Quantum mode for centralized and cloud-based real-time shelf merchandise inventory monitoring system

Abstract

The invention provides a quantum system of a cloud-based centralized real-time commodity inventory monitoring system. The system comprises a plurality of layers or shell layers [alpha] p, [alpha] a, [alpha] d, [alpha] e, [alpha] c, and [alpha] cl that represent perception, analog signals, digital signals, Ethernet processing, cloud databases, and client applications and processes, respectively. Connection [beta] pa, [beta] ad, [beta] de, [beta] ec, and [beta] ccl respectively represent (1) analog signal generation, (2) analog signal to digital signal conversion, (3) signal processing and real-time inventory data generation in Ethernet, (4) data transmitted from Ethernet space to a cloud database, and (5) client applications and processes that send data and service requests to the cloud database over a common API. And the connections are respectively expressed according to a sequence from low to high. The node is composed of alpha p, alpha a and alpha d. Multiple layers or shell layers form a quantum structure for data generation and communication. An embodiment of a shelf system composed of independent tracks with photoresistors is used to show how the system monitors real-time merchandise inventory on a shelf.

Description

Copyright Notice

[0001] Portions of the disclosure of this patent document contain material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever. Field of the Invention

[0002] This document discloses product inventory monitoring and the collection and management of inventory data. Real-time shelf inventory is a critical issue in the retail industry. Cash registers or self-checkout carts with built-in cameras and scales record items through a scanning process, but exclude items that were not scanned due to theft, improper placement, or other factors. Empty shelves result in significant revenue losses. Furthermore, in the e-commerce era, efficient cash flow requires effective inventory levels, not just sufficient inventory levels. Furthermore, the interaction between real-time shelf inventory, sales revenue, and cash flow is a key focus of smart shelf and display systems. These systems involve real-time interaction, centralized and cloud-based systems, cashierless stores, driverless supply chains, learning processes, the Internet of Things (IoT), fifth- or sixth-generation (5G or 6G) networks, blockchain, robotics, warehouse automation, new chips, sensors, or sensing devices, big data, data analytics and data mining, just-in-time delivery, and other artificial intelligence (AI) technologies. These technologies represent a new era in smart shelf / network / cloud systems. This document addresses the characteristics of these technologies and the system architecture and design.

[0003] The following papers briefly introduce the latest technologies for e-commerce applications in retail chains: (1) Distributed Computing and Artificial Intelligence, edited by Kenji Matsui et al., DCAI, 2021, states in the preface that "distributed computing plays an increasingly important role in modern signal / data processing, information fusion, and electronic engineering (e.g., e-commerce, mobile communications, and wireless devices). In particular, the application of artificial intelligence in distributed environments... for the Internet of Things, Industrial Internet of Things (IIoT), big data, blockchain... from personal laptops to edge / fog / cloud computing systems for parallel and distributed computing"; (2) "A Review of Cloud Computing Trends and Applications in Healthcare Ecosystems", Mbasa Joaquim Moto et al., Volume 2021, ID1843671, describes the new challenges and opportunities facing the Internet of Things, edge computing, fog computing, and cloud computing; (3) "Digitalization of the World from Edge to Core", David Reinsel et al., November 2018, describes the growing global data and digital transformation capabilities; (4) From June 2014 to September 2015, Jurgen A German team led by Sturm uses robotics to study indoor navigation and virtual shopping; (5) "Multi-task Learning with Sequence-Conditioned Transport Networks", to be published in IEEE in 2022, describes an end-to-end vision-based system architecture and sequence-conditional transport networks that significantly improve pick-and-place performance on a novel 10-task benchmark problem; (6) "Implicit Behavior Cloning", Adrian Wong et al., CORL, 2021, reveals that robots with implicit policies can learn complex and very subtle behaviors for contact-rich tasks from human demonstrations, including high combinatorial complexity with 1 mm accuracy; (7) "Faster R-CNN: Towards Real-Time Object Detection via Region Proposal Networks", Shaoqing Ren, Kaiming He et al., (arXiv:1506.01497, June 4, 2015 (v1), last revised January 6, 2016 (v3), shows that a trained end-to-end RPN can generate high-quality region proposals. Background of the Invention

[0004] The approaches described in this section are approaches that could be pursued, but not necessarily approaches that have been previously conceived or pursued. Therefore, the approaches described in this section may not be prior art to the claims in this application and are not admitted to be prior art by inclusion in this section.

[0005] US 11,222,306 relates to a method for real-time shelf inventory monitoring using sensors or photoresistors under light.

[0006] US 11,064,816 relates to a discrete gravity-fed commodity propulsion system assembled for use in a real-time on-orbit commodity inventory monitoring system.

[0007] US 8,376,154 relates to a gravity-fed rolling rack system that is used by robots for merchandise replenishment.

[0008] US 10,660,435 relates to an indoor refrigerated rack / shelf system for a real-time shelf merchandise inventory monitoring system.

[0009] US 9,420,901 relates to a power supply, typically for use in low voltage plug-and-play display systems.

[0010] US 9,375,098 relates to a weight-propelled rolling rack assembly, the system being used for merchandise inventory measurement. Summary of the Invention

[0011] The invention provides a quantum system for a cloud-based centralized real-time commodity inventory monitoring system. The system comprises multiple layers or shells α p , α a , α d , α e , α c and α cl , representing perception, analog signal, digital signal, Ethernet processing, cloud database, and client application and process respectively. pa , β ad , β de , β ec and β ccl They represent (1) analog signal generation, (2) analog signal to digital signal conversion, (3) signal processing and real-time inventory data generation in Ethernet, (4) data transmission from Ethernet space to cloud database, and (5) client applications and processes that send data and service requests to cloud database through public API. These connections are represented in order from low to high. The nodes are represented by α p , α a and α d Multiple layers or shells form a quantum structure for data generation and communication. An example of a shelving system consisting of independent tracks with photoresistors is used to demonstrate how the system can monitor real-time inventory on shelves.

[0012] This document also discloses a centralized, cloud-based system for real-time monitoring of merchandise on shelves. The system, as an embodiment, includes: The racking system has: (a) The shelf has: Tracks, on which goods are movably placed; A main control board having a network port for communicating with an in-store server (referred to herein as a control server); A plug-and-play system, including data and power conductors, for communicating information between the main control board and the track; (b) a plurality of photoresistors disposed adjacent to the track, wherein the photoresistors generate analog signals indicating the number of items placed on the track at any given moment; (c) a unique identifier associated with at least one component selected from the group consisting of: a photoresistor, a group of photoresistors, a track, a main control board, a plug-and-play system, and a shelf; (d) a plurality of binary decoders disposed on the track for polling the analog signal from each photoresistor; and (e) a microcontroller unit (MCU) disposed on the track, the MCU having a track application program interface (API), an analog-to-digital converter port for analog signal input, and an analog-to-digital signal converter, wherein the main control board is disposed on the shelf, for the shelf to be connected to Ethernet networking and to the track via a network bus and a power bus through a plug-and-play system; a control server connected to the shelf through a network port of the shelf main control board, wherein the control server sends a request to scan each photoresistor to the MCU through the shelf main control board, and then the MCU obtains an analog signal from the photoresistor according to the addressing of the binary decoder, and then the MCU converts the analog signal into a digital signal through the analog-to-digital signal converter, so that the digital signal is transmitted to the main control board through the plug-and-play system, and then transmitted to the access control server through the network port, The control server includes a system that processes data and writes the processed data to a cloud database, and client applications and processes (such as RT-POG) can access the processed data from the cloud database through an API.

[0013] This document also discloses a system for tracking merchandise on a shelf, the system comprising: (a) a propulsion device on which merchandise is placed and which moves the merchandise toward the front of the shelf by gravity; (b) a photosensor that generates an analog signal having an amplitude that is indicative, over a continuous range, of the intensity of light (ambient or non-ambient) incident thereon, wherein the intensity of the light depends on the position of the article relative to the photosensor; (c) a circuit that converts the analog signal into digital data, the value of the digital data indicating the amplitude; and (d) “Control Server”, where: (i) receiving digital data from the circuit; (ii) processing the digital data and the digital data indicating the intensity of light incident on other light-sensitive elements on the shelf to produce a processed amplitude; (iii) determining the number of products on the shelf and the shelf inventory percentage based on the processed measurement value and mapping the photosensitive element to the product size; (iv) The processed values, quantities, and percentages are written to a central cloud database, and client applications and processes can access the processed data through APIs.

[0014] This document discloses a centralized, cloud-based system for real-time monitoring of items on shelves and tracking them in the form of a quantum system: (a) The product is movably placed on a track. The photoresistor is used as a sensor to sense the layer / shell α p the existence of the Middle Light; (b) In layer / shell α a Generate an analog signal on the track to indicate the number of goods placed on the track at any time; (c) Layer / shell α p and layer / shell α a The connection between them is the generative connection β pa ; (d) The analog signal on the track is converted into a digital signal in layer / shell αd, wherein a plurality of binary decoders arranged on the track poll the analog signals of the respective photoresistors; a microcontroller unit (MCU) arranged on the track has a track application program interface (API), an analog-to-digital conversion port for inputting analog signals, and an analog-to-digital converter; (e) Layer / shell α a The connection between the layer / shell αd is the conversion connection β a d; (f) The main control board is set on the rack for layer / shell α e The rack Ethernet networking in the,and rail connection is carried out through the network bus and power bus through the plug and play system,,that is, the processing connection βde; Among them, the MCU has converted the analog signal into a digital signal through an analog signal to digital signal converter, thereby transmitting the digital signal to the main control board through a plug-and-play system; (g) The main control board set on the rack also passes the Ethernet layer / shell α e The network port of the rack main control board is connected to the control server; (h) In addition, the main control board installed on the rack is also connected to the camera, access point (AP) and / or 5G / 6G extender through the network port to connect to the Ethernet layer / shell e electronic devices in; (i) The control server sends a request to the MCU through the main control board to scan each photoresistor on each track on the shelf. The MCU obtains the analog signal from the photoresistor and is addressed by the binary decoder, where the main control board is connected to the Ethernet layer / shell α e The network port in sends the digital signal to the control server. The control server is a system that processes the signal and converts it into processed data, and then writes the processed data to the layer / shell α c In the cloud database, client applications and processes (such as RT-POG) can connect to the cloud through the β cc l and β ec Processed data is accessed from the cloud database via a public HTTP API. (j) The control server is a physical device in the edge layer, and the associated network is the edge network, where edge computing occurs locally where the data is generated, right at the "edge" of a given application network. Fog computing is a layer between cloud computing and edge computing. The fog computing space can receive edge layer data before it reaches the cloud layer and can selectively filter irrelevant data, analyze the data for remote access, or inform localized learning models. (k) Client applications and processes send requests to the cloud database through the public API in the API layer / shell. αAPI connects to the cloud β ec and β cc l Connect to Ethernet layer / shell α e The control server program in the store, the signal handler or client application (such as RT-POG), such as Figure 0 shown. Attachment picture illustrate

[0015] Figure 0 It is a quantum system for centralized, cloud-based, real-time shelf merchandise monitoring systems.

[0016] Figure 1 A diagram of a centralized, cloud-based, real-time shelf monitoring system picture .

[0017] Figure 1A yes Figure 1 The racking system box picture .

[0018] Figure 1B yes Figure 1 Schematic diagram of the system picture , shows the process from the analog signal generated by the photoresistor to the processed data.

[0019] Figure 2 yes Figure 1 The process flow of the system execution picture, including two blocks, one block represents the shelf and the other block represents the operations in the control server.

[0020] Figure 2A It is the process of tracking the execution of the API when controlling the server polling picture .

[0021] Figure 3 Refers to the forms and processes for the global identification code of the global retail system picture .

[0022] Figure 4 The unique Base-62 global identification code of the real-time inventory system component is shown.

[0023] Figure 5 is an isometric view of the photoresistor array on the shelf rail picture

[0024] Figure 6 Is the system network topology picture (Networking 600).

[0025] Figure 7 is the function of the binary decoder picture .

[0026] Figure 8 Is the function of electrical equipment on the shelf picture

[0027] Figure 9 Is the program that controls the server picture .

[0028] Figure 10 Exploded top view of roller track with printed circuit board picture , which shows the photoresistors of the printed circuit board distributed and connected along the length of the track.

[0029] Figure 11 Exploded side view of the track and main control panel with plug and play system picture .

[0030] Figure 12 yes Figure 1 Details of the plug-and-play system in the system picture .

[0031] Figure 13 It is the bottom view of the track picture .

[0032] Figure 14 It is the frame of the shelf network picture .

[0033] Figure 15 The bottom view of the shelf picture .

[0034] Figure 16 It is the track navigation for robots to transport goods to the track picture Show.

[0035] Figure 17 It is planned picture Represents real-time shelf inventory.

[0036] Figure 18 It is real-time layout picture Demonstration.

[0037] Figure 19 is a report on shelf inventory changes over time.

[0038] Figure 20 Reports on stock changes over time for several typical time-day periods. Figure 21 Real-time inventory on shelves nationwide picture . DETAILED DESCRIPTION

[0039] Figure 0 Quantum Systems is a cloud-based centralized real-time store inventory monitoring system.

[0040] Quantum systems are used to simulate the processes involved in inventory monitoring, from analog signal generation to data transmission and communication. In the nervous system, sensory receptors process stimuli, nerve impulses, and action potentials to complete sensory transmission. Simply put, sensory neurons, or afferent neurons, are neurons in the central nervous system, where the nerve endings and afferent nerve fibers of sensory nerves form the receptors.

[0041] Based on this concept of neural system, an inventory monitoring system is built that includes a sensor to sense the presence of goods. Then, the sensor is connected through the connection β in the layer / shell αa. pa Produces a certain amount of value, thus generating an analog signal. Furthermore, the analog signal passes through the connection β in the layer / shell αd a d is converted into a digital signal. The node is defined as including the sensor α p , simulated layer / shell α a and the "electronic unit" of the conversion layer / shell αd. In this model, Figure 0 There is at least one node in .

[0042] Based on the above description, the node has a digital signal that is transmitted to a specific device in the Ethernet space or environment via a connection βde as a transmission. Shelves, shelf systems, and other displays are located in this Ethernet space. The signal is transmitted within the Ethernet space to the local control server, and each store is equipped with at least one local control server. The local control server provides the digitized signal from the shelf to the signal processing program, which converts these signals into real-time inventory data. Finally, the real-time inventory data is transmitted through the internal API via the connection βde. ec Transfer from signal handler to cloud database.

[0043] In layer / shell αd, i.e., in the Ethernet space, there are two sub-layers / shells, namely, edge layer / shell and fog layer / shell, which are used to transmit data from physical devices in layer / shell αd. Usually, the edge server will first send data to the fog layer through the local network to decide whether it is worth sending to the cloud to reduce traffic, especially for complex information or large domains, such as picture Images or videos can be captured without sacrificing bandwidth and latency. Edge computing and fog computing are used in cloud databases to store and process relevant data with remarkable efficiency through cloud computing. An edge layer / enclosure is required, but a fog layer / enclosure is not necessary.

[0044] Additionally, the client layer / shell α c l is located in the cloud layer / shell α c In addition, both layers / shells are connected by β cc l Linked via a communication protocol (such as HTTP).

[0045] The two APIs, the internal API and the public API, can be represented by concentric circles in the cloud layer / shell. The internal API communicates with the handlers of the Ethernet layer / shell, while the public API communicates with the client layer / shell.

[0046] Requests from client applications and processes are sent to the cloud database through a public API.

[0047] Requests to read and write to the database from the Ethernet layer's handlers are sent to the cloud database via an internal API.

[0048] Layer / Shell α p , α a ,αd,α e , α c and α c l represents perception, analog signal, digital signal, Ethernet processing, cloud database, and client application and process respectively.

[0049] Connect Beta pa , β a d, βde, β ecand β cc l represents (1) analog signal generation, (2) analog signal to digital signal conversion, (3) signal processing and real-time inventory data generation in Ethernet, (4) data transmission from Ethernet space to cloud database, and (5) client applications and processes sending data and service requests to cloud database through public API. These connections are represented from low to high.

[0050] Multiple layers / shells act as quantum structures, operating from low to high levels, represented by signals, data, and data quality. The cloud database, serving as data storage, represents the highest level of connectivity. Below this layer is Ethernet with local servers for data creation, manipulation, and communication. Nodes are the basic units for sensing and generating analog signals and processing data. The client layer / shell consists of six layers / shells, five connections, and three sublayers / shells. The cloud database, serving as data storage, connects to the cloud through β ec with Ethernet layer / shell alpha e It communicates with servers in the Apache ActiveMQ Artemis server and satisfies requests from client applications and processes through a public API.

[0051] Sensing and connectivity can consist of any device made of biological, organic, chemical, electronic, electrical, thermal, magnetic, acoustic, mechanical or other elements / materials, using non-AI techniques or AI techniques such as CNN (Convolutional Neural Network).

[0052] The structure can be expressed by mathematical formulas. They are: (α p +α a +αd) = node (e unit); STRU=ε stru (Node,PCB,ECP,PS,FM) Among them, STRU represents the structure of the device, including nodes, PCB (printed circuit board) and other elements: ECP stands for electronic / electrical components. PS stands for power supply, FM stands for firmware.

[0053] The networking and processing procedures in an Ethernet environment are expressed as follows: in, ID is defined as an assigned identification number. SP is a software process for sensor detection, conversion and additional processing of data, as well as communication with local networks and cloud databases. EC stands for edge computing. FC stands for fog computing. CC stands for cloud computing. API is an application programming interface, including public APIs and internal APIs. An APP is an application program that resides on a computing device (e.g., laptop, mobile device, etc.). CMS is a centralized and cloud-based monitoring system.

[0054] The network NT is described as NT=μnt(ID,Stru,Con,SV,OS,ES) in, Stru is a structure built from nodes, circuit boards and other elements. Con is the communication link, SV is the server, OS stands for operating system, PS stands for power supply, and ID stands for identification code.

[0055] The following example illustrates the Quantum system using a centralized, cloud-based, real-time inventory monitoring system for store shelf systems built with electronic and electrical components.

[0056] Figure 1 、 1A , 1B, 2 and 2A are blocks describing the above real-time inventory monitoring system picture .These picture Shown are the above mentioned perceptions, layers / shells and connections.

[0057] Figures 3 to 16 The inventory monitoring system of this embodiment is described in detail. Figures 17 to 20 The results of real-time inventory counting, i.e. RT inventory counting on the shelves, are shown.

[0058] Shelf layout picture (POG) is a picture A table or model used to indicate where retail products (i.e., merchandise) should be placed on a shelf to maximize sales. POGs typically display the product, brand, specifications, weight, price, advertising items, location, etc.

[0059] A real-time POG (RT-POG) is a POG that displays real-time shelf inventory and inventory / replenishment data and conditions using the POG format. As used herein, "real-time" refers to shelf inventory data and conditions that are collected instantly, or data and conditions that are no more than a few seconds, minutes, or at most an hour old.

[0060] Figure 1 is a framework for a real-time shelf inventory monitoring system (i.e., system 100 as an embodiment). picture .

[0061] System 100 includes (a) a shelf frame 102 and a control server 150 in a store 101, (b) a cloud server 170, (c) multiple client applications and processes 145, and (d) a network system 600 (see Figure 6 ).

[0062] Store 101 is a retail establishment.

[0063] The shelving system 102 is a fixture for displaying items 105 (eg, merchandise) and includes a shelving assembly, namely the shelving system 108 .

[0064] Figure 1A It is the frame of the shelf system 108 picture . picture The components shown in FIG include a racking system 108 . Figure 1 and Figure 1A are generalized representations of the components described in this document and are not drawn to scale but are drawn to illustrate the functional relationships among them.

[0065] The racking system 108 includes a plurality of racks 109 , a representative one of which is designated as a single rack shelf 110 .

[0066] A single shelf layer 110 includes a main control panel 120, plug-and-play components (plug-and-play system 121), and a plurality of rails 115, a representative one of which is designated as rail 115A.

[0067] The main control board 120 is a circuit board including circuits for power management, network management, and serial port to Ethernet conversion.

[0068] The plug-and-play system 121 is used for non-point-to-point communication to achieve high efficiency. The plug-and-play system 121 includes (1) at least four conductive wire channels, such as 4 copper wire channels 1005 (see Figure 10 ), (2) track plug pins corresponding to the number of conductive wires, such as 4 power and data plug pins 1120 (see Figure 11 ), (3) power supply bus 1450 (see Figure 14 ), and (4) a data link based on the connection of N nodes in a multipoint network, such as an RS-485 bus 1420 (see Figure 14 ).

[0069] The track 115A includes a plurality of rollers 111A and a track circuit board 117A.

[0070] Rollers 111A are located on the top surface of track 115A. Rollers 111A are generally gravity-fed propulsion devices on which merchandise 105A (i.e., a subset of merchandise 105) is movably arranged so that gravity forces merchandise 105A to move toward the front of a single shelf shelf 110 so that people can see and access merchandise 105A. Rollers 111A are configured with multiple individual rollers or sliding ribs such that gaps exist between adjacent rollers or sliding ribs, allowing light to pass through the gaps.

[0071] The track's circuit board 117A includes circuitry 122, which in turn includes (a) a plurality of photoresistors 125A, a representative of which is designated as photoresistor 125A-1, (b) a plurality of binary decoders 130A, a representative of which is designated as binary decoder 130A-1, and (c) a microcontroller unit (MCU) 135A, which includes a track application programming interface (API) 136A.

[0072] The photoresistor 125A-1 is a light-sensitive element whose resistance continuously changes according to the amount or intensity of light incident on the photoresistor 125A-1. For example, as the amount of light increases, the resistance of the photoresistor 125A-1 decreases. The output of the photoresistor 125A-1 is an analog signal, such as a current or voltage, where the magnitude of the signal represents the amount or intensity of light incident on the photoresistor 125A-1. Therefore, when the photoresistor 125A-1 is exposed to light, it generates an electrical signal, such as a current or voltage, and the electrical signal is a linearly varying analog signal. The operation of the other photoresistors 125A is similar to that of the photoresistor 125A-1.

[0073] A photodiode can be used as a replacement for photoresistor 125A-1. A photodiode is a semiconductor that converts light into an electric current, where the current varies depending on the amount or intensity of light incident on the photodiode. Photoresistor 125A can be implemented as any light-sensitive element that has a characteristic that continuously changes depending on the intensity of light incident on it.

[0074] Binary decoder 130A polls and communicates with photoresistor 125A.

[0075] The MCU 135A communicates with the binary decoder 130A and controls the binary decoder 130A.

[0076] The other tracks in track 115 are configured similarly to track 115A.

[0077] The track's circuit board 117A is located below the rollers 111A. As described above, gaps exist between adjacent rollers 111A, allowing light to pass through these gaps. The light is sensed by one or more photoresistors 125A located below the gaps. The photoresistors detect the light and generate analog signals indicating the intensity of the light, thereby indicating the area or volume of a single shelf shelf 110, whether the area or volume is occupied or unoccupied by the merchandise 105A on the single shelf shelf 110.

[0078] Reference again Figure 1 .

[0079] The control server 150 includes a control server program 153 .

[0080] The cloud server 170 includes a cloud database 178 , a cloud database API 177 , and an application API 179 , including an RT-POG API and a robot navigation API, etc., accessible to client applications and processes.

[0081] Client applications and processes 145 include, but are not limited to, real-time inventory monitoring applications 145A, robotic algorithms 145B, delivery logistics applications 145C, and data mining and analysis systems 145D, which run on desktop, mobile, wearable and dedicated devices or any authorized device that can access the application API 179 and process the received data.

[0082] In general, cloud technology is suitable for very large data volumes and very efficient high-speed data transactions, such as real-time performance. It also ensures a high level of data normalization, including no data duplication, eliminates data synchronization issues, and allows remote client applications and processes to access data in real time.

[0083] The control server 150 provides the data to the cloud database 178 through the cloud database API 177 .

[0084] The control server program 153 manages the processes of the control server 150 , including polling the track to obtain data, applying thresholds to convert digital signals into covered / uncovered values, mapping photoresistors to match product dimensions, calculating percentages and determining product conditions, and uploading this processed data to the cloud database 178 .

[0085] Cloud server 170 provides data for store 101 to client applications and processes 145. In practice, system 100 may include multiple stores configured similarly to store 101, and thus, cloud server 170 will provide data to client applications and processes 145 for multiple stores.

[0086] Figure 1B yes Figure 1 System diagram picture , shows the process from the analog signal generated by the photoresistor to the processed data.

[0087] A single shelf 110 accommodates N×M photoresistors 125A, where N is the number of photoresistors 125A along the depth of the track 115A, and M is the number of tracks 115 along the width of the single shelf 110. In system 100, N is less than or equal to 768. Therefore, the photoresistors 125A are located on the track wiring board 117A, and the maximum number of photoresistors 125A per track 115A is 768, although in practice, this number can vary depending on the design. If desired, the number of photoresistors 125A can exceed 768.

[0088] As described above, the output of the photoresistor 125A-1 is an analog signal, such as a voltage, the magnitude of which represents the amount or intensity of light incident on the photoresistor 125A-1. Meanwhile, the voltage common collector (VCC) 182 of the MCU 135A is a power supply voltage provided to the MCU 135A and is used to compare with the voltage of the photoresistor 125A-1 as a reference voltage. VCC 180 is a power supply voltage provided to the binary decoder 130A-1. The photoresistor 125A-1 is affected by incident light, causing the output voltage to drop. The relationship between light intensity and voltage output is linear. The power supply voltages of VCC 180 and VCC 182 can be +3.3V, +5V, or other voltages, but the power supply voltages of VCC 180 and VCC 182 are the same.

[0089] The analog signal generated by the photoresistor 125A is converted into a digital signal through a conversion formula 135C, and the binary decoder 130A1 controlled by the MCU 135A performs data processing through a polling process.

[0090] Components of the MCU 135A include a general purpose input / output (GPIO) port 184 and an analog-to-digital converter (ADC) port 186. Additionally, the MCU 135A includes a rail API 136A.

[0091] The input to ADC port 186 is a series of voltage measurements 188 (analog), which are converted from the measurement range of 0-VCC (analog) to 1-4095 (digital) by conversion formula 135C. Therefore, the output of track API 136A is digital signal 138.

[0092] The track API 136A operates upon receiving a read request 134 or command from the control server program 153 and responds to the control server 150 with a digital signal 138 .

[0093] The digital signal is polled and managed by the control server program 153 and further converted into processed data 160 after going through the calibration process 157, the threshold process 159 and the photoresistor mapping process 240A.

[0094] In practice, store 101 will include multiple shelves configured similarly to shelf system 102, and each shelf will include multiple shelves configured similarly to single shelf shelf 110, and each shelf will include multiple tracks configured similarly to track 115A, with rollers, PCBs, photoresistors, binary decoders, and MCUs. In addition, expanded embodiments of system 100 will cover multiple stores configured similarly to store 101.

[0095] Figure 2 Is the process flow of the real-time shelf inventory monitoring system picture , i.e. process 141. To describe process 141, Figure 2 Shown are a main control panel 120, a single shelf shelf 110, a rail API 136A, merchandise 105A, and rails 115A.

[0096] exist Figure 2 There are two blocks in FIG. 1 , one block, block 201 , represents a single shelf layer 110 , and the other block, process 202 , represents an operation in the control server 150 .

[0097] Figure 2 Also shown are (a) rail APIs 136B, 136C, 136D, and 136E, which are similar in operation to rail API 136A, (b) rails 115B, 115C, 115D, and 115E, which are similar in operation to rail 115A, and (c) merchandise 105B, 105C, 105D, and 105E, which are a subset of merchandise 105. APIs 136B through 136E are associated with rails 115B through 115E, respectively, and merchandise 105B, 115C, 115D, and 115E are a subset of merchandise 105. Merchandise 105C, 105D, and 105E are movably disposed on rails 115B, 115C, 115D, and 115E, respectively. Additionally, block 201 shows lighting 250 managed by the rail API 136F for many important applications, including ambient lighting adjustments, alarm indicators when merchandise inventory on shelves falls below a warning level, lighting for cameras to recognize bar codes, applications for disinfection, and the like.

[0098] Process 202 outlines the process of converting analog signals generated by photoresistors mounted on track 115 into digital data in control server 150, as process start 207 of process 202 for ease of description, and further converts the data into covered / uncovered values, maps photoresistors to merchandise items, calculates percentages and determines merchandise sales conditions, and then uploads it to cloud database 178. This data can then be accessed by client applications or processes (e.g., RT-POG) through application APIs 179 (including RT-POG APIs and robot navigation APIs, etc.).

[0099] In system 100, a unique global identification number is assigned to each of the following: (a) a photoresistor, such as photoresistor 125A-1; (b) a photoresistor assembly, such as photoresistor 125A; (c) a track, such as track 115A; (d) a main control board, such as the main control board 120; (e) plug-and-play systems, such as plug-and-play system 121; and (f) Shelves, such as individual shelf shelves 110.

[0100] As described above with respect to block 201, a single shelf layer 110 includes a plurality of rails 115, and on rails 115A, photoresistors 125A are mounted on the rail circuit board 117A. Items 105A, ie, a subset of items 105, are placed on rails 115A.

[0101] On the track 115A, there is a gap between two adjacent rollers 111A. Light can pass through the gap and be sensed by one or more photoresistors 125A located just below the gap, thereby generating an analog signal.

[0102] Track APs 136A, 136B, 136C, 136D, and 136E manage the receipt and response of signals and data communications between the various components of tracks 115A, 115B, 115C, 115D, and 115E and cloud server 170 .

[0103] Under the management of the rail APs 136A, 136B, 136C, 136D, and 136E, the digital signal released from the MCU 135A is transmitted to the main control board 120 through the RS-485 bus by using the plug-and-play system 121 .

[0104] The main control board 120 converts the digital signal transmitted through the RS-485 bus into Ethernet, through which TCP / IP (Ethernet) can be used for communication between the main control board 120 and the control server 150. In addition, the main control board 120 has a 4-port network PoE switch function.

[0105] Figure 8 The display shows that four network ports 805 can be designed: one for input and output, and two for other applications, such as a Power over Ethernet (PoE) wireless access point (AP) 825, a PoE camera 830, and / or a 5G extender (832). Each shelf, such as a single shelf layer 110, has a main control board, such as main control board 120. If each shelf has 6 shelves, there are 12 network ports 805 for input and output, and 12 other network ports 805 for various applications.

[0106] Reference again Figure 2 , process 202 is in the control server 150, which scans the photoresistors of each track 115 on a single shelf layer 110 through the track API, for example, scanning the photoresistor 125A on the track 115A through the track API 136A, and the control server program 153 performs mapping, conversion, formatting processes 240A and 240B, and uploads process 240C to the cloud database 178 through the hypertext transfer protocol (HTTP).

[0107] In the control server 150 , there are an initial process 207 and three process blocks, namely, process blocks 210 , 230 , and 240 .

[0108] Process 207 initiates a scan of the photoresistors of each track on the shelf, for example, scanning photoresistor 125A on track 115A.

[0109] Process block 210 performs track API polling for each track, for example, performs track API polling 136A for track 115A. Each track API is controlled by its corresponding MCU 135, for example, track API 136A is controlled by MCU 135A.

[0110] Process block 230 includes three operations, namely: 115B, 115C, 115D and 115E are similarly configured with rollers and photoresistors. Operation 230A, retrieving a threshold value generated by machine learning (ML); Operation 230B, retrieving an initial value of the photoresistor and adjusting the threshold value; and Operation 230C converts the data from the digital signal into discrete covered / uncovered values.

[0111] Process block 240 executes: Operation 240A, mapping the photoresistor to match the product; Operation 240B, calculating the percentage and determining the condition of the product; and Operation 240C, uploading the data processed in Operations 240A and 240B to the cloud database 178 through the cloud database API 177 .

[0112] Process blocks 210 , 230 , and 240 are components of the data collection network under the control server 150 .

[0113] Finally, the cloud database 178 connects all client applications and processes through application APIs 179 (including RT-POG API and robot navigation API, etc.) for use by chain store headquarters (HQ), brand manufacturers, warehouses, and robots via HTTP.

[0114] In particular, Figure 2 Live product display near the top / center picture Monitor 3500 is the client application used by store managers / stockists.

[0115] Figure 2A This is the process of tracking the process executed by API 136A when the control server 150 polls picture .

[0116] Figure 2A Two processes are shown that are handled by a track API, such as track API 136A, namely, a photoresistor polling process 260 and a lighting adjustment process 270. During execution of process block 210 within control server 150, processes 260 and 270 are handled by track API 136A.

[0117] The photoresistor polling process 260 measures the amplitude of the analog signal as it is affected by the intensity of light shining on the photoresistor 125A on the track 115A, converts the analog signal into a digital signal, and transmits the digital signal to the control server 150 via the main control board 120. On the track 115A, these operations are controlled by the MCU 135A.

[0118] The dimming process 270 adjusts the light to a certain brightness level, such as 85% of the maximum level. Similar to the polling process 260, there are "paired" and "unpaired" states because the LED lights of the track 115A are implemented by the lighting on the single shelf layer 110 above the track 115A, and the lighting of the track 115 and the single shelf layer 110 are "paired" together in the database.

[0119] The photoresistor polling process 260 and the dimming process 270 communicate with the associated client applications and processes using a plug-and-play RS-485 bus and Ethernet TCP / IP. Furthermore, processes 260 and 270 are available in "RS-485 address paired" and "RS-485 address unpaired" versions.

[0120] Figure 2 and 2A Summary.

[0121] Figure 2 and Figure 2A Process 141 , the real-time shelf inventory monitoring system process, is described.

[0122] Figure 2 and 2A Innovations include: (a) Data collection network; (b) Plug and Play system; (c) Photoresistor polling via binary decoder under MCU control; (d) threshold calculation via initialization and machine learning (“ML”); (e) Photoresistor mapping to match the product; (f) Processing data related to calculating product percentages and product conditions, which can be accessed to create many applications and visualizations, including but not limited to inventory reports over time, RT-POG for inventory control, delivery logistics, robotic operations, etc.; (g) RT-POG format; (h) control server program 153; and (i) Generally, the rack can be used for power and data communication processes anywhere, such as a store, warehouse, or office equipped with Ethernet.

[0123] Each of the control server 150 and the cloud server 170 can be implemented on a device including a processor and a memory. A processor is an electronic device composed of logic circuits that respond to and execute instructions. A memory is a tangible, non-transitory, computer-readable storage device encoded with a computer program. In this regard, the memory stores data and instructions, i.e., program code, which can be read and executed by the processor to control the operation of the processor. The memory can be implemented as random access memory (RAM), a hard drive, a read-only memory (ROM), or a combination thereof. The memory in the control server 150 stores program modules for controlling the processor in the control server 150 to perform operations on behalf of the control server 150. The memory in the cloud server 170 stores program modules for controlling the processor in the cloud server 170 to perform operations on behalf of the cloud server 170.

[0124] The term "module" is used herein to refer to a functional operation that can be embodied as an independent component or an integrated configuration of multiple subordinate components. Therefore, a program module can be implemented as a single module or multiple modules that operate in coordination with each other. In addition, a program module can be implemented as software, hardware (e.g., electronic circuitry), firmware, or a combination thereof.

[0125] Program modules may also be configured on a tangible, non-transitory external storage device for subsequent loading into the memory of the control server 150 and the cloud server 170. Examples of such external storage devices include (a) optical disks, (b) magnetic tapes, (c) read-only memory, (d) optical storage media, (e) hard disk drives, (f) a storage unit consisting of multiple parallel hard disk drives, (g) universal serial bus (USB) flash drives, (h) random access memory, and (i) electronic storage devices coupled to the control server 150 and the cloud server 170 via a network.

[0126] Although each of the control server 150 and the cloud server 170 are represented herein as standalone devices, they are not limited thereto and may be coupled to other devices in a distributed processing system.

[0127] Furthermore, although the cloud database 178 is represented herein as a standalone component, it is not limited thereto and may be implemented in multiple storage devices in a distributed database system.

[0128] Data communication between the control server 150, the cloud server 170, and the client applications and processes 145 is carried out over a network, which can be a private network or a public network and can include any or all of the following: (a) a personal area network, such as covering a room, (b) a local area network or wireless local area network, such as covering a building, (c) a campus area network, such as covering a campus, (d) a metropolitan area network, such as covering a city, (e) a wide area network, such as covering an area across a metropolitan area, a regional, or a national border, (t) the Internet, (g) a telephone network or a wireless communication network (e.g., Bluetooth, etc.). Long Term Evolution (LTE) is a standard for wireless broadband communication for mobile devices and data terminals. 5G is the successor to LTE. Communication is carried out over a network using electronic and optical signals that are transmitted and received wirelessly or by wire or fiber.

[0129] As mentioned above, POG is a picture A table or model used to indicate where retail products (i.e., merchandise) should be placed on a shelf to maximize sales, and typically displays the product, brand, specifications, weight, price, advertising items, location, etc. A POG typically consists of three parts: the first part is a color rendering showing the product with a location number. pictureThe second part is a layout showing the product location using black and white position numbers picture ; The third part is the product specification table, which lists each product's store ID number, UPC, weight, size, name, brand, packaging, etc.

[0130] Reference again Figure 2 , the following paragraphs summarize the operations performed by process 202.

[0131] Process 202 includes or employs the following features: (a) Data collection network; (b) data operating system; (c) Plug and play system; (d) Under the control of MCU 135A, binary decoder 130A-1 performs photoresistor polling and retrieval; (e) Threshold calculation through initial value and machine learning (ML); (t) Photoresistor mapping to match the product; and (g) RT-POG, which displays item percentage, condition, and time reports for items 105 (eg, items) on a single track (eg, track 115A).

[0132] As described above, there are three process blocks in the data collection network running under the control server 150. These are: process block 210: main control board gating and track polling; process block 240: mapping photoresistors to match products, calculating percentages, determining the conditions of the products, and uploading the processed data to the cloud database 178.

[0133] The process 202 begins at operation 207. The control server 150 scans the photoresistor of each track on the shelf.

[0134] Operation 210A. Control Server 150 - a main control board that controls each shelf via an Ethernet address.

[0135] Operation 210B: Control server 150 polls the track of each serial address of the shelf master board via the plug and play system.

[0136] Operation 230A. Control server 150 - retrieves the ML generation threshold for each photoresistor.

[0137] Operation 230B: Retrieve the initial value (uncovered digital voltage value) of each photoresistor and calibrate the threshold.

[0138] Operation 230C: Convert the digital signal into a cover / uncover value (discrete) in the control server.

[0139] Operation 240 A. Map the photoresistor to match the product using mapping software.

[0140] Operation 240B: Calculate the percentage and determine the condition of the item.

[0141] Operation 240C: The control server 150 uploads the processed data to the cloud database 178 .

[0142] End process 240.

[0143] Process 202 populates the cloud database 178 with formatted data for the following use cases: (a) Reporting module used by the chain's headquarters; (b) Based on the layout used by store managers and / or stockists picture (RT POG) real-time inventory monitoring; (c) Real-time inventory monitoring across store brands based on brand manufacturer usage; (d) the logistics systems used in warehouse operations; and (e) Robotic algorithms for product replenishment.

[0144] Each photoresistor, track, shelf, or component (eg, MCU 135A, binary decoder 130A-1, and track's circuit board 117A) is assigned a unique global identification (ID) number.

[0145] Figure 3 A table, Form 305, and a global identification number indicating the global retail system process picture ,The system consists of multiple layers: chains, countries, shelves, tracks and photoresistors.

[0146] Some communications in system 100 (eg, plug and play system 121) are conducted according to RS-485, which is a standard that defines the electrical characteristics of drivers and receivers for serial communication systems.

[0147] Features of this global retail system include providing each element from the chain store to each photoresistor with a unique cloud-based global ID in real time at multiple levels; and managed by cloud-based communication between this global system and each unique element.

[0148] Figure 4 A unique Base-62 global identification number for a component of the real-time inventory system is shown, which will be referred to in this document as a "unique global ID." The unique global ID is a 10-character string representing a base-62 number that uniquely identifies each component in cloud database 178. The unique global ID allows for quick identification of the component and allows for easy retrieval from cloud database 178.

[0149] The basic logical components in the cloud database 178 are stores, display units, shelves, facings, and photoresistors. Figure 5 , side #22) is a row of merchandise 105 arranged on track 115A so that customers can Figure 10 The first product 105 is seen at the front 1001, while all other products 105 in the face are arranged behind it and extend to the end of a single shelf panel 110. Figure 10 All of these components have unique global IDs. Furthermore, network-related components (e.g., routers, main control boards, rails, and other equipment) are assigned unique global IDs. In addition to these items, other entities and components can also have unique global IDs. For example, cameras, robots, replenishment pallets, warehouses, regions, suppliers, customers, and so on can all have global IDs.

[0150] The unique global ID has 10 characters. The unique global ID uses base 62, which uses all numeric characters 0-9, all 26 lowercase letters, and all 26 uppercase letters of the English alphabet. The current format of the unique global ID and the use of base 62 allow for 3,844 different components, each with 3.5 trillion possible unique global IDs, as well as a checksum digit, all within the IO Digital ID.

[0151] System network topology

[0152] Figure 5 Is the isometric view of a photoresistor array (e.g. photoresistor 125A) picture , the photoresistor array is located on a track (eg, track 115 ) and on a shelf (eg, a single shelf layer 110 ).

[0153] The leftmost display surface (or track) is designated as display surface #1, and the rightmost display surface (or track) is designated as display surface #22. All display surfaces are installed in a row from display surface #1 to display surface #22. The maximum number of display surface addresses and lighting device addresses that can be selected along the width of the shelf is 255. Lighting devices include but are not limited to Figure 15 The LEDs on LED light board 1530 are shown in Figure 2. 256 is the maximum number of LEDs in the current standard chip resolution. Of the 256, one number is assigned to FF. The remaining number is a maximum of 255. The maximum number of photoresistor arrays mounted on a PCB (e.g., track board 117A) is 768.

[0154] Figure 5The physical structure of the track's circuit board 117A is shown. The photoresistor 125A is placed on the top surface of the track's circuit board 117A, while the binary decoder 130A-1 and the MCU 135A are located on the bottom surface of the track's circuit board 117A.

[0155] also, Figure 5 Figure 5 shows the physical structure of the circuit board 117A of the track. Figure 5 shows the process of describing the binary decoder 130A as a polling access to communicate with all the photoresistors 125A one by one in a cycle. picture The MCU 135A controls the communication with the binary decoder 130A to perform the polling process.

[0156] Figure 6 It is a framework for the system network topology picture , where the face IDs and photoresistor array IDs along the width of the shelf form an ID system in the two-dimensional xy plane, which further forms a system network. This system network works in conjunction with the RT-POG and provides each element with an independent communication ID address as well as an actual physical address. By using mapping technology, the sensor location of any product on the shelf can be found.

[0157] The resolution of items on the shelf depends on the density of the photoresistors 125A along the width and depth, as well as the software capabilities.

[0158] Each photoresistor on track 115A, such as photoresistor 125A-1, is connected to the main control board 120 of a single shelf layer 110 via RS-485, and then the main control board 120 is connected to the control server 150 via Ethernet, and the control server 150 is connected to the cloud database API 177 via HTTP. The operations from the cloud database 178 to the photoresistor, such as photoresistor 125A-1, are described here.

[0159] First, the cloud database API 177 connects to the store Internet gateway 605 (router) via HTTP over the Internet (see Figure 6 ). The router is a device that communicates between the Internet and devices in the system network (such as the main control board 120).

[0160] Second, refer again Figure 6The store's Internet gateway 605 (router) and Ethernet gateway 610 (router) are connected via Ethernet (e.g., Ethernet 615). The number of Ethernet gateways (routers) can range from 1 to 253. These Ethernet gateways (routers) are isolated from each other by different IP ranges and network masks to reduce network traffic and the risk of broadcast storms, although virtual local area networks (VLANs) are another option. Considering the risks involved, we chose to use multiple routers to construct subnets or network masks for the system network in the store.

[0161] Third, the IP addresses and network masks of each control server (e.g., control server 150) and multiple rack control boards (e.g., control board 120) are set within the same IP range. The number of rack control boards per control server depends on the processing speed of that control server. As hardware and software advance, the number of rack control boards that can be managed will increase, but in practice, this may ultimately be limited. For example, the total capacity of Ethernet IP addresses is 65,535. With a subnet mask or network mask, one route can be used for 253 racks.

[0162] Fourth, shelves do not have IP addresses. Shelves are numbered by store, i.e., shelf IDs recorded in the cloud database 178.

[0163] Fifth, each control server, such as control server 150, communicates directly with cloud database 178. Each control server, such as control server 150, is assigned an ID in cloud database 178. Typically, communication between controller program 153 of control server 150 of store 101 and cloud database API 177 is via an HTTP link.

[0164] Sixth, the shelf main control board, such as the main control board 120, communicates with each display surface or each LED light board 1530 (see Figure 15 ) and its devices, acting as an RS-485 server, with each display surface or LED light panel acting as its RS-485 client. Each display surface or LED light panel uses a hexadecimal RS-485 address as its display surface or lighting ID, and the total number of RS-485 clients is 254 (01-FE).

[0165] Seventh, an MCU (e.g., MCU 135A) on each circuit board (e.g., track circuit board 117A) controls a multi-level binary decoder (e.g., binary decoder 130A) to sequentially switch the circuit's high and low levels, thereby switching each photoresistor on and off. The maximum number of photoresistors 125A selectable on track circuit board 117A (and therefore track 115A) is 768. Each photoresistor is assigned an ID starting with 0x0000.

[0166] Figure 7 is a block of a binary decoder (e.g., binary decoder 130A-1). picture , shows the working process of the binary decoder. Figure 7 In the MCU 135A, there are 12 first binary decoders, namely first binary decoders #1, #2, #3, and #12. Each first binary decoder interfaces with eight second binary decoders, and each second binary decoder interfaces with eight photoresistors, such as eight photoresistors 125A. When GPIO port 184 polls each binary decoder, each photoresistor under the polled binary decoder is selected and connected to ADC port 186. All analog signals generated by photoresistors 125A are converted into digital signals by ADC port 186. ADC port 186 converts the analog electrical signals into data processing signals controlled by MCU 135A. ADC port 186 and GPIO port 184 are both part of MCU 135A. The analog or digital signals generated by the photoresistors are all voltage-based.

[0167] Let's review Figure 3 The global ID system in FIG. 1 shows the lowest level component, a photoresistor 125A, placed on the circuit board 117A of a given track (e.g., track 115A) that senses changes in the intensity of the light received. When the track (e.g., track 115A) is inserted into the front and rear track extrusions, the track's four pins are connected to four conductive (e.g., copper) wires, and the data is immediately conducted to the main control board (e.g., main control board 120) (see FIG. 1 ). Figure 10 ).

[0168] In addition, when there is no product on the track, the initial value of the photoresistor 125A is recorded (see Figure 15 On the left picture The control server 150 sends the order and transmits the data to the track with the given ID via TCP / IP and RS-485 communication protocols.

[0169] The track, such as track 115A, records the current resistance (or voltage) value of each photoresistor, such as each photoresistor 125A, in turn through the ADC port 186 and converts it into a corresponding value within a range from 0 to VCC scaled by 4095 (i.e., 0 to 4095) as a series of voltage measurements 188. The data is then sent back to the control server 150 as a digital signal of the photoresistor on the track. The track API process is now complete.

[0170] The data conversion process of the photoresistor 125A is described here. The gap between the two rollers (for example, adjacent rollers of a portion of roller 111A) provides sufficient illumination for the photoresistor 125A under the track 115A, and each photoresistor 125A generates a corresponding linear resistance change according to the received light flux. The photoresistor 125A is connected in series with the 3.3V / 1A DC circuit VCC 180, thereby causing voltage / current changes with the light flux. When the received light flux is the largest, the resistance value becomes the lowest and the voltage becomes the highest. Conversely, when the photoresistor (for example, photoresistor 125A-1) is blocked by a product (for example, product 105), the resistance value becomes the highest and the DC voltage drops accordingly. In short, a DC signal (i.e., an analog signal) is generated in response to the amplitude change of the incident light flux.

[0171] Figure 7 The diagram shows MCU 135A controlling binary decoder 130A to cyclically switch the address of photoresistor 125A (e.g., switching to photoresistor #0x0765 while simultaneously measuring the voltage difference between the DC current of photoresistor #0x0765 and a standard DC voltage of a low dropout (LDO), such as DC+(3.3V IA)LDO, VCC 182). Based on the acquisition accuracy of ADC port 186, such as 12-bit resolution, MCU 135A divides the maximum voltage intensity of 3.3V by 4095, i.e., a range of 1 to 4095, and records the relative voltage difference, i.e., photoresistor #0x0765, during the polling process. During the polling process, a series of voltage measurements 188 of all photoresistors 125A are recorded (and compared with the MCU's DC+(3.3V IA)LDO, VCC 182).

[0172] The photoresistor 125A is composed of multiple binary decoder 130A stages controlled by an MCU 135A. The number of stages is used to adapt to a certain number of photoresistors 125A. The second-stage binary decoder (for example, #1 to #8 of the second stage) is linked to one of the first-stage binary decoders (level 1). If necessary, a third-stage binary decoder can be constructed as level 3 to link to level 2. Similarly, level 4, level 5, and more levels can be constructed. Figure 7 As shown, each binary decoder at level 1 is linked to eight binary decoders at level 2, and each binary decoder at the lowest level is linked to eight photoresistors. This structure can be extended to more photoresistors by using multiple levels of linking.

[0173] Figure 8 It is a frame for electronic equipment on a single shelf layer 110 picture , including two functional blocks, namely the main control board and the track.

[0174] Main control board

[0175] The main control board 120 represents a single shelf layer board 110, because each shelf has only one main control board. The power adapter 835 provides power to the main control board 120. The main control board 120 maintains Ethernet communication with the control server 150.

[0176] An access point (AP) is a network hardware device that allows wireless devices to connect to a wired network. The AP connects directly to a local area network (LAN), such as an Ethernet network, and then uses wireless LAN technology to provide wireless connectivity for other devices.

[0177] Power over Ethernet (PoE) is a technology that provides power to devices via Ethernet data cables. The main control board 120 provides power and communication links for the track 115, the PoE camera 830, and the PoE AP 825.

[0178] Network port 805, such as an RJ45 port (TCP / IP), is an 8-pin / 8-position plug or jack typically used to connect a computer to an Ethernet-based local area network. Network port 805 can be referred to as a Transmission Control Protocol / Internet Protocol (TCP / IP) port or an Ethernet port. The network switch component 810 built into the main control board 120 extends the Ethernet network to multiple RJ45 ports. PoE cameras 830 and PoE APs 825 use standard RJ45 ports.

[0179] The serial-to-Ethernet component 815 built into the main control board 120 acts as a TCP server, and one port of the network switch component 810 forwards to the RS-485 bus and acts as a serial port server, providing an Ethernet link for the track 115 and lighting through the conductive track.

[0180] The power management component 820 built into the main control board 120 serves as the power management for the entire shelf layer board 110, and performs multi-layer conversion on the DC power transmitted by the power adapter 835. Figure 8 middle: (a) Line A4 provides PSE (power sourcing equipment) power control for RJ45 port 805 according to the 802.3af / at protocol; Lines A1, A2, and A3 connect to 5G / 6G extender 832, PoE camera 830, and PoE access point 825, respectively; (b) Line B1 connects the power management component 845 of the track 115A to the power management component 820 of the main control board 120 through a conductive track; Line B2 provides DC power to the light emitting diode (LED) light bar / lighting board 855 at the bottom of the rack through a cable; and (c) Line C1 connects the serial port to Ethernet conversion component 815 of the main control board 120 and the serial port management component 840 of the track 115A; Line C2 provides RS-485 communication for the light-emitting diode (LED) light strip / light board 855 at the bottom of the shelf through a cable.

[0181] track

[0182] The built-in serial port management component 840 of the track 115A acts as a serial client of the RS-485 bus, which is connected to the MCU 135A.

[0183] The built-in power management component 845 of the track 115A receives power from the main control board 120 through line B1 to power all electronic components on the track 115A.

[0184] As a core component, MCU 135A performs the following operations: A: records the display surface track ID of track 115A; B: Controlling the pulse width modulation (PWM) of the LED controller 850 via the digital-to-analog converter (DAC) port 860, which works with the built-in LED chip 852 as a merchandise inventory indicator to control the brightness of the track 115A via line D2; C: Controlling binary decoder 130A via GPIO port 184 to manage address decoding of 1-768 photoresistor 125A via line DI; and D: Polling is processed through line D1 through the following steps: (i) ADC port 186 receives and measures the voltage signal from one of the 1-768 photoresistors 125A as input 186 of the ADC port; (ii) the voltage signal is converted into a digital signal through the ADC, and (iii) the other voltage signal is polled.

[0185] LED chip 852 is used to sound an alarm when the inventory percentage (the inventory percentage of a given track) is equal to or below an alarm level (e.g., 20%).

[0186] Control server program

[0187] Figure 9 This is the control server program 153 frame picture , which transmits data signals from the display surface / track to the cloud database 178. Figure 9 An overview of the software concept of the control server 150 , its basic internal processes and the connection of the software to signals from installed and activated control boards (eg, the control board 120 on site) and the online cloud database 178 is shown.

[0188] Control server 150 is a physical computer or other device whose function is to poll the status of all or a subset of all active control boards (e.g., control board 120) in a single store (e.g., store 101), detect changes in the status of any track on the control board (e.g., changes in the state of the photoresistor), process and convert these changes into meaningful shelf inventory data, and write these changes to cloud database 178. A store can have one or more control servers, such as control server 150, depending on the number of active control boards in the store. Each store has a unique ID in the entire cloud database 178, and each control server has an active server ID (asID) unique to that store.

[0189] When the control server program 153 is activated, the active facing matrix generator module 902 is started. The active facing matrix generator module 902 periodically retrieves a dataset of shelves, facings, and products within those facings from the cloud database 178, which have active dashboards and facing / photoresistor chains registered in the cloud database 178. A photoresistor chain is defined as a collection of photoresistors in a track. The scope of the dataset is limited to shelves and facings linked to the store and the asID (active server ID) specified in the configuration file 901. The active facing matrix generator module 902 is responsible for updating the internal active shelf / facing list 906 used by the dashboard status polling module 904.

[0190] The control board status polling module 904 continuously polls the collective shelf / display surface 903 of the shelf / control board-display surface / photoresistor chain specified in the internal active shelf / display surface list. A shelf / control board 110 / 120 refers to a pairing of a single shelf layer board 110 and a control board 120. All active single shelf layers 110 have a single control board 120 associated with them, and all control boards 120 process a single shelf layer board 110. This pairing is called "single shelf layer board / control board 110 / 120". All active single shelf layers / control boards 110 / 120 are polled asynchronously and concurrently, and each display surface / photoresistor chain is polled either synchronously or asynchronously and concurrently, depending on the architecture of the control board 120 and its firmware. Concurrency is the accurate term to describe the process because it does not refer to the computer execution, but to the process of polling the control board / display surface. There is enough "wait time" between requesting a signal and receiving it to "accommodate" the activity of polling another component. If this were done sequentially, it would be incredibly slow, and would get significantly slower as more boards were added to the network (i.e., it wouldn't scale well).

[0191] The board status polling module 904 passes the raw photoresistor chain data to the data mapping and conversion module 905. The data mapping and conversion module 905 formats the raw data according to the photoresistor chain mapping procedure and may add metadata or otherwise manipulate the raw data's positioning to conform to the structure of the cloud database 178. The board status polling module 904 then checks the latest photoresistor value list 906, which will be empty at module startup, to see if the current data differs from the latest photoresistor value for the currently processed active display. The board status polling module 904 passes the data to the data mapping and conversion module 905, which processes the data and writes the changed value to the cloud database 178. The board status polling module 904 then updates the latest photoresistor value list 970, which the board status polling module 904 will read at the beginning of the next board status polling cycle.

[0192] Plug and Play

[0193] Figure 10 This is an exploded top view of track 115A picture , which has a track circuit board 117A, showing that the photoresistors 125A of the track circuit board 117A are distributed and connected along the length of the track 115A. picture The features of a single shelf layer 110 are introduced, including 4 copper wire channels 1005, rear extrusion 1007, track connection holes 1008, rear baffle 1010, vertical support rods 1015, end circuit board 1020, end cover 1025; porous plate 1030, 4 power and data link conductors 1035, front extrusion 1040, landing area 1045, drainage holes 1050, shelf surface 1055 and roller track base 1060.

[0194] The photoresistor 125A is located on the track's circuit board 117A, which is located below the roller 111A so that the photoresistor 125A does not come into contact with any merchandise placed on the roller 111A.

[0195] Figure 11 This is an exploded side view of the track 115A with the main control board 120 picture . Figure 11 Features described include locking flange 1102, LAN cable plug 1105, LAN bus 1107, locking lever 1110, rear extrusion base 1115, locking pin slot 1117, 4 power and data plug pins 1120 and roller support rod 1130.

[0196] The rear extrusion 1007 includes a rear extrusion base 1115, a rear baffle 1010, four copper wire channels 1005, an end circuit board 1020, an end cap 1025, a locking lever 1110, and a locking flange 1102. The end circuit board 1020, along with four power and data link conductors 1035, connects to the main control board 120 as an integral unit. The end cap 1025 is a container for the integral unit, which is intended to be placed on a shelf surface 1055.

[0197] refer to Figure 10 and Figure 11 Photoresistor 125A generates a signal based on the presence of merchandise on roller 111A. The magnitude of the signal changes as the merchandise on roller 111A moves from its original position. All signals (data) from track circuit board 117A communicate with main control board 120 via four power and data plug pins 1120, four power and data link conductors 1035, and terminal circuit board 1020.

[0198] The 4 copper wire channels 1005 are configured to receive the 4 power and data link conductors 1035 for close insertion, for the 4 power and data plug pins 1120, i.e., one pair for power and one pair for data transmission, for a total of 4 pins. In practice, the number of pin pairs can be 6, 7, 8 or more, odd or even. The circuit board 117A of the track is connected to the 4 power and data link conductors 1035. In addition, Figure 11 The main control board 120 is shown fixed to the rear panel 1010 and is positioned between the rear panel 1010 and the shelf nail plate (i.e., the multi-hole plate 1030). The main control board 120 has four PoE LAN cable plugs 1105 for data communication and power supply.

[0199] Figure 12 Some features of the plug and play system 121 picture The plug and play system 121 connects the four power and data plug pins 1120 from the roller 111A and the photoresistor ( Figure 12 ) are electronically connected to four horizontally placed power and data link conductors 1035. In this regard, Figure 12 The four power and data plug pins 1120 are shown, including a pair of DC power pins 1215, namely a DC- plug and a DC+ plug, and a pair of data pins 1210, namely an RS-485A plug and an RS-485B plug. Thus, within the four copper wire channels 1005, there are four slots for each of the four plugs. Furthermore, the RJ45 cable 1205 for the PoE camera 830 is connected to the LAN cable plug 1105. The terminal circuit board 1020 is designed to connect to the main control board 120.

[0200] Figure 13 This is the bottom view of track 115A picture , with electronic components, adjacent to the front bezel 1305. Two pairs of locking flanges 1318 lock onto the rear extrusion 1007 ( Figure 10 and secure the rail 115A to the front extrusion 1040 ( Figure 10 Status LED 1310, MCU 135A and binary decoder 130A-1 are located on the bottom of the track's circuit board 117A and above the bottom of the track 115A ( Figure 13 (not shown), there are two sets of power and data plug pins 1120 at the bottom of the track 115A.

[0201] Figure 14 It is the framework of the rack network cabling system picture , which specifies the main control board 120, track 115, LED light bar / light board 855, PoE camera 830 and PoE AP 825. PoE AP 825 is used to locate and provide WLAN (wireless local area network) links for other electronic devices (such as mobile phones, etc.).

[0202] Figure 14 Shows rack network communications, including cabling and protocols.

[0203] There are three types of racking networks: Type A, Type B, and Type C. Type A is used for power and data communication from the track to the main control board. Figure 14 As shown on the left, the power bus and RS-485 bus include a DC+ power supply, a DC- power supply, an RS-485A, and an RS-485B. All data is transmitted from the main control board 120 to the control server 150, and then transmitted to the cloud database 178 through the network switch component 810 and the RJ45 port 805 (TCP / IP), as shown in FIG. Figure 8 shown. Type B is used for power supply and lighting data communication on the shelf, such as Figure 14 As shown in the middle, a DC+ power supply, a DC- power supply, an RS-485A, and an RS-485B are included through the power bus and RS-485 bus, respectively. Similarly, all data are transmitted from the main control board 120 to the control server 150, and then transmitted to the cloud database 178 through the network switch component 810 and the RJ45 port 805 (TCP / IP), as shown in FIG. Figure 8 shown. Type C is suitable for electrical / electronic devices, such as PoE Camera 830 and PoE AP 825, such as Figure 14As shown on the right, all data is connected to the RJ45 port 805 via the Ethernet bus and then transmitted to the cloud database 178 via the control server 150.

[0204] Figure 14 The detailed structure of the RS-485 bus 1420 and the power bus 1450 is shown in Figure 12 As a feature of the plug-and-play system.

[0205] Real-time shelves picture (RT-POG) is the result of a centralized and cloud database, real-time, shelf merchandise inventory monitoring system in stores.

[0206] Access points (APs), 5G / 6G extenders, and cameras

[0207] Figure 15 is a bottom view of a shelf 1501 with electronic equipment picture Shelf 1501 includes cameras 1550 and 1505 , AP 1510 , main control board 1512 , 5G extender 1515 , main control board 1520 , shelf base 1525 , LED light board 1530 , shelf 1535 , merchandise 1540 , and shelf 1545 .

[0208] AP 1510 is located at the bottom of shelf 1545. It is connected to the main control board 1512 via an Ethernet cable. The communication protocol is the PoE standard of 802.3af / at. The MAC (media access control) of AP 1510 is recorded in the control server 150 and associated with the shelf ID so that when a wireless device (desktop, mobile device, wearable device, or other dedicated device) connects to the Ethernet via IP, it can record that the wireless device is near shelf 1501. However, a feasible distance is required to calculate the RSSI (received signal strength indicator) of the wireless communication network signal, such as based on IEEE 802.11. APs (such as AP 1510) do not need to be installed on every shelf, but only on a subset of the shelves.

[0209] Digital cameras 1550 and 1505 with normal or wide-angle lenses are placed on the bottom or surface of shelf 1545, on its rear or front stop edge. The number of cameras per shelf can be zero, one, or more, or one camera can operate vertically across several shelves in a row. The camera is used to monitor the front-most product on the shelf, either on the same side or on the other side of the aisle intersection between two rows of shelves. The digital camera visualizes and records the front-most product as picture Like, these pictureImages can be recognized and classified with 99% accuracy through ML (machine learning) AI, such as CNN (convolutional neural network), where the ML process consists of three steps: convolution, max polling, and full connection. picture A digital camera with image and data processing can be integrated into one unit or divided into two parts, one of which is smaller and includes the lens and picture Like the sensor, the other part consists of a sensor and a picture It can be embedded near the front block, while the other part includes picture Devices like converters, Ethernet converters, DC-DC converters and others can be placed in another location, such as the bottom shelf. picture The image needs to have enough pixels, for example, 2M to 5M, so that the result of the above ML processing can achieve a high accuracy. Figure 8 In the embodiment, there is a PoE camera 830, which processes the data and transmits the data to the control server 150 via a network port 805 (eg, an RJ45 port (TCP / IP)).

[0210] When the 5G extender 1515 is connected to the RJ45 port of the main control board 1520, the 5G extender 1515 uses 802.3af / at power and only serves as a PD (power device) to extend the outdoor 5G signal into the store (e.g., store 101).

[0211] Figure 16 This is a diagram of the track navigation of robot 1645 using ultra-high frequency radio frequency identification (UHF RFID) technology to transport product 1605 to the track. picture . Each PCB has a UHF RFID tag chip and antenna, such as the passive tag antenna 1620 on PCB 1615, which operates in the 902-928 MHz frequency band. UHF RFID can use other frequency bands, but must comply with local government regulations, such as 865-868 MHz in Europe. The UHF RFID tag of each track (such as the passive tag antenna 1620) is recorded with an independent ID in the central cloud database 178 and is associated with the position of the track, that is, facing the robot arm 1645 to replenish the goods according to the order of the cloud database API 177. A robot that can pick and place with sufficient accuracy is required.

[0212] When robot 1645 arrives near a cargo hold that needs replenishment, it turns on UHF RFID antenna 1610 embedded in robot arm 1625 and performs an active group scan of nearby UHF RFID tags (e.g., nearby passive tag antennas, such as passive tag antenna 1620). It then finds the RFID tag ID that needs replenishment and those on adjacent tracks through communication with control server 150. Three-point positioning calculations are performed based on the RSSIs obtained by scanning these three or more RFID tags, allowing robot 1645 to accurately find the track and replenish the required items on the track.

[0213] The data provided to robot 1645 includes: (A) Accessing the navigation map through the Robot Navigation API of the Application API 179 picture , so that the robot 1645 can replenish the goods; (B) access the real-time inventory of the goods through the RT-POG APT of the application API 179; (C) shelf location signals generated by APs / 5G extenders; and (D) RFID passively responds to track positioning signals from the robot's search signal associated with the track.

[0214] Navigation and software can be based on shelf ID, rack ID and track ID.

[0215] Figure 17 It is the real-time shelf inventory represented in RT-POG picture Show.

[0216] Figure 18 It is the RT-POG demo.

[0217] When presenting real-time shelf inventory to retail store managers, the data must be organized in a readable and easily relatable format. Product type and location are crucial data for store associates involved in restocking products and maintaining shelf and display conditions. This information must be presented in a format that is familiar and immediately understandable to those who need it.

[0218] By displaying real-time shelf inventory within the framework of a standard POG, the information is presented in a format familiar to retail professionals, enabling store personnel involved in replenishing and maintaining shelf and display conditions to easily replenish inventory as needed and correct shelf and display conditions such as misplaced items, gaps between items, and missing items from the front of displays.

[0219] Real-time shelf inventory is displayed as a percentage, and gaps between product items and missing product items on the front of the display are displayed as circles whose size and position correspond to the real-time shelf and display conditions.

[0220] Additionally, shelf stock percentages are color-coded so that low stock levels can be easily seen.

[0221] Figure 17 Shows a cross-sectional view of a shelf with 70% inventory (by shelf capacity / product) picture , view 1710, a gap state with a large space between two product items, view 1720, and a state where the product items are lying flat on the shelf rather than in an upright position, view 1730. In each case, the percentage in stock is displayed, although the percentage may be inaccurate or misleading in the conditions represented by views 1720 and 1730.

[0222] Figure 18 The simplified user interface of the RT-POG 1800 is shown. Units, shelves, and display surfaces are displayed, along with photos and text indicating the products assigned to the shelves and display surfaces by the store POG. Inventory is color-coded, with cool colors indicating sufficient stock and warm colors indicating a need for restocking.

[0223] Inventory levels are defined as the percentage of inventory in a given category / section within a given store that is not empty or has not received a warning. For example, if the single-serve beverage section has 100 shelves, each 48 inches wide and 22 inches deep, with 100 rails per shelf, and there are 10,000 rails in the single-serve beverage section, then 20% means that 80% of the shelves are empty or have inventory levels below a warning level, such as 20%. For consistency, use the same colors defined in the RT-POG: red for 0% to 20%, orange for 21% to 40%, yellow for 41% to 60%, green for 61% to 80%, and blue for 81% to 100%.

[0224] A given point (circle) represents a store, and selecting that point (eg, a point on a touch screen display) will display the store number.

[0225] You can select a category or section for a given store. Here, a section is defined as a subcategory under a category. For example, under the Health & Wellness Products category, there are subcategories for Allergies & Sinus, Cough & Cold, Diabetes OTC, Eye Care, Oral Care, Sleep & Snoring, and Vitamins & Supplements.

[0226] By using RT-POG, information is delivered to the headquarters and relevant departments selected by the headquarters.

[0227] Some of the data collected are listed below, for example: (1) Real-time inventory based on RT-POG; (2) Display the status of goods in real time; (3) Real-time sales and revenue; (4) cash flow; and (5) Sales history.

[0228] Figure 19 This report is a 1900 report on shelf inventory over time. It displays inventory status at regular intervals within a specified timeframe. Data can be selected, grouped, and sorted based on user-specified parameters. For example, you can group and sort the report by products across all display areas in all stores, or by selected shelves within a single store.

[0229] exist Figure 19 , select the report for all display surfaces in the store, ungroup them, and sort by fixture, shelf, and display surface. The time range starts at 1:00 PM on May 17, 2022, and ends at 2:00 PM on the same day. Report inventory for all 15-minute intervals between the start and end time ranges. Figure 19 This is a report for a single business day at a store in 1900, grouped and sorted by product, with a line break showing stock changes over time. picture by picture Represented in a shaped manner.

[0230] exist Figure 20 In the report, the daily shelf inventory changes over time by product are reported, with a broken line picture Format 2000 shows several exemplary time-based inventory on-shelf curves.

[0231] Figure 21 It is a real-time inventory map of the country 2100, based on the continental United States. picture The real-time shelf inventory level is displayed in the form of RT-POG and real-time shelf inventory location. picture Teleport to the chain's headquarters. Inventory Location picture is geographic (national or regional), displays all stores (including chains or branded items) by store ID, whether owned by the user or selling for the brand, and displays real-time shelf inventory of the item by color percentage. land picture Some components of the structure include: (1) Geographic hierarchy: country, region, state, county, city, and / or town; (2) Users: chain stores, brand manufacturers. (3) Supply chain: suppliers, vendors, warehouses, delivery services, etc.; (4) Product: store, category, section, brand name, packaging; (5) Display: racking system, shelf system, walk-in cooler, etc. (6)POG:RT-POG; (7) Inventory levels on shelves: indicated by color; and (8) ID number: Each product has a unique ID number.

[0232] The characteristics of the Quantum system have been described above.

[0233] From the above description based on the photoresistor as a sensor, we can see that through vision picture The complete structure of the cloud-based Quantum system explained, as well as its powerful function of monitoring the real-time inventory of goods on the shelf system.

[0234] In the Quantum system, analog / digital signals and real-time inventory data are connected through β pa , β a d, βde, β ec or β ccl Transmitted from lower level to higher level. Digital signals are sent to local control server and signal processing program, where they are converted into real-time inventory data and transmitted to cloud layer / shell α c A cloud database where requests for real-time inventory data from client applications and processes can be received by the cloud database through a public API.

[0235] The techniques described herein are exemplary and should not be construed as implying any particular limitations on the disclosure herein. It should be understood that various alternatives, combinations, and modifications may be devised by those skilled in the art. For example, the steps associated with the processes described herein may be performed in any order unless otherwise specified or provided for in the steps themselves. The disclosure herein is intended to encompass all such alternatives, modifications, and variations that fall within the scope of the appended claims.

[0236] The terms "comprise" or "comprising" should be interpreted as specifying the presence of stated features, integers, steps or components, but do not exclude the presence of one or more other features, integers, steps or components or groups thereof. The terms "a" and "an" are indefinite articles and therefore do not exclude embodiments having a plural article.

Claims

1. A centralized, cloud-based system for real-time monitoring of shelf merchandise inventory and condition, comprising: Shelf components, including: (a) Shelves, including: a track on which the merchandise is movably placed; A main control board with a network port; and a plug-and-play assembly including data and power conductors to enable information communication between the main control board and the track; (b) a plurality of photoresistors positioned proximate to the track, wherein the photoresistors generate analog signals indicative of the number of items placed on the track at any given time; (c) a unique identifier associated with at least one component selected from the group consisting of one of the photoresistors, the group of photoresistors, the track, the main control board, the plug-and-play component, and the shelf; (d) a binary decoder disposed on the track for polling the analog signal from each of the photoresistors; and (e) a microcontroller unit (MCU) located on the track, the MCU including a track application program interface (API), an analog-to-digital converter port for inputting analog signals, and an analog-to-digital converter, wherein the main control board is located on the shelf and is used for Ethernet networking of the shelf and connecting to the rails via the network bus and the power bus through the plug-and-play system; and Connected to the control server of the shelf assembly through the network port of the main control board, The control server sends a request to scan each photoresistor to the MCU through the main control board, and then the control server sends a request to the MCU through the main control board to retrieve an analog signal from the photoresistor. wherein the MCU then converts the analog signal into a digital signal through the analog signal to digital signal converter, so that the digital signal is transmitted to the main control board through the plug-and-play component and then transmitted to the control server through the network port; wherein the control server includes a system for writing the processed shelf inventory data into a cloud database; and The cloud database stores and organizes data from the control server and allows client applications and processes, through an API server or multiple API servers, to access the data by sending requests to and receiving data from the API server associated with the cloud database.

2. The system of claim 1 , wherein the control server then performs the following operations: Retrieving the generated threshold value of each photoresistor from the cloud database; retrieving the initial uncovered digitized voltage value of each photoresistor and calibrating its threshold value; Converting the digitized voltage value received from the main control board into a covered / uncovered value in the control server; mapping the covered / uncovered value of the photoresistor to match the physical size of the product placed on the track; Adjusting the anomaly mapping of the covered / uncovered values using machine learning-based artificial intelligence techniques; as well as The percentage of the goods on the track is calculated to form processed data; and the processed data is uploaded to the cloud database.

3. The system according to claim 2, further comprising: a cloud server accommodating the cloud database; The cloud server includes a cloud database API and an application API, which connects to the control server software via the Hypertext Transfer Protocol (HTTP) and connects to client applications and processes via HTTP.

4. The system of claim 3, wherein the processed data and various permutations of the processed data can be retrieved via the cloud database API by at least one client selected from the group consisting of a remote inventory monitoring application, a product inventory robot, a delivery logistics application, and a data mining and analysis system.

5. The system according to claim 1, wherein the track is configured as a single longitudinal track having a bottom, a left side, a right side, a front end, and a rear end, and The shelf further comprises a printed circuit board (PCB) arranged along the individual longitudinal tracks, on which the photoresistor, the binary decoder and the MCU are mounted, and is accessed through the track API in ambient light or non-ambient light to track the goods.

6. The system according to claim 1, wherein the shelf comprises a surface, a bottom, a left side, a right side, a front end, a rear end, a front stopper, a rear stopper, at least two rails, at least two dividers and a support member, The support member comprises: (i) at least one beam spanning the width of the base, (ii) a left support, (iii) a right support, and (iv) a vertical wall and a pair of uprights as a rack assembly or the like, wherein said supports comprise: (i) a wire mesh as a surface, or at least one beam across the width of said base, (ii) a left wire or beam at the left edge, (iii) a right wire or beam at the right edge and (iv) vertical uprights as supports for a racking system or the like, wherein the main control board includes electronic components for data communication, wherein the at least four embedded data and power conductors are located on the surface of the shelf for transmitting data bus between the at least two rails and the main control board via serial communication, The plug-and-play assembly is adapted for use in at least one configuration selected from a plurality of separate longitudinal tracks of photoresistors and a plurality of non-separate longitudinal tracks with photoresistors.

7. The system according to claim 1, in, The analog signal generated by the photoresistor is converted into the digital signal through a retrieval process by the binary decoder under the control of the MCU for data processing. wherein the shelf accommodates a number of said photoresistors equal to N x M, where N is the number of said photoresistors along the depth of said track, and where M is the number of tracks along the width of said shelf, and Where N can be any integer greater than 0, where N greater than or equal to 768 is sufficient in most cases.

8. The system according to claim 1, wherein the photoresistors are organized by a binary decoder of multiple levels controlled by the MCU, and The number of the multiple levels is used to adapt the number of photoresistors.

9. The system according to claim 1, wherein the system utilizes a data communication network configured as a serial communication bus for the rails and an Ethernet network for the rack main control board; wherein the plug-and-play component is configured to communicate with the shelf main control board and the track on the shelf; wherein said point is defined as said rack main control board or said track as one of said track's serial communication buses; wherein the track is identified by methods such as scanning a unique ID encoded in a strip-like element such as a barcode affixed to the track; wherein the communication address of the track is predetermined and is independent of the ordinal position of the track on the shelf; and The communication between the shelf main control board and the track is defined as a non-point-to-point communication form.

10. The system according to claim 1, wherein the system utilizes a data communications network configured for a serial communications bus of the track and an Ethernet network of the rack main control board; wherein the plug-and-play component is configured to communicate with a main control board of the shelf and tracks on the shelf; Wherein the point is defined as the main control panel of the shelf or one of a plurality of fixed segment devices (e.g., outlets along the width of the shelf); wherein the track is identified by polling the fixed socket or segment device, and for a newly inserted track, a unique ID stored in an internal memory of the track is read by communicating with the fixed socket or segment device; wherein the communication address of the track is automatically set to the ordinal position of the fixed socket or segment device, so that the ordinal position of the track can be determined, as well as its approximate position relative to the leftmost edge or rightmost edge of the shelf; and The communication between the main control board of the shelf and the track through the fixed socket or segment device is defined as a point-to-point communication form.

11. The system of claim 9, wherein the plug-and-play component comprises: Plug and play main channel, including: (i) At least four built-in channels; (ii) at least four embedded data and power conductors; (iii) at least four pins on a connector of the rail of the shelf for transmitting information between the main control board and each photoresistor on the rail; (iv) a power bus; and (v) a data link based on the connection of N nodes in a multipoint network, such as an RS-485 bus, The plug-and-play channel includes: (i) at least four conductive wire channels; (ii) track plug pins corresponding to the number of conductive wires; (iii) a power bus; and (iv) a data link based on the connection of N nodes in a multi-point network, such as an RS-485 bus, and (v) a terminal PCB for connecting to a main control board of the track. The plug-and-play assembly electronically connects the four data and power plug pins from the roller and photoresistor to the four horizontally placed data and power conductors. The 4 data and power plug pins include a pair of DC power pins and a pair of data pins, and Among them, in the 4 copper wire channels, there are four slots or four pairs of slots to accept four plugs.

12. The system according to claim 1, The method for mapping the photoresistor comprises: Cloud, edge, or on-premises databases, including: Data relating to the specifications of the track, plan data and data relating to the physical dimensions of the merchandise; a data stream comprising a signal from said track; and an algorithm that maps photoresistors to product dimensions using the cloud, edge, or local database and its data, wherein the method of mapping photoresistors allows accurate shelf inventory percentage readings for various product items using a fixed and limited number of photoresistors on a track, and Among them, AI technologies such as supervised machine learning, unsupervised machine learning and semi-supervised machine learning are used to detect and adjust abnormal conditions of the product items on the shelves, such as gaps between product items and chaotic arrangement of the product items.

13. The system according to claim 1, The main control board includes: Network switch components; and power management components, The network switch component: Through the serial port to Ethernet conversion component of the MCU, it is connected to the serial port management component and the light emitting diode (LED) chip on the track, and Connecting a Power over Ethernet (PoE) access point, a PoE camera, a 5G extender, and a 6G extender to a power source equipment (PSE) component and at least one component selected from the group consisting of the following components through the network port, wherein the power management component on the main control board transmits power from the shelf to the rail, and transmits power to at least one of the LED light strip, lighting panel, power adapter, network port and serial port management component of the MCU through the PSE for data communication, and The system further includes a link between the shelf and the access server via the network port.

14. The system of claim 1 further comprising a wiring and protocol system for network communication by: A serial communication bus between the track and the main control board; an Ethernet bus between the main control board and the control server; as well as A component for communicating between the access server and the cloud server using the Hypertext Transfer Protocol (HTTP), wherein: The network communication includes: (1) Power and data communication from the track to the main control board; (2) power and lighting data on the shelf; and (3) Power supplies for electrical / electronic equipment.

15. The system of claim 1, wherein the system utilizes: Serial communication bus between the rails and the shelves; Ethernet between the shelves and the control server; Hypertext Transfer Protocol (HTTP) between the control server and the cloud server; and HTTP between the cloud server and the client application or process.

16. The system of claim 1, wherein the shelf further comprises: Electrical / electronic components for data collection and communication in an Ethernet environment; wherein the electrical / electronic component comprises at least one device selected from the group consisting of an access point (AP), a 5G / 6G extender, the main control board, a lighting panel, and a camera; and a serial communication bus for transmitting data collected from the track to the rack via the plug-and-play components, wherein the data is generated by the photoresistor, and The main control board includes the network port, which is used for communication between the rack and the control server via an Ethernet bus.

17. The system according to claim 1, The shelves include digital cameras placed at relevant locations to identify the products, and each track is matched with the POG by using AI technology including CNN (convolutional neural network). wherein at least one of the digital cameras monitors one or more rails of the rack at a fixed position, wherein at least one of said digital cameras is placed in a relevant location, either on the same shelf or on a shelf across the aisle, One of the cameras visualizes and photographs merchandise at at least one shelf intersection mounted across the aisle. One of the digital cameras can be placed near the bottom of the front fender to monitor the front-most item on each track. A movable digital camera replaces the fixed position camera to scan the width of the shelf, One of the cameras captures an image with a sufficient number of pixels, The image of the frontmost product on each track can be captured by a wide-angle lens and reconstructed. Wherein a single digital camera comprises two parts: a lens and an image sensor, and another part comprises an image converter, an Ethernet converter, a DC-DC converter and / or other devices; and A single digital camera is an integral unit comprising the two parts mentioned above.

18. The system according to claim 1, wherein the rack further comprises an access point (AP) for data communication via the network port, wherein said rack houses wireless devices for connecting to an Ethernet environment via said access point, wherein the control server records the media access control associated with the shelf ID, wherein the wireless device is connected to an Ethernet network via an Internet Protocol, and The control server records that the wireless device is in the vicinity of the shelf, and The AP is used for navigation of the wireless device.

19. The system according to claim 1, The shelf also includes a 5G / 6G extender for data communication via the network port. wherein the rack accommodates wireless devices to connect to an Ethernet environment via the 5G / 6G extender, wherein the 5G / 6G extender is connected to the network port and serves as a power supply device to extend the 5G / 6G signal from outside the store to inside the store, and The 5G / 6G extender is used for navigation of the wireless device.

20. The system according to claim 1, The shelf also includes a printed circuit board with an ultra-high frequency radio frequency identification (UHF RFID) tag chip and an antenna, wherein the UHF RFID tag chip has an independent identification recorded in the cloud database and is associated with the location of the track, The system further comprises: robot: receiving an instruction from the cloud database to deliver the product to the shelf; Turning on a UHF RFID antenna embedded in the robotic arm; Active group scanning of UHF RFID tags; The UHF RFID tag chip of the shelf is found based on received signal strength indicators (RSSIs) obtained from three or more of the UHF RFID tags, thereby locating the track; and the product is transported to the track.

21. The system of claim 20, wherein the robot utilizes at least one data item view selected from the group consisting of: (a) Navigation map; (b) real-time shelf inventory of goods; (c) Shelf location signals generated by access points (APs) and / or 5G / 6G extenders; (d) a track location signal generated by an RFID passively responding to a search signal of the robot associated with the track; (e) Shelf ID; (f) Shelf ID or main control board ID; and (g) Track ID.

22. The system according to claim 1, wherein the system uses unique global identification (ID) of components in the system, The unique global identifier is configured with two base-62 digits determined by the component type of the component, followed by seven base-62 digits randomly selected and tested for uniqueness, and a base-62 checksum to ensure data integrity during network transmission.

23. The system of claim 1, further comprising: a polling / scanning module for retrieving a digital signal from the photoresistor; a photoresistor mapping module, configured to map the photoresistor to the physical size of the commodity currently on the track; as well as Module 100 includes determining a threshold value for converting the digital signal into a discrete covered / uncovered value, wherein the threshold value is determined using at least one data set selected from the group consisting of: (a) predetermined thresholds for product types; (b) using machine learning derived from labeled training data to adjust AI thresholds; (c) adjusting a threshold value of a photoresistor based on the track not being covered at all times; (d) information from a calibration module that further adjusts the thresholds and adds the initial uncovered value of each of said photoresistors recorded when said track was installed; as well as (e) Information from a data writing module that applies mapping and thresholding to data and writes the data to the cloud database.

24. The system of claim 1, wherein the cloud database comprises at least one data item view, the data item view selected from the group consisting of: (a) real-time shelf inventory data expressed as a percentage or non-percentage value; (b) historical shelf inventory presented in tables, charts, or other formats; (c) real-time shelf inventory by store, region, department, country, or global geography; (d) real-time shelf merchandise inventory for individual stores by category, section, department, shelving system, or other relevant grouping; (e) commodity, product and brand data; (t) Client data; (g) Main control board, track, MCU, binary decoding, photoresistor or sensor metadata; (h) Plug and play, cameras, LED light strips / panels, access points, RFID, and other device metadata; as well as (i) Other supporting physical device metadata.

25. The system of claim 1 , wherein the system provides a set of predefined views to a user device that deliver data to the user device to facilitate visualization of real-time shelf inventory in at least one format selected from the group consisting of a layout format used in notifications, a historical format, and any number of user-defined or requested analytical reports.

26. The system of claim 25, wherein the system employs an application program interface for the cloud database that allows a user of a client application or process to access data from the cloud database and view the accessed data according to the predefined view.

27. The system of claim 1 , wherein the system employs an application program interface for the cloud database that provides a list of replenishment commands to robotic equipment, unmanned delivery systems, human operators, or any other process associated with replenishing shelf inventory, the commands being prioritized by geographic location, movement of specific items, or other factors.

28. The system according to claim 1, wherein the network port is a Transmission Control Protocol / Internet Protocol (TCP / IP) port, and The network port is an Ethernet port.

29. The system of claim 1, configured according to a general purpose Quantum system.

30. A centralized, cloud-based system for real-time commodity inventory monitoring, configured in accordance with the Quantum system, comprising multiple layers or shells α p , α a , α d , α e , α c and α cl , respectively represent perception, analog signal, digital signal, Ethernet processing, cloud database and client, wherein the layers or shelves are defined by their functions, The connection between two adjacent layers is as follows: (a) Connect β pa Located between the perception layer and the analog signal layer; (b) Connect β ad Located between the analog signal layer and the digital signal layer; (c) Connect β de Located between the digital signal layer and the Ethernet processing layer; (d) Connect β ec located between the Ethernet processing layer and the cloud layer; and (e) Connect β ccl Located between the cloud layer and the client layer, where the layer represents: (a) β for analog signal generation pa ; (b)β ad Used for conversion of analog signals to digital signals; (c)β de For signal processing in Ethernet (e.g., real-time inventory data generation); (d)β ec For data transferred from the Ethereum space to the cloud database; (e)β cc1 Used by the client to send data and service requests to the cloud database through the public API. The connections are represented from lower to higher levels, where nodes are defined as "e-units", including α p , α a and αd, The multiple layers act as a quantum structure for data generation and communication.

31. The system according to claim 30, in, The photoresistor on the track on which the goods are movably placed is used as a sensor to sense the layer α p The existence of light in Among them, the analog signal on the track is in layer α a is generated to indicate the number of items placed on the track at any moment, Among them, layer α p and layer α a The connection between the two is to generate a connection β pa , Among them, the analog signal on the track is in layer α d is converted into a digital signal, A plurality of binary decoders placed on the track poll the analog signal of each photoresistor, and a microcontroller unit (MCU) placed on the track has a track application program interface (API), an analog-to-digital converter port for analog signal input, and an analog signal to digital signal converter. Among them, layer α a and layer α d The connection between is a transition connection β ad , Among them, the main control board is placed on the shelf for layer α e In-the-rack Ethernet networking and rail connectivity via plug-and-play components (i.e. processing) via network bus and power bus connections β de , where the MCU has converted the analog signal into a digital signal through an analog signal to digital signal converter, thereby transmitting the digital signal to the main control board through plug-and-play components, The main control board set on the rack also communicates with the network port of the rack main control board at the Ethernet layer α e Connect to the control server, The main control board set on the shelf is also connected to the camera, access point (AP) and / or 5G / 6G extender through the network port to connect to the Ethernet layer α e electronic devices in The access server sends a request to the MCU through the main control board to scan each photoresistor on each track on the shelf, and the MCU obtains an analog signal from the photoresistor and addresses it through a binary decoder, wherein the main control board receives the analog signal through the Ethernet layer α e The network port in the control server sends a digital signal, The control server is a system that processes the signal and converts it into processed data, and then writes the processed data into layer α c In the cloud database, Client applications and processes (such as RT-POG) can connect to the cloud via HTTP. ccl and β ec accessing the processed data from the cloud database via a public API, and The client application and process send requests to the cloud database through the public API of the API layer, API Connect via Cloud β ec and β ccl Connection access Ethernet layer α e Server software in, i.e., signal processing software, in-store or client applications or processes (e.g., RT-POG).

32. The system according to claim 30, where α p , α a and α d are the components of the node, where α p is the perception layer, where α a is the analog signal layer, where α d is the digital layer, and The node is located in the commodity inventory monitoring system.

33. The system according to claim 30, wherein the system has the ability to sense the presence of light, The perception is passed through layer α a The connection β in pa Generate a certain amount of value to generate an analog signal, The analog signal passes through layer α d The connection β in ad Converted into digital signal, The nodes are defined as including the perception α p , simulation layer α a and conversion layer α d "e-unit", and The sensing and generation of the analog signal is characteristic of the quantum system and reveals the generation of data.

34. The system according to claim 30, The node has a digital signal, which is connected to the β de Transmitted to a specific device in the Ethernet space or environment, as a transmission, Wherein the racks and rack systems and other displays are located in the Ethernet space, wherein the signals are transmitted within the Ethernet space to the local access server, each The store is equipped with at least one local control server, The local control server provides digital signals from the shelves to the signal processing software, which converts the signals into application data, such as real-time inventory data. The application data, such as real-time inventory data, is connected to the internal API through the beta ec From the signal processing software to the cloud database, Data transmission and development are the characteristics of the quantum system, and there are growth stages of data, step by step, with driving force.

35. The system according to claim 30, where the system is one of multiple layers that work as a quantum structure from low to high, represented by signal and data and data quality, in, Among the multiple layers, the cloud database as data storage represents the highest level of all connections, and subordinate to this layer is the Ethernet with local servers for data creation, operation and communication. A node is a basic unit that generates and processes data from analog signals sensed and generated. There are six layers, five connections, and three sublayers in the client layer. The cloud database used as data storage is connected to the cloud through the β ec with Ethernet layer / shell α e communicates with servers in the , and satisfies requests from client applications and processes through a public API, where low-to-high processing is characteristic of the quantum system, And when data flows through an ordered process, there is a hierarchical data series.

36. The system according to claim 30, Two of the APIs, the internal API and the public API, can be represented by concentric circles in the cloud. The internal API communicates with the processing software at the Ethernet layer, while the public API communicates with the client layer. Requests from client applications and processes are sent to the cloud database through public APIs. The Ethernet layer processing software sends the database read and write requests to the cloud database through the internal API. The dual API in the cloud layer is a feature of the quantum system that controls data exchange operations.

37. The system of claim 30, wherein the sensing, layers / shells and connections can be composed of any devices made of biological, organic, chemical, electronic, electrical, thermal, magnetic, acoustic, mechanical or other elements / materials, using non-AI technology or AI technology such as CNN (Convolutional Neural Network).

Images