US20190156270A1 - Distributed Sensor System and Method for Inventory Management and Predictive Replenishment - Google Patents

Abstract

An system and a method for scoring products viewed by a user prior to making a purchase decision and also identify the products considered most before ultimately purchasing. The system uses sensors placed on the shelves to identify the item viewed before making a purchase decision and also track the time spent per decision. The tracking will allow viewed and ultimately bought scores and item similarity scores to be determined. The scores in conjunction with point-of-sale terminal data to identify times determinative of users' interest levels prior to a purchase. The system predicatively replenishes inventory of high interest items based on historical patterns of interest and historical point-of-sale terminal data.

Description

RELATED APPLICATIONS
• This application claims priority to Indian Patent Application No. 201711041326 entitled “DISTRIBUTED SENSOR SYSTEMS AND METHODS FOR INVENTORY MANAGEMENT AND PREDICTIVE REPLENISHMENT,” filed on Nov. 18, 2017, the content of which is hereby incorporated by reference in its entirety.
BACKGROUND
• Online shopping the shopping pattern and buying behavior of the users are easily understood by tracking the similar items viewed and the product ultimately bought through session tracking. The viewed and ultimately bought (VUB) scores can be used to determine suitable substitute items. The viewed and also viewed scores help in determining similar items or complements. Also the list of items viewed together before making a shopping decision helps in better understanding of the product attributes a user is looking at before buying a product. Through these scores, the key attributes leading to a purchase can be determined which can help us to better understand the market trend and improve our assortment. Unfortunately, such scores are unavailable at a retail store level.
BRIEF DESCRIPTION OF DRAWINGS
• Illustrative embodiments are shown by way of example in the accompanying drawings and should not be considered as a limitation of the present disclosure:
• FIG. 1 is a block diagram illustrating a distributed sensor system according to an exemplary embodiment.
• FIG. 2 is diagram illustrating a deployment of an array of sensor modules in the distributed sensor system according to an exemplary embodiment.
• FIG. 3 is a diagram illustrating a deployment of an array of sensor modules in a shelving unit according to an exemplary embodiment.
• FIG. 4 is a flowchart illustrating a process implemented by the distributed sensor system for determining continuous detection of a user according to an exemplary embodiment.
• FIG. 5 is a flowchart illustrating a process implemented the distributed sensor system for modifying inventory records based on sensor outputs according to an exemplary embodiment.
• FIG. 6 is a flowchart illustrating a process for updating inventory records according to an exemplary embodiment.
• FIG. 7 is a block diagram illustrating an electronic device operable to support the distributed sensor system according to an exemplary embodiment.
• FIG. 8 is a block diagram of an example sensor module to support the distributed sensor system according to an exemplary embodiment.
DETAILED DESCRIPTION
• Described in detail herein is an system for inventory management. The system for inventory management provides for the identification of which products are viewed by a user before making a purchase decision and also the identification of the products considered most before purchasing ultimately. Additionally, the system can utilize the identification of products considered most before purchasing ultimately to predictively manage inventor of the considered product.
• FIG. 1 is a block diagram illustrating a distributedsensor system100 for controlling and managing inventory according to an exemplary embodiment. Embodiments of thesystem100 can include an array ofsensor modules102,servers104 and108, one or more point-of-sale terminals106,databases110A,110B, and at least onenetwork access point112.
• Each sensor module in the array ofsensor modules102 can be considered a node and each sensor module in thearray102 can be in selective communication with adjacent or neighboring nodes in the array. The sensor modules can communicate with adjacent or neighboring nodes through wired and/or wireless communication. Communication support to facilitate communication between adjacent nodes in the array of sensors can include Bluetooth®, Zigbee, WiFi (e.g., as specified by IEEE 802.11a-n), near-field communications (NFC) or other comparable wireless communication stacks. Additionally, the array ofsensors102 can be implemented within an Internet of Things framework such as ioTivity or Zephyr, which support a development stack with underlying communication application programming interfaces (APIs) already implemented. The array ofsensor modules102 can be implemented with sensing elements such as optical sensors, acoustic sensors, weight sensors, radio-frequency identification (RFID), inductance sensors, capacitive sensors, cameras (e.g., CMOS imagers or CCD sensors) or a combination of these sensors.
• The array ofsensor modules102 can be in selective communication with theserver104 via the network access point. The network access point is an electronic device that provides the sensor modules access to a network including theserver104. The network access point can route or relay data communications between devices in a network (e.g., between the sensor modules and the server104). In exemplary embodiments of the network access point can be embodied as a relay, a router, a gateway, a hub, a server, a modem, or any other device suitable for facilitating connection of the sensor modules to the network and routing data communications between devices on the network. As a non-limiting example, the network access point can be a WiFi access point.
• Theserver104 receives data from the array ofsensors modules102 via thenetwork access point112. The data can include information relating to the placement/location of the sensor module(s) transmitting the data, information regarding objects detected by the sensor module(s), and other relevant data corresponding to the type of sensor(s) implemented by the sensor module(s) (e.g., images from an optical sensors, weights measurements from a weight sensor, data from an IR or acoustic sensor, etc.). Theserver104 utilizes the data from the array ofsensor modules102 to determine user interest in an item placed near to the sensor. Theserver104 can also determine attributes of items selected or, alternatively, not selected by a user, and provide recommendations on item placement to improve the likelihood of an item being selected
• The one or more point ofsale terminals106 facilitate transactions of viewed and selected items by the users. Point ofsale terminals106 transmit transaction data to the server to provide correlative data between items viewed by a user, time spent viewing an item by a user, and the purchasing of an item by a user.
• Theserver108 can be an inventory server that manages interaction between inventory andproduct databases110A,110B. Theinventory server108 can also facilitate any inventory replenishment activities as well. Theinventory server108 can be a logically separate entity from theserver104 and thedatabases110A,110B, but physically theinventory server108 can coexist on the same hardware as theserver104. As theinventory server108 can service other disparate systems, theserver108 can provide additional computational resources to avoid overutilization of theinventory server108. Theserver108 provides the computational resources to decode data from the array ofsensor modules102, as well as interfacing with the point-of-sale terminal106. To minimize computational burden on theinventory server108, theserver104 only interfaces with the inventory server when information inventory or product information is necessary for the process.
• Thedatabases110A,110B can include rows or records corresponding to items in a facility, items of interest being viewed and/or selected by users, items being monitored by the array ofsensor modules102. Theinventory server108 provides an interface for accessing thedatabases110A,110B. As theinventory databases110A,110B can be accessed by other disparate systems, theinventory server108 can provide support for all query requests for the database. Theinventory server108 can queue query requests to mitigatedatabase110A,110B query overloads. At least some of the records can be generated on a set interval to display trends relating to the items being analyzed by the array ofsensor modules102. Thedatabases110A,110B can include records corresponding to past item purchase history and trends. Additionally thedatabases110A,110B can include latent attributes corresponding to each item in thedatabases110A,110B where each latent attribute corresponds to a characteristic of the product. Like theinventory server108, thedatabases110A,110B can be logical instances of databases, and can physically reside on theserver104.
• FIG. 2 is diagram illustrating the deployment of an embodiment of the array ofsensors102 in a retail environment according to an exemplary embodiment. In the present example embodiments, the sensor modules in the array ofsensor modules102 can wirelessly communicate with adjacent/neighboring sensor modules. As mentioned above the array ofsensor modules102 can utilize one or more wireless communication stacks or protocols.
• Each of the sensor modules in thearray102 can have three modes of operation based on object detection by the sensor in each sensor module and/or based on the object detection by the sensors in one or more adjacent or neighboring sensor modules. As one example each sensor module can have sleep state (e.g., a first mode of operation), an active state (e.g., a second mode of operation), and an intermediate state (e.g., a third mode of operation). In exemplary embodiments, each of the sensor modules can normally be in the sleep state of operation (i.e. the default state)
• In an exemplary operation, when the sensor of afirst sensor module102A detects an object (e.g., a user standing in front of and within range of the sensor), thefirst sensor module102A transitions from the sleep state to the active mode. Thefirst sensor module102A determines the existence and/or movement of a user based on sensed data output by the sensor. In another embodiment, thefirst sensor module102A, transmits the data to theserver104, which determines the existence and/or movement of the user.
• In response to the first sensor module detecting the user, the first sensor module transmits a wake message to its adjacent/neighboring sensor modules, which can include asecond sensor module102B as well as other sensor modules. Additionally, or in the alternative, as described herein, the first sensor module can transmit a message to the server104 (FIG. 1) if the sensor module continuously detects the user for a specified time threshold. In response to thesecond sensor module102B receiving the wake message fromfirst sensor module102A, thesecond sensor module102B, transitions from the sleep state to the intermediate state. While thesecond sensor module102B is in the intermediate state, thesecond sensor module102B recordation of sensing data is enabled in the second sensor module and the wireless transmitter is disabled.
• In one example operation, the first thesensor module102A also transmits sensing data to the server upon continuous detection of the first user within range of the first one of the sensor modules for a period of time that is greater than the threshold time or reverts back to the sleep state upon detection of the user within range of the first one of the sensor modules for a period of time that is less than the threshold time.
• Upon detection of the second user by thesecond sensor module102B, thesecond sensor module102B, transitions from the intermediate state to the active state. The second sensor module transmits a wake message to its adjacent/neighboring sensor modules. Additionally, or in the alternative, as described herein, the first sensor module can transmit a message to the server104 (FIG. 1) if the sensor module continuously detects the user for a specified time threshold.
• As the user moves out of range of the sensor modules in that are in the active state and a specified amount of time has passed, the sensor modules return to the sleep state unless the sensor modules receive a wake message from a neighboring sensor module in which case the sensor module(s) enter the intermediate state.
• FIG. 3 is a diagram illustrating the deployment of an array of sensor modules in a shelving unit according to an exemplary embodiment.
• The deployment of an array ofsensor modules102 can include affixing of each of the array ofsensor modules102 to ashelving unit300. The array ofsensor modules102 can be deployed on the shelves of theshelving unit300. Alternatively, the array ofsensor modules102 can be deployed at the top of theshelving unit300 or the base of theshelving unit300. The array ofsensor modules102 can be placed correspondingly to different products located on the shelves of theshelving unit300, where one of the sensor modules in the array ofsensor modules102 corresponds to a specific item stock. Each of the sensors in the array ofsensor modules102 can be spaced by the average dimension of items in a category corresponding to the items.
• In this embodiment thefirst sensor module102A can be placed correspondingly to aspecific item A302. Thefirst sensor module102A can detect and transmit any data corresponding to interactions with item A.
• Thesecond sensor module102B can be placed corresponding to aspecific item B304. Thesecond sensor module102B can detect and transmit any data corresponding to interactions with item B.
• Thethird sensor module102C can be placed corresponding to aspecific item C306. Thesecond sensor module102C can detect and transmit any data corresponding to interactions with item C.
• In this embodiment, as the user viewsitem A302, moves toitem B304, and finally toitem C306, each of the sensors in the arrays ofsensors102 changes power states corresponding to the item being observed. In this embodiment, historical data about the item being viewed can be determined by the changing of power states of the modules.
• FIG. 4 is a flowchart illustrating a process determining continuous detection of a user according to an exemplary embodiment.
• Atstep402, sensor modules can be placed equidistantly in shelves of the store. The sensors can be an array of sensors. The sensors can be implemented as optical sensors, acoustic sensors, weight sensors, RFID tags/readers, inductive sensors, capacitive sensors, cameras or a combination of these. The sensors detect and record the amount of time spent by the user in front of the item. The adjacent sensors communicated through Bluetooth, or comparable wireless protocol, to detect and track movement of users from one item to another and to detect double-counting.
• Atstep404, a product viewed by the user is detected by one or more of the sensor modules based on the height and position. A time spent by a user i in front of an item j can be denoted T_ijLet Y_ijdenote an indicator variable such that: Y_ij=1, if the user i and purchases the item j. Otherwise Y_ij=0. Then for the j^thitem, a logistic regression is performed as follows:
• 
  P_ij=Probability (Y_ij=1)
• 
  log(PW_ij/(1−P_ij)=α+β*T_ij+e_ij
• 
  Pij=1(1+exp(−(α+⊕*T_ij))+e_ij−1
• • where e_ij˜N(0,Σ)
• Fitting a least squares solution by minimizing the sum of squares e_ij, estimated values of the unknown values α and β can be obtained. Utilizing the estimated values of α and β, a predicted value for P_{ij can}be obtained from the observed T_ijvalues.
• Atstep406, based on logistic curve, a cut-off time can be determined by the sensor module that detected the user or by the server (in response to the sensor module transmitting sensed data to the server) to facilitate a decision indicating that the product is considered to be bought. Starting with an arbitrary cut-off value c₀, if predicted P₁₃>c₀, then the predicted indicator variable equals one, Y_ij=1. A sensitivity can be defined as a proportion of times that actual Y_ij=1. Specificity can then be defined as a proportion of times that actual Y_ij=0. A plot of specificity and its inverse, e.g., 1-specificity, forms a receiver operation characteristic (ROC) curve.
• The area under the curve (AUC) corresponding to all possible values of c₀, (0<c₀<1) can be computed and the optimal value of c₀=c₁which is the AUC maximum can be selected. Substituting Pij with c₁and using the estimated values of α and β, the optimal cut-off for T_ijcan be obtained. The optimal cut-off values are obtained for each item in a category of items available (e.g., on the shelves and/or in a facility) and updated periodically, e.g., at a daily, weekly, monthly, or yearly intervals. When the time spent in front of an item is greater than the optimal value (T_ij>c₁) for a given user/item combination is observed, the system can determine that the user is interested in that item.
• Atstep408, the server records the list of items viewed and time of consideration corresponding to each item added to the cart. The server can store the viewed and time of consideration for each item in thedatabases110A,110B. Thedatabases110A,110B can create a row corresponding to the item, where a new row represents a new observed time of consideration corresponding to each item added to the cart.
• Atstep410, based on the proportion of users viewing the item before buying another one, and time of decision, the sensor module or the server can calculate the viewed and ultimately bought (VUB) score for the item. Based on the past point of sale (POS) sales records and user data recorded by the server, the following can be calculated:
• 
  P_b|a=(number of users who are interested in itemb)/(number of users who purchase itema)
• 
  T_b|a=(average time spent by the users who purchase itema, to consider itemb)/(average time spent by the users to consider itemb)
• The VUB score of item b relative to item a is V(b|a)=P_b|a*T_b|a. Best substitutes of item a are given by those with maximum values of V(b|a).
• Atstep412, the server identifies suitable substituted items based on VUB scores. Best substitutes of item a are given by those with maximum values of V(b|a). The list of items considered by a user i immediately after putting item a in the cart can be taken into account. The also bought score (ABS) of item b relative to item a can be determined by dividing the number of users who add item b to the cart after adding item a by the number of users adding item a to the cart (ABS(b|a)=(number of users who add item b to the cart after adding item a)/(number of users adding item a to the cart)).
• The Also Viewed Score (AVS) of an item b relative to item a can be determined dividing the number of users who are interested in item b after adding item a to cart and before the next addition to cart by the number of users adding item a to cart (AVS(b|a)=(number of users who are interested in item b after adding item a to cart and before the next addition to cart)/(number of users adding item a to cart)).
• The base item similarity (BIS) score=BIS(b|a)=al*ABS(b|a)+AVS(b|a). al denotes a characteristic based on a proportion of items viewed over items purchased, where al=(average number of items viewed by users after purchasing item A)/(Average number of items purchased by users after purchasing item a). The item similarity score is TIS(b|a)=(averate time spent by users in viewing item b after purchasing item a)/(average time spent by users in viewing all items after purchasing item a before the next purchase). The item similarity score I of item b relative to item a is I(b|a)=BIS(b|a)*TIS(b|a). Best complements of item a are given by those with maximum values of I(b|a).
• Atstep414, key item attributes are decided by the server based on the VUB scores. Using the VUB scores, the list of best substitutes for each item can be obtained. Latent item attributes like price, weight, pack size, color, flavor, nutrition index can be utilized as attributes depending on the category of items. Utilizing a model V(b|a)˜11, 12, . . . where 11 and 12 are latent attributes, the variable importance score of the Random Forest model, can determine the most important distinguishable attributes L1, L2, L3, etc. A classification model based upon the any sensed data for each item can be used to determine key driver attributes.
• Atstep416, suitable substitutes with more desirable attributes are identified by the server and are recommended to replace the underperforming items. A list of substitutes can be identified and ordered based on past sales. Comparing important attributes L1, L2, . . . etc. between the highest and lowest selling items can give the causal levels of the important attributes which can cause the variation in item performance. For example, item B1 can be the highest selling good with higher levels of attributes L11, L21, etc. and item B5 can be the lowest selling good with higher attribute levels L15, L25, etc. Low performing items can be substituted by those items which have high VUB scores and the causal levels of important attributes which cause high performances. In the above example item B5 can be substituted with another item, B7, with high attribute levels L11, L21, L35, and L45.
• Atstep418, if price is the major driver for product choice, suitably modified prices are recommended for underperforming products. If L1=price and the remaining attributes are less important, then decreasing the price of the low performing item can lead to better sales of the item. Items with high similarity scores should be placed adjacently to increase the cross-sales.
• Atstep420, assortment, pricing and product placement recommendations are generated by the server and sent regularly to store manager and category managers via mobile application alerts.
• In parallel to the process described above, atstep422, the server identifies the next items considered after each successful purchase. The server utilizes the next considered items in a feedback loop upon the purchase of an item.
• Atstep424, based on the proportion of users viewing the item immediately after buying another one, and a time of decision, the item similarity score is calculated by the server. The similarity score is dependent upon the time frame in which the user removes the item from the shelf, places it in the cart, and views another item.
• Atstep426, based on item similarity, suitable complement items are determined by the server. The items viewed immediately after the removed item, can be used to determine the latent attributes in the suitability scores.
• Atstep428, adjacent placement of complementary items is recommended by the server.
• FIG. 5 is a flowchart illustrating a process determining continuous detection of a user according to an exemplary embodiment.
• Atstep502, an array of sensor modules distributed along a shelving unit in a facility are controlled according to a first operation mode by default in which recordation of sensor data acquired by a sensing element in each of the sensor modules is disabled and a wireless transmitter in each of the sensor modules is disabled. The array of sensor modules, when inactive, exist in a low power state or sleep state.
• Atstep504, in response to detection of a user within range of a first one of the sensor modules, the first one of the sensor modules in the array transitions from the first operation mode to a second operation mode in which recordation of the sensing data acquired by the sensing element is enabled and the wireless transmitter is enabled. The first sensor modules activates the corresponding sensing element and begins recording data from the sensing element. Additionally, the first user is identified, from the sensing data, removing at least one or more of the items from the shelving unit. A preference score is calculated for each item using the sensing data and whether the user removes the item from the shelving unit.
• Additionally, a message request can be transmitted from sensor modules to the server. The server transmits a threshold time to the sensors on receipt of the message request. Another message can be transmitted from the sensor modules to the first one of the sensor modules upon transitioning to the second mode of operation. In response to receipt of the message, the first one of the sensor modules transitions to the third operation mode in which recordation of sensing data is enabled and the wireless transmitter is disabled.
• Atstep506, a first message is transmitted from the first one of the sensor modules to a server in selective communication with the sensor modules, the server configured to receive the first message from the first one of the sensor modules and to transmit a threshold time to the first one of the sensor modules based on receipt of the first message. The threshold time corresponds to a time long enough in duration to determine that a user is interested in a product located adjacent or near the first one of the sensor modules. The threshold time associated with each item can also be extracted from the sensing data. Based on the preference score, a recommendation is provided to a second user for at least one of the items. The recommendation can include replacing a low-performing item with a different item, rearranging locations of item on the shelving unit or changing the pricing for an item.
• Atstep508, upon transitioning to the second mode of operation, a second message can be transmitted from the first one of the sensor modules to one or more of the sensor modules in the array that are located adjacent to the first one of the sensor modules. The first one of the sensor modules notifies the adjacent sensor modules of an impending change of state.
• Atstep510, in response to receipt of the second message, the one or more sensor modules in the array located adjacent to the first one of sensor modules transition to a third operation mode in which recordation of sensing data is enabled in the one or more sensor modules and the wireless transmitter is disabled. The first sensor module notifies the adjacent modules to begin to record sensing data,
• Atstep512, in response to continuous detection of the first user within range of the first one of the sensor modules for a period of time greater than the threshold time, the sensing data is transmitted from the first one of the sensor modules to the server. Upon the user lingering for the threshold time, the first sensor propagates its recorded data to theserver104.
• Atstep514, in response to continuous detection of the first user within range of the first one of the sensor modules for a period of time less than the threshold time, the first one of the sensor modules reverts back to the first operation mode. As the user can be transiting in front of the one or more sensor modules, the failure to meet the threshold allows the first one of the sensor modules to revert to a low power, sleep or disabled state.
• FIG. 6 is a flowchart illustrating a process for updating inventory records according to an exemplary embodiment.
• Atstep602, sensing data can be received at a first server from one or more of sensor modules in an array in response to detection of one or more users by the one or more of the sensor modules. The sensing data can be transmitted over a wired or wireless connection or through standard networking infrastructure. Alternatively, the sensing data can be received at the server through mesh IoT networking infrastructure, including a Bluetooth low energy (LE) mesh network.
• Atstep604, a continued presence of the one or more users within range of one or more of the sensor modules, for a first period of time, can be detected from the sensing data. The server determines from the sensing data that one or more users are present in the vicinity of the sensor and related items. The sensing data can be optical in nature (e.g., an image), acoustic, or can take the form of weight values take from scales on which users transit.
• Atstep606, the first one of the sensor module can identify, from the sensing data, that at least one of the one or more users is located in front of a first item for a specified amount of time and removes the first item from the shelving unit. Similar to as described above, the sensing data can persist for a period greater than the specified amount of time, and the sensing elements can detect a change relative to their sensing methods (e.g., change in weight on a scale corresponding to an item being added to a cart).
• Atstep608, an inventory record in a database can be updated via a second server in response to a message from the first server corresponding to the removal of the first item, the inventory record corresponding to the first item. The second server can take the form of an inventory server. The inventory server facilitates the procurement of additional product stock. The inventory server monitors the stock of the items in store, as well as stock in the distribution centers, and peer retail stores. Additionally the inventory server can procure stock from the vendor of the items. The inventory server can determine a probability that the first item is viewed and ultimately removed from the facility. The inventory server can update the record corresponding to the item when the probability is determined to be greater than a threshold. Alternatively, the server can monitor the POS terminals to identify transactions corresponding to the item after the removal from the shelving unit. The server can determine whether the time between the users interest, and the removal satisfies a threshold, and then determine the shortest period of time between the continuous detection of the user within range of the sensor and the point of sale transaction.
• Atstep610, the second server can determine whether the inventory record satisfies an inventory threshold in response to the inventory record being updated. Once the inventory record is depleted of sufficient amount to sustain adequate stock, the inventory server can take action to remedy the stock.
• Atstep612, the second server can generate a replenishment signal to notify a third party system to provide a quantity of the first item to the facility in response to determining that the inventory record satisfies the inventory threshold. The inventory server can notify a peer store, distribution center or vendor to prepare an order of the item. The inventory threshold can be determined based in part on the capacity for transport of the items or a transport time.
• FIG. 7 is a block diagram illustrating an electronic device operable to support a system for inventory management according to an exemplary embodiment.
• Computing device700 process product scoring and inventory management. Thecomputing device700 includes one or more non-transitory computer-readable media for storing one or more computer-executable instructions or software for implementing exemplary embodiments. The non-transitory computer-readable media can include, but are not limited to, one or more types of hardware memory, non-transitory tangible media (for example, one or more magnetic storage disks, one or more optical disks, one or more flash drives, one or more solid state disks), and the like. For example,volatile memory704 included in thecomputing device700 can store computer-readable and computer-executable instructions or software for implementing exemplary operations of thecomputing device700. Thecomputing device700 also includes configurable and/orprogrammable processor702 for executing computer-readable and computer-executable instructions or software stored in thevolatile memory704 and other programs for implementing exemplary embodiments of the present disclosure.Processor702 can be a single core processor or a multiple core processor.Processor702 can be configured to execute one or more of the instructions described in connection withcomputing device700.
• Volatile memory704 can include a computer system memory or random access memory, such as DRAM, SRAM, EDO RAM, and the like.Volatile memory704 can include other types of memory as well, or combinations thereof.
• A user can interact with thecomputing device700 through adisplay710, such as a computer monitor, which can display one or more graphical user interfaces supplemented by I/O devices708, which can include a multi-touch interface, a pointing device, an image capturing device and a reader.
• Thecomputing device700 can also includestorage706, such as a hard-drive, CD-ROM, or other computer-readable media, for storing data and computer-readable instructions and/or software that implement exemplary embodiments of the present disclosure (e.g., applications). For example,storage706 can include one or more storage mechanisms for storing information associated with viewed and ultimately bought scores and latent attributes and can be indexed accordingly. The storage mechanism can be updated manually or automatically at any suitable time to add, delete, and/or update one or more data items in thedatabases110A,110B.
• Thecomputing device700 can include anetwork interface712 configured to interface via one or more network devices with one or more networks, for example, Local Area Network (LAN), Wide Area Network (WAN) or the Internet through a variety of connections including, but not limited to, standard telephone lines, LAN or WAN links (for example, 802.11, T1, T3, 56kb, X.25), broadband connections (for example, ISDN, Frame Relay, ATM), wireless connections, controller area network (CAN), or some combination of any or all of the above. In exemplary embodiments, thenetwork interface712 can include one or more antennas to facilitate wireless communication between thecomputing device700 and a network and/or between thecomputing device700 and other computing devices. Thenetwork interface712 can include a built-in network adapter, network interface card, PCMCIA network card, card bus network adapter, wireless network adapter, USB network adapter, modem or any other device suitable for interfacing thecomputing device700 to any type of network capable of communication and performing the operations described herein.
• FIG. 8 is a block diagram of an example embodiment of thesensor module102A to support the distributed sensor system according to an exemplary embodiment. Each sensor module in the array of sensor modules described herein can be implemented according to the example embodiment of thesensor module102A. Thesensor module102A includes amicroprocessor802 communicatively coupled to amemory804, abattery816 to provide power to thesensor module102A, a first power level (FPL)radiofrequency communication module806, and a second power level (SPL)radiofrequency communication module812. The FPLradiofrequency communication module806 and the SPLradiofrequency communication module812 are configured to transmit and receive data at different power levels to control a range of communication with theserver104 andother sensor module102A, as described herein.
• The FPLradiofrequency communication module806 includes areceiver808 configured to receive signals at a first power level via wireless communication transmitted from theserver104 via one or more wireless access points and convert the signals to electrical signals that can be read by themicroprocessor802. The FPLradiofrequency communication module806 further includes atransmitter810 configured to transmit signals at a first power level via wireless communication through one or more wireless access points to theserver104. In the example embodiment, the wireless communication between an electronic shelf label and the wireless access point(s) uses a Wi-Fi communication protocol with the first power level. In some embodiments, thereceiver808 andtransmitter810 are combined into a transceiver unit.
• The SPLradiofrequency communication module812 includes areceiver814 configured to receive signals at a second power level via wireless communication transmitted from one or more of thesensor modules102. Thereceiver814 can convert the signals to electrical signals that can be read by themicroprocessor802. The SPLradiofrequency communication module812 further includes atransmitter816 configured to transmit signals at a second power level via wireless communication. The SPLradiofrequency communication module812 is configured to transmit data at the second power level using at least one of low level Wi-Fi, Bluetooth, low energy Bluetooth, near field communication (NFC), or RFID capabilities. In an exemplary embodiment, thetransmitter816 can be configured to transmit signals at the second power level through RFID connections to one ormore sensor modules102A. In some embodiments, thereceiver814 andtransmitter816 are combined into a transceiver unit.
• While the non-limiting example embodiment shown inFIG. 8 includes the first power levelradiofrequency communication module806 and the second powerlevel communication module812, exemplary embodiments of the electronic shelf labels can include a single communication module that is configured to selectively transmit data/information at the first power level and to selectively transmit data/information at the second power level. For example, in exemplary embodiments, a communication module of asensor module102A can be configured to selectively transmit data/information to the wireless access point(s) at the first power level and to selectively transmit data/information to other electronic shelf labels at the second power level. In some embodiments, the frequency at which the data/information is transmitted is identical for transmissions at the first and second power levels. In some embodiments, the frequency at which the data/information is transmitted is different for transmissions at the first and second power levels.
• Thesensor module102A can further include at least onesensor818 controlled by themicroprocessor802. For example, thesensor818 may include optical sensors, acoustic sensors, weight sensors, radio-frequency identification (RFID), inductance sensors, capacitive sensors, cameras (e.g., CMOS imagers or CCD sensors). In one embodiment, thesensor module102A includes a sensor driver, and themicroprocessor802 is arranged to control updates of thesensor818 by controlling the sensor driver.
• Themicroprocessor802 of thesensor module102A is configured to receive programming instruction from theserver104 or another one of the sensor modules to configure thesensor module102A to detect activity in front of thesensor818, such as a customer viewing a product on the shelf associated with thesensor module102A. More particularly, themicroprocessor802 is configured to process the received programming instructions and relate the instructions to the operation of thesensor818.
• Thesensor module102A can operate in a sleep mode of operation, an intermediate mode of operation, and an active mode of operation. In the sleep mode of operation, themicroprocessor802 andcommunication module806 are in a low power state (e.g., drawing micro-Amps), while thesensor818 is operation to detect objects and theSPL communication module812 is functional to receive messages from other sensor modules, but not to transmit messages. In the intermediate mode of operation, themicroprocessor802 andcommunication module806 are in a low power state (e.g., drawing micro-Amps), while thesensor818 is operating to detect objects and thecommunication module812 is functional to transmit and receive messages. In the active mode of operation, each component of thesensor module102A is fully operational/functional to perform the functions described herein with respect to the active mode.
• Themicroprocessor802 can operate in the sleep mode for an indefinite period of time, and can periodically transition to operate in an awake/active mode (activate from the sleep mode) to identify whether one of thesensor modules102A has detected a consumer within sensing area. Alternatively, or in addition, an output of the sensor can be connected to the microprocessor, and a change in the output of the sensor can cause the microprocessor to transition to the active mode. The microprocessor can transition from the sleep mode of operation to the intermediate mode of operation in response to an adjacent/neighboring sensor module detecting an object/user and transmitting a message to the sensor module, which can be received via theSPL communication module812, which can receive the message and output it to the microprocessor.
• When thesensor module102A is in the sleep mode operation, and thesensor818 detects an object/user, the microprocessor transitions from the sleep mode to the active mode. In response to transitioning to the active mode, the microprocessor can control theSPL communication module812 to transmit a message at the second power level to other adjacent or neighboring sensor modules (e.g., as defined by those sensor modules within range of the transmission from thecommunication module812 at the second power level or those sensor modules having specified addresses/identifiers). The neighboring/adjacent sensor modules can receive the message and transition from the sleep mode to the intermediate mode. When in an active mode, themicroprocessor802 controls the communication channels, for example, by controlling when to send transmissions toother sensor modules102A and receive data from theserver104 including data for other sensor modules. In this way, sufficient processor capacity is available when needed, for example, for information updates. At the same time, since themicroprocessor802 is only active and consuming power when it is needed (e.g., to discharge the duties of an actively detecting sensor), excess or unnecessary power consumption is minimized. After a predefined period of time or after the sensor module no longer detects a consumer, themicroprocessor802 returns to sleep mode and power consumption will be reduced.
• In describing exemplary embodiments, specific terminology is used for the sake of clarity. For purposes of description, each specific term is intended to at least include all technical and functional equivalents that operate in a similar manner to accomplish a similar purpose. Additionally, in some instances where a particular exemplary embodiment includes multiple system elements, device components or method steps, those elements, components, or steps can be replaced with a single element, component, or step Likewise, a single element, component, or step can be replaced with multiple elements, components, or steps that serve the same purpose. Moreover, while exemplary embodiments have been shown and described with references to particular embodiments thereof, those of ordinary skill in the art will understand that various substitutions and alterations in form and detail can be made therein without departing from the scope of the present disclosure. Further, still, other aspects, functions, and advantages are also within the scope of the present disclosure.
• Exemplary flowcharts are provided herein for illustrative purposes and are non-limiting examples of methods. One of ordinary skill in the art will recognize that exemplary methods can include more or fewer steps than those illustrated in the exemplary flowcharts and that the steps in the exemplary flowcharts can be performed in a different order than the order shown in the illustrative flowcharts.

Claims (20)

We claim:

1. A distributed sensor inventory management system comprising:

an array of sensor modules distributed along a shelving unit in a facility, each of the sensor modules in the array being associated with items on the shelving unit;

a first server in selective communication with the sensor modules, the first server configured to:

receive sensing data from one or more of the sensor modules in the array in response to detection of one or more users by the one or more of the sensor modules;

detect, from the sensing data, a continued presence of the one or more users within range of the one or more of the sensor modules for a first period of time;

identify, from the sensing data, that at least one of the one or more users is located in front of a first item for a specified amount of time and removes the first item from the shelving unit; and

a second server in selective communication with the first server, the second server being configured to:

receive a message from the first server in response to the removal of the first item;

update an inventory record in a database, the inventory record corresponding to the first item;

determine, responsive to the updating of the inventory record, whether inventory record satisfies an inventory threshold; and

generate a replenishment signal, responsive to the determining, to notify a third party system to provide of a quantity of the first item to the facility.

2. The system ofclaim 1, wherein the second server is configured to:

determine a probability that the first item is viewed and ultimately removed from the facility; and

update the inventory record in a database, responsive to the determining, when the probability is greater than a probability threshold;

3. The system ofclaim 2, wherein the server configured to:

monitor an operation of a plurality of point-of-sale terminals to identify point of sale transactions corresponding to the first item during a second period of time after detection of the removal of the first item from the shelving unit;

determine whether the second period of time satisfies a first time threshold;

identify, responsive to the determining, a second time threshold corresponding to a shortest period of time of continuous detection of the one or more users within range of the first one of the sensor modules and a point of sale transaction within the first threshold.

4. The system ofclaim 1, wherein the third party system is associated with a peer retailer, a distribution center, or a vendor.

5. The system ofclaim 1, wherein the each of the sensor modules includes a near-field communications transmitter for intercommunication between neighboring ones of the sensor modules in the array.

6. The system ofclaim 1, wherein a spacing between the sensor modules on the shelving unit is determined by an average dimension of items in a category corresponding to the items.

7. The system ofclaim 1, wherein the inventory threshold is based at least in part on a capacity for transport and a transport time.

8. An inventory management method in a distributed sensor system comprising:

receiving, at a first server, sensing data from one or more of sensor modules in an array in response to detection of one or more users by the one or more of the sensor modules;

detecting, from the sensing data, a continued presence of the one or more users within range of one or more of the sensor modules, for a first period of time;

identifying, from the sensing data, that at least one of the one or more users is located in front of a first item for a specified amount of time and removes the first item from the shelving unit;

updating, an inventory record in a database via a second server in response to a message from the first server corresponding to the removal of the first item, the inventory record corresponding to the first item;

determining, via the second server, whether the inventory record satisfies an inventory threshold in response to the inventory record being updated; and

generating, via the second server, a replenishment signal to notify a third party system to provide a quantity of the first item to the facility in response to determining that the inventory record satisfies the inventory threshold.

9. The method ofclaim 8, wherein updating the inventory record comprises:

determining a probability that the first item is viewed and removed from the facility; and

updating the inventory record in the database, responsive to the determining, when the probability is greater than a probability threshold;

10. The method ofclaim 9, wherein determining the probability comprises:

monitoring an operation of a plurality of point-of-sale terminals to identify transactions corresponding to the first item during a second period of time after detection of the removal of the first item from the shelving unit;

determining whether the second period of time satisfies a first time threshold;

identifying, responsive to the determining, a second time threshold corresponding to a shortest period of time of continuous detection of the one or more users within range of the first one of the sensor modules and a point of sale transaction within the first threshold.

11. The method ofclaim 8, wherein the third party system is associated with a peer retailer, a distribution center, or a vendor.

12. The method ofclaim 8, wherein each of the sensor modules includes a near-field communications transmitter for intercommunication between neighboring ones of the sensors modules in the array.

13. The method ofclaim 8, wherein a spacing between the sensor modules on the shelving unit is determined by an average dimension of items in a category corresponding to the items.

14. The method ofclaim 8, wherein the inventory threshold is based at least in part on a capacity for transport and a transport time.

15. A non-transitory computer readable medium, having stored thereon, instructions that when executed in a computing system, cause the computing system to perform operations comprising:

receive sensing data from one or more of sensor modules in an array in response to detection of one or more users by the one or more of the sensor modules;

detect, from the sensing data, a continued presence of the one or more users within range of the one or more of the sensor modules, for a first period of time;

identify, from the sensing data, that at least one of the one or more users is located in front of a first item for a specified amount of time and removes the first item from the shelving unit;

update an inventory record in a database, the inventory record corresponding to the first item;

determine, responsive to the updating of the inventory record, whether inventory record satisfies an inventory threshold; and

generate a replenishment signal, responsive to the determining, to notify a third party system to provide a quantity of the first item to the facility.

16. The computer readable medium ofclaim 15, wherein the instructions to update an inventory record comprises instructions to:

determine a probability that the first item is viewed and removed from the facility; and

update the inventory record in the database, responsive to the determining, when the probability is greater than a probability threshold;

17. The computer readable medium ofclaim 16, wherein the instructions to determine a probability comprises instructions to:

monitor an operation of a plurality of point-of-sale terminals to identify transactions corresponding to the first item during a second period of time after the detection of the removal of the first item from the shelving unit;

determine whether the second period of time satisfies a first time threshold;

identify, responsive to the determining, a second time threshold corresponding to a shortest period of time of continuous detection of the one or more users within range of the first one of the sensor modules and a point of sale transaction within the first threshold.

18. The computer readable medium ofclaim 15, wherein the third party system is associated with a peer retailer, a distribution center, or a vendor.

19. The computer readable medium ofclaim 15, wherein each of the sensor modules includes a near-field communications transmitter for intercommunication between neighboring ones of the sensors modules in the array.

20. The computer readable medium ofclaim 15, wherein a spacing between the sensor modules on the shelving unit is determined by an average dimension of items in a category corresponding to the items.

Images