period, there are 8 consumption events for the same consumer. If all conditions are assumed to be equal to the above example, the remaining amount is predicted to be 1.00309-0.0985 × 8 — 21.5%. Given that the average length of the consumer's tracking period for detergent in the past three months (in units of time) is T, the time remaining until the detergent is completely depleted is about 0.215T.

Additionally, where tag 104 includes environmental sensors, tag 104 may record the preservation conditions of object 105. The server 102 may use the preservation conditions of the target 105 compared to the optimal preservation conditions to evaluate and determine an inferred expiration time of the target 105. Server 102 may store the predicted remaining life of target 105 as T₂. As the tracking period continues, the server 102 may update the model to predict when the target 105 has expired.

The parameters of the consumption prediction model are frequently updated and modified by the system. For example, if the CPG target 105 is ice cream, the coefficient b may be less during the winter season than during the summer season. When the number of households changes, the coefficient b is likely to change. Thus, the system may recalibrate itself using the latest input from the sensors and the consumer, as well as the data available from the server 102, to maintain high accuracy. The system may re-estimate the model parameters using the updated data shortly after the end of the tracking period. The calibrated model may be used during a new tracking period.

The above examples illustrate only a specific model build in application of commodity consumption prediction by pose detection. In practice, multiple models may be constructed and evaluated to produce more accurate predictions. The server 102 has the ability to track data and evaluate multiple models simultaneously, however, based on their evaluation, one model may be used to make predictions at a particular time. The server 102 may self-calibrate and automatically make the estimation continuously.

Fig. 4 illustrates a process flow diagram for monitoring the CPG target 105 and determining when the CPG target 105 requires replenishment 424. The system selects a consumption prediction model 402 for a particular target 105. The consumption prediction model may use past consumption history stored on the server 102 or use data entered by the consumer to develop parameters 404 associated with the CPG target 105. Tag 104 may use the preconfigured list to determine 406 how to recognize gestures associated with consumption events of target 105. Tag 104 connects with target 105 and a tracking period is initiated 408.

In operation, tag 104 tracks 410 consumption events of target 105 during a tracking period. Tag 104 updates 416 the consumption prediction model with the consumption events and associated timestamps during the tracking period. The consumer may end the tracking period 412. Tag 104 ends tracking 414 when R is estimated to be equal to 0 or the consumer ends tracking period 412. The model updates parameters (e.g., for model calibration) based on the tracking period data 418. The consumption/depletion model runs 420 to determine an estimated R value (i.e., the percentage of target commodity remaining) 422. The model determines 424 when replenishment of inventory/replenishment is required (e.g., when a threshold level set for replenishment is reached). If the model determines that replenishment of inventory/replenishment is required, the supplier may be placed an order to replenish the inventory/replenishment target 104. When the model determines that replenishment of inventory/replenishment is required, the server 102 can issue alerts to the customer and/or provider, such as via Short Message Service (SMS), Enhanced Message Service (EMS), Multimedia Message Service (MMS), instant message, email notification, and the like.

In parallel, a calibration process may be initiated to allow the consumer to provide a set of gestures that are most likely to be used in a typical consumption event. The motion sensor may record these preconfigured poses with thresholds (e.g., the thresholds described in table 3). After obtaining initial estimates of the model parameters, these parameters may be used to perform predictions during the real-time tracking period. While the tracking period is ongoing, motion sensors, such as accelerometer 204, compass 206, and gyroscope 208, are used to detect consumption events using a rule set by recognizing a set of predefined gestures performed over time. As the tracking proceeds, server 102 updates its model variables with information about the consumption events and their timestamps. The server 102 performs predictions using the model to derive the time remaining until the CPG target becomes fully depleted (or reaches a threshold consumption level). At the end of the tracking period, the model self-calibrates its parameters based on its recorded information and any consumer input. After calibration, the updated model is ready for another tracking period. The process is iterative.

Fig. 5 illustrates a process flow diagram for monitoring the quantity and quality of the CPG target 105 according to the present disclosure. The system selects a consumption prediction model 502 for a particular target 105. The model may use past consumption history stored on the server 102 or use data entered by the consumer (e.g., training) to develop parameters 504 associated with the CPG target 105. Tag 104 is connected to target 105 and a tracking period 508 is initiated, for example, by the consumer. The tag begins a tracking period 510. The consumer may end the tracking period 512. The tag may be detached from the target 514 and the time period data used to calibrate 528 the model parameters.

Tag 104 tracks 510 environmental conditions 516 and consumption events 522 of target 105 during a tracking period. Tag 104 or server 102 may run a number of models, including a quality model 518 for tracking the quality of target 105 and a consumption prediction model 524 for monitoring the remaining amount of target 105. The expiration 520 of the target 105 is estimated by the quality model 518. The remaining amount 526 of the target is estimated by the quantity model 524. The system uses the values calculated for the target's estimated expiration 520 and/or consumption 526 to determine 530 when replenishment of inventory/replenishment is needed. If the system determines that replenishment of inventory/replenishment is required, it may place an order with the supplier to replenish the inventory/replenishment target 104. When the model determines that replenishment of inventory/replenishment is required, the server 102 may issue a reminder to the consumer and/or supplier.

Although an exemplary server is described in the embodiments herein, it should be understood that the processes for consumption prediction and quality monitoring described herein may be implemented in program code by a microcontroller, for example, configured in an appliance such as a refrigerator, freezer, storage cabinet, etc., and one skilled in the art will appreciate that discrete control electronics, large scale integrated circuits, or other control techniques may be used to implement the functions described herein.

Although exemplary sensors are disclosed in embodiments herein, including accelerometers, compasses, and/or gyroscopes for monitoring gestures or consumption events, it should be understood that other types of sensors, such as ultrasonic, vibration, infrared, microwave, etc. sensors may be implemented in accordance with the disclosure.

While various embodiments are disclosed herein, it is to be understood that the disclosure is not so limited and that modifications may be made without departing from the disclosure. The scope of the present disclosure is defined by the appended claims, and all devices that come within the meaning of the claims are intended to be embraced therein either literally or equivalently.

Claims

1. A method of monitoring the remaining amount of a consumable good, the method comprising the steps of:

attaching a tag comprising at least one motion sensor to a consumer good;

configuring the tag to identify a gesture associated with a consumption event of the consumer good;

initiating a tracking period using the at least one motion sensor to detect motion of the consumer good to monitor a gesture associated with the consumption event of the consumer good;

providing information from the at least one motion sensor included on the tag to a consumption prediction model configured to determine a remaining amount of the consumable good;

determining whether the motion is a recognized gesture associated with a consumption event of the consumer good; and

predicting a remaining amount of the consumable good based on the consumption event of the consumable good.

2. The method of monitoring the remaining amount of a consumable good as recited in claim 1, further comprising:

receiving a signal from the at least one motion sensor;

buffering the signal;

comparing the signal to a preconfigured gesture list, the preconfigured gesture list associated with the consumption event of the consumer good; and

determining whether a recognized gesture associated with a consumption event of the consumer good has been detected.

3. The method of monitoring the remaining amount of a consumable good as recited in claim 1, further comprising:

updating, with the consumption event and associated timestamp, a consumption prediction model configured to determine a remaining amount of a consumed commodity model.

4. The method of monitoring the remaining amount of a consumable good as recited in claim 1, further comprising:

and reminding the consumer of the residual quantity of the consumed commodity.

5. The method of monitoring the remaining amount of a consumable good as recited in claim 1, further comprising:

predicting when the consumer good will be depleted;

generating an order to replenish the consumer good; and

notifying the supplier of replenishment of said consumer good.

6. The method of monitoring the remaining amount of a consumable good of claim 5, further comprising:

placing an order to replenish the consumer good before the consumer good is completely depleted.

7. The method of monitoring the remaining amount of a consumable good as recited in claim 1, further comprising:

ending the tracking period for the consumer good;

updating the model using the elapsed time for the tracking period and the number of consumption events that occurred during the tracking period; and

calibrating the consumption prediction model.

8. The method of monitoring the remaining amount of a consumer good as recited in claim 1, wherein the at least one motion sensor is selected from the group consisting of an accelerometer, a compass, and a gyroscope.

9. The method of monitoring the remaining amount of a consumable good as recited in claim 1, further comprising the steps of:

determining a coefficient of a consumption rate of the consumable good with respect to each consumption event;

determining an average length of a tracking period for the consumer good;

predicting an estimated time remaining until the consumer good is depleted; and

placing an order to replenish the consumer good before estimating the consumer good to be depleted.

10. A system for monitoring the remaining amount of a consumable good, comprising:

a server;

at least one consumption prediction model configured to predict a remaining amount of the consumed commodity, the at least one consumption prediction model running on the server and comprising at least one parameter representing the remaining amount of the consumed commodity;

at least one tag disposed on at least one consumer good, the at least one tag including at least one motion sensor and configured to detect motion of the at least one consumer good and generate a signal from the at least one motion sensor;

a signal accumulator operatively connected to the at least one motion sensor, the signal accumulator configured to receive signals from the at least one motion sensor; and

a pattern recognition processor configured to compare the signal from the at least one motion sensor to a preconfigured table defining motions associated with consumption events.

11. The system for monitoring the remaining amount of the at least one consumer good as recited in claim 10, further comprising:

a consumer device operatively connected to the server;

the consumer device is configured to instruct the server to start or end a tracking period; and is

The consumer device is further configured to receive an alert from the server, the alert including a remaining amount of the at least one consumable good.

12. The system for monitoring the remaining amount of the at least one consumer good as recited in claim 11, wherein the server is configured to calibrate the model after a tracking period ends.

13. The system for monitoring the remaining amount of the at least one consumable good as recited in claim 11, wherein the server is configured to send a reminder to a supplier to replenish the consumable good when the at least one consumption prediction model predicts a need for replenishing the consumable good.

14. The system for monitoring the remaining amount of the at least one consumable good as recited in claim 11, wherein the tag is configured to communicate with the server to update the at least one consumption prediction model when the pattern recognition processor detects that a consumption event has occurred.

15. A label system for monitoring the remaining amount of a consumer good, comprising:

at least one motion sensor integrated with the consumer good;

a pattern recognition processor operatively connected to the at least one motion sensor, the pattern recognition processor configured to receive at least one signal from the at least one motion sensor and determine whether a consumption event has occurred;

a consumption prediction model for predicting a remaining amount of the consumable good, the consumption prediction model configured to be updated when the pattern recognition processor determines that a consumption event has occurred.

16. The tagging system for monitoring the remaining amount of a consumable good of claim 15, wherein the consumption prediction model outputs a reminder including the predicted remaining amount of the consumable good.

17. The label system for monitoring the remaining amount of a consumer good as recited in claim 16, wherein the alert is received by a consumer on a consumer device.

18. The tag system for monitoring the remaining amount of a consumer good as recited in claim 16, wherein the tag is configured to send a replenishment notification to a supplier when the amount of the consumer good needs to be replenished.

19. The tagging system for monitoring the remaining amount of a consumable good of claim 16, wherein the tag is configured to create a replenishment notice to replenish the consumable good before the consumption prediction model predicts a time at which the consumable good will be depleted.

20. The labeling system for monitoring the residual amount of a consumer good as defined in claim 15, wherein the at least one sensor is selected from a group of sensors comprising an accelerometer, a compass and a gyroscope.