BR102020024530A2 - VENDING MACHINE - Google Patents

Abstract

vending machine. The present invention patent describes a technology applied to points of sale by means of breweries, refrigerators or freezers or any other device that can be used as an exhibitor or storage place for products that allows the sale of items in an automated way without the need for a billing unit attached to the equipment for payment in cash or credit card where the point of sale identifies which items were removed by the user and when closing the door of the point of sale such items are automatically charged to the user by the system.

Description

VENDING MACHINE APPLICATION FIELD

[0001] The present invention patent describes a technology applied to points of sale through breweries, refrigerators or freezers, or any other device that can be used as an exhibitor or product storage place that allows the sale of items in an automated way. without the need for a billing unit attached to the equipment for payment in cash or credit card where the point of sale identifies which items were removed by the user and when closing the door of the point of sale such items are automatically charged to the user by the system .

STATE OF ART

[0002] Automatic item vending machines are already well known in the state of the art, however, such machines require the user to carry out a payment procedure in cash or by credit or debit card, tokens or any other form of payment, in machine before being able to pick up the chosen item. The purchase operation on these machines is a little time consuming and such machines have many mechanical and electromechanical mechanisms that make such equipment expensive and requiring frequent maintenance.

[0003] Document US20150235202A1 (Eazy Coin Corp., published 08/20/2015) discloses a system and method for enabling cashless purchases at a vending machine. A conventional vending machine can be modified by connecting a Bluetooth transceiver to it that can communicate with a mobile phone's Bluetooth interface. The vending machine's Bluetooth interface can be connected to an MDB bus to communicate with the vending machine's VMC processor. The VMC processor can be reprogrammed so it can query the vending machine's Bluetooth reader. The VMC can then poll the Bluetooth interface to detect a payment signal sent by the mobile phone and deliver an item, requested via the mobile phone's graphical user interface, to a customer.

[0004] However, this art system differs from the present invention because it uses the connection with the user's mobile device via Bluetooth. In addition to the need for the user's mobile device to have a chip dedicated to communication via bluetooth, by document US20150235202A1 it is possible to observe that the user "authenticates" in the vending machine through this type of connection, which may represent a vulnerability for the system as there is an open door for direct connection between the user and the equipment.

[0005] A technical limitation of the bluetooth connection is also that it supports up to 7 devices connected at the same time to a "master" device (https://en.wikipedia.org/wiki/Bluetooth). Therefore, if the system allows multiple connections at the same time, in a scenario where 7 users remain within the range of the vending machine, all connected by some eventuality, other users could not connect to the same point of sale.

[0006] If the connection of a new user to the equipment forces the termination of a previously existing connection of another user to the same equipment, this behavior can be used to impair the functioning of this connection process between users and equipment.

[0007] Another point that differs from the technology of the present invention is that in the system described in the document US20150235202A1, the user has an account linked to the registration of the same, with a "recharge" system via means of payment to consolidate a balance necessary for use. Before withdrawing the desired product, the user must select the amount to be used to purchase the desired product(s).

[0008] At the points of sale of the present invention, this charge occurs automatically on the user's card, linked at the time of registration, without the need to choose values or products, right after the withdrawal of the desired object(s)( s).

[0009] In the system described in the document US20150235202A1, after selecting the value for use, the user selects the product and it falls into an accessible compartment, without the exposure glass of the equipment being opened, unlike the system of the present patent in which the equipment door is unlocked so that the user can choose and remove the desired product. Given the complication between "authenticating the purchase" in the system and removing the desired product(s), the present invention sought to facilitate this purchase process for the convenience of the user.

[0010] Document CN207503332U (QINGDAO HIRON COMMERCIAL COLD CHAIN CO LTD, published 6/15/2018) describes a smart vending chest freezer that allows the consumer to register as a member or to discern the member's identity through module configuration management and information acquisition module, main control unit and automatic switching door with door opening control system, get response from the goods selection system and obtain goods control action at the time of pre-determining whether the door is closed or if it is kept open, price calculation of goods, and finally, perform smart vending processing in order to receive the door closing signal computing the amount of goods, improving the flexibility of selling soft drinks and snacks , but also convenience, had higher security performance.

[0011] The document CN207503332U describes a horizontal freezer, different from the equipment used in the present invention (but the company is not limited to this type of equipment). Another specification in document CN207503332U is that this equipment is intended for cold foods, such as ice cream. In the equipment in question, a device is used that detects that an item has been removed from inside it, automatically triggering the door to close. The system of this patent comprises inductive sensors and a motor for automatic movement of the door, which can entail a high cost for the equipment, in addition to offering risks of contamination for food and beverages due to elements such as greases and oils, mechanical devices such as motors tend to have more frequent failures and demand higher maintenance costs.

[0012] In this sense, the present invention aims to meet a need to provide an intelligent solution and at the same time simplify the equipment needed to provide a good user experience combined with low cost of manufacturing/maintenance of the equipment.

[0013] Document US10037509B1 (Amazon Technologies Inc., published 7/31/2018) describes a system and method for efficiently tracking inventory using RFID tags (e.g. RFID tags). For example, an RFID tag reader may perform an inventory mode to receive RFID tag identifiers from RFID tags located within a range of an antenna element of the RFID tag reader. These RFID tag identifiers can be included in an inventory table associated with the antenna element. In a second mode, the RFID tag reader can perform a targeted reading. During a directed read mode, the RFID tag reader, using the antenna element, can issue a request for a selected RFID tag identified in the inventory table to respond to the request.

• 1) A necessidade de etiquetar cada um dos produtos a serem identificados adiciona um processo custoso de preparação dos mesmos, o que pode inviabilizar determinadas operações, principalmente em escalas maiores e/ou produtos com alto giro;
• 2) Este sistema é dependente de que a TAG do produto o acompanhe na movimentação que ocorre devido ao consumo do mesmo. No caso da TAG ser (intencionalmente ou não) removida e permanecer no mesmo local pode acarretar em prejuízos para o comercializador, visto que nesse o sistema não detectaria o consumo do produto em questão. Considerando que a presente invenção atua com bebidas e produtos refrigerados, que podem ser molhados ou apresentar umidade na embalagem, este é um fator importante a ser considerado visto que a etiqueta pode se desprender mais fácil do objeto consumido nessas condições; e
• 3) A tecnologia de RFID é sensivelmente afetada por líquidos. Em caso de derramamento de líquido no local onde a TAG de respectivo produto foi adesivada, por exemplo, pode ocorrer detecção incorreta ou não detecção.

[0014] The system developed by Amazon uses RFID tags, which allow the identification of products on shelves or other locations. For such product identification to occur, each product in the inventory must have a tag (TAG) that allows its identification. With the correct identification of each product, it is possible to monitor the stock and the output of a given product since, with its removal, the label is no longer recognized by the inventory identification system. Thus, two major problems can be cited for the use of this technology in moderate and large scale operations:

• 1) The need to label each of the products to be identified adds a costly process of preparing them, which can make certain operations unfeasible, especially on larger scales and/or products with high turnover;
• 2) This system is dependent on the product's TAG accompanying it in the movement that occurs due to its consumption. In the event that the TAG is (intentionally or not) removed and remains in the same place, it may result in losses for the trader, since in that case the system would not detect the consumption of the product in question. Considering that the present invention works with beverages and refrigerated products, which can be wet or have moisture in the packaging, this is an important factor to be considered, since the label can easily come off the object consumed under these conditions; and
• 3) RFID technology is sensitively affected by liquids. In case of spillage of liquid in the place where the TAG of the respective product was affixed, for example, incorrect detection or no detection may occur.

[0015] Document US10140483B1 (Amazon Technologies Inc., published 11/27/2018)
describes the ability to efficiently track inventory using RFID tags (eg RFID tags). For example, a stock shelf located within a material handling facility may include a series of antenna elements arranged in rows built into the shelf that are coupled to an RFID reader. Antenna elements are configured to establish electromagnetic fields oriented along rows of shelves. Fields can be used to read RFID tags included or attached to inventory items placed on the shelf row.

[0016] The same considerations made in relation to document US10037509B1 above apply to document US10140483B1 on the use of RFID.

[0017] Document US10268983B2 (Amazon Technologies Inc., published 4/23/2019) describes a system for tracking the removal or placement of items in inventory locations with a material handling facility. In some cases, a user may remove an item from an inventory location and the inventory management system may detect this removal and update a user's item list associated with the user to include an item identifier representative of the removed item. Likewise, if the user places an item in an inventory location, the inventory management system can detect this placement and update the user's item list to remove an item identifier representative of the placed item.

• 1) Análises constantes de imagens ou análises de vídeos demandam sistemas computacionais mais robustos e geralmente mais caros, além de integrações mais complexas no caso de espaços maiores que demandem mais de uma câmera para detecção; e
• 2) Dependendo do ângulo de captura das imagens ou vídeos, podem ocorrer casos em que a própria mão do usuário esconda (involuntariamente ou não) o objeto retirado, impedindo sua detecção pelo sistema.

[0018] In this patent document, a system capable of detecting the removal of a certain object through cameras was presented. As soon as the user's hand passes through a previously defined region or plane, the system analyzes and through computational resources processes such images (or videos) to analyze whether the user's hand leaves the region of interest with a certain object in hand and, in this In this case, adjusts the respective quantity of the product in the inventory. Some limiting factors for this type of system are:

• 1) Constant image analysis or video analysis demand more robust and generally more expensive computer systems, in addition to more complex integrations in the case of larger spaces that require more than one camera for detection; and
• 2) Depending on the angle of capture of images or videos, there may be cases in which the user's own hand hides (involuntarily or not) the removed object, preventing its detection by the system.

[0019] In order to overcome such technical problems, the present invention implements a solution that does not depend on video processing or otherwise that requires more robust processing and solved this problem of inappropriate positions and angles for detection by creating an isolated system, in which the images are always captured with the equipment door closed, without external interventions, so that the detection takes place efficiently.

[0020] Document CN209819974U (published on 12/20/2019) proposes a smart vending horizontal freezer and a smart vending system. The smart selling horizontal freezer includes a cabinet body, a door body and a wireless network camera. The cabinet body has an accommodation chamber and a delivery port. Above the chamber, the port is set at the output, the wireless network camera is fixedly connected to the cabinet, there are several wireless network cameras, and the various wireless network cameras are arranged along the circumferential gap of the side wall of the cabinet. . The wireless network camera has an image acquisition unit and a network signal output unit. The network signal output unit is in communication with the image acquisition unit and the network signal output unit is in connection with the cloud to send the acquired image signal to the cloud; The smart vending system includes the above-mentioned smart vending horizontal refrigerator. The utility model improves the reliability of product identification by the smart vending chest freezer, thus reducing the equipment and operating costs of the smart vending chest freezer.

[0021] The technology described in the document CN209819974U can be considered the closest to the present invention. However, it is used in a horizontal freezer and which uses 4 cameras on the same plane to detect products removed from the main chamber. The present invention applies to points of sale that sell, mainly, beverages and as in this segment the products are presented vertically (bottles and cans "standing"), it is necessary to use shelves. In this scenario, it becomes economically unfeasible to use 4 cameras per shelf for the proper detection of consumed objects, as in the case of this Chinese patent document.

[0022] Document US2020041200A1 (LG ELECTRONICS INC [KR], published 06/02/2020) describes an artificial intelligence refrigerator including a communication unit, a camera configured to capture an image of a tray including a plurality of slots for store food and a processor configured to control the communication unit and camera. The processor extracts a marker hidden by the food from a plurality of markers respectively attached to the plurality of slots of the captured image from the tray, acquires a position identifier of a slot corresponding to the hidden marker and a reception time indicating that the food is received in the slot , transmits the acquired position identifier and reception time to a server via the communication unit, receives certain color information from the server based on the reception time, and controls a light emitting element provided in the slot to send a color to be displayed in the slot included in the received color information.

[0023] In this system of document US2020041200A1 cameras detect hidden marks in a set of grooves on shelves. Although the slots are optional, this patent is based on these markings to detect products entering or leaving the refrigerator. For this, it is necessary that each product is placed in a specific location, above the respective marking, which, in an operation with users removing items from the refrigerator, would force the shelf to have some kind of delimiter so that the objects do not move and leave the shelves. markings. Such delimiters can make system shelves and operation more expensive. Also, in a system that mainly sells beverages, items that are arranged vertically for better marketing, the tags can be hidden by items between the tags and the cameras. One solution would be to work with cameras in an upright position, facing downwards, but due to camera limitations and focal lengths needed to capture the entire region of interest (shelf with products), the shelves would need to be at a considerable distance from the camera, which it would reduce the capacity of the refrigerator and generate a more expensive process for the commercialization of these products.

SUMMARY OF THE INVENTION

[0024] The present invention aims to detect, identify, classify and account for different items removed from a point of sale implemented through breweries, refrigerators or freezers and charge the items removed by the customer without having to stop to carry out the payment transaction.

[0025] The present invention was developed with a technology that makes it possible to transform a conventional brewery, refrigerator or freezer, or other type of equipment, into a fully automated point of sale, capable of recognizing, accounting and charging the items withdrawn directly by the end user. .

[0026] Such technology has as main objective to offer a differentiated consumption experience to the users of the equipment and system, combining a safe, efficient and low cost solution.

[0027] For the present invention to be able to deliver an adequate operation, the technology has its own equipment, a computer, at each point of sale ("hardware" and "software") capable of controlling access to the products, and such a computer the point of sale is connected, through the internet, to a central server that coordinates and records the entire operation.

[0028] In this operation, such points of sale are connected to the server via the internet at all times and can be monitored in real time.

[0029] Although the present invention also makes use of QR code for user authorization, unlike other existing technologies, the system of the present invention uses detection via Computer Vision (type of applied Artificial Intelligence) for the correct detection of removed items.

BRIEF DESCRIPTION OF THE FIGURES

[0030] Figure 1 illustrates the point of sale implemented by a brewery, refrigerator or freezer of the present invention and its components.

[0031] Figure 2 is a diagram of how the points of sale implemented through breweries, refrigerators or freezers of the present invention interact with the system that controls the operations of all points of sale.

[0032] Figure 3 illustrates the flowchart of processes at points of sale.

[0033] Figure 4 illustrates a block diagram with the hardware representation of each point of sale.

DETAILED DESCRIPTION OF THE INVENTION

[0034] The smart point of sale (1) of the present invention is composed of: a refrigerated chamber (2);
a door (3);
a magnetic lock (4);
shelves (5);
indication lights (6);
a computer (7);
a QR code (8); and
cameras (9).

[0035] The refrigerated chamber (2) of the intelligent point of sale (1) is closed by a door (3) equipped with a magnetic lock (4).

[0036] On each level of shelves (5) of the point of sale (1) there are cameras (9) that record images of the items arranged on each shelf.

[0037] Such images are processed by an internal computer (7) that processes the images to be transmitted to the server (12) by means of communication with the internet (11).

[0038] The door (3) of the point of sale (1) is equipped with a QR code (8) that allows the identification and connection between the point of sale (1) and the user's cell phone (10). This connection between the point of sale (1) and the cell phone (10) is made through the internet (11) where the point of sale (1) is connected to the internet (11) by a means of communication and the cell phone (10) ) of the user is connected to the internet by mobile protocol, such as 3G, 4G or 5G or similar, or by a public or private network of the place where the point of sale (1) is installed.

[0039] When the user scans the QR code (8) with his cell phone (10), the cell phone (10) communicates with the point of sale (1) and after user recognition, the point of sale (1) opens the door (3) allowing the user to remove the products from inside the point of sale (1).

[0040] In general, the functionality of the control system at each point of sale (1) is summarized to the authentication of users, for the proper opening of the door (3), and exchange of information with a central server (12).

[0041] During this process, several integrity tests are periodically carried out at all points of sale (1), among other activities controlled jointly by the "hardware" and "software". In an overview, the process flow can be demonstrated by the flowchart in Figure 3.

[0042] When connecting the point of sale (1) the system is initialized, step (30), and the images of the shelves (5) are stored to measure the items in the point of sale and the control variables of the point of sale software (1) running on the internal computer (7) are initialized, the software of the internal computer (7) closes the magnetic latch (4) of the door (3) and turns on the red light of the indicator lights (6) informing that the point of sale (1) has the door (3) closed.

[0043] When a user approaches his cell phone (10) and reads the point of sale QR code (1) the application installed on the cell phone connects to the computer software (7) of the point of sale (1) and this checks if the user is registered and enabled to use the point of sale (1), step (31). If the user is not registered and/or enabled, the software is waiting for a new reading of the QR code, step (32).

[0044] If the user is enabled, the software of the internal computer (7) activates the magnetic lock (4) of the door (3) of the point of sale (1) allowing the user to open the door (3), step (33) ) . To signal the user that the door has already been opened, the internal computer software (7) illuminates the indicator lights (6) green.

[0045] While the user takes the items from the point of sale (1) the system is in a loop checking the status of the door (3), step (34), if it is still open, the system is waiting, step (35) . When the user finishes removing the items and closes the door (3), the system captures the images of the shelves (5) with the cameras (9), step (36). The images are processed by the computer software (7) of the point of sale (1).

[0046] After processing the images, they are sent to the server (12) via the internet (11), step (37).

[0047] The "hardware" responsible for the point of sale control system (1) and the respective accessories necessary for the system to function, is detailed in Figure 04.

[0048] The point of sale (1) consists of a brewery, refrigerator, or freezer, consisting of a refrigerated chamber (2) and a door (3), in which a power supply (40), a computer (7), a control board (41), cameras (9) and a lock or magnetic latch (4).

[0049] The flow of information related to the Computer Vision system, which can be executed either locally, on the computer (7) at each point of sale (1), or centrally, on the server (12) responsible for several points of sale (1), is illustrated in figure 05.

[0050] When the door (3) is closed, the cameras (9) capture the images of the shelves (5) of the point of sale (1) and these images are treated and analyzed by the Computer Vision System (50) implemented through artificial intelligence, which can be installed on the computer (7) internal to the point of sale (1) when there is only one point of sale or installed on the server (12) when there is more than one point of sale (1) being monitored and connected in network. The Computer Vision System (50) identifies which and how many items were removed from the point of sale (1) by the user and sends this information to the Collection System (51) which charges the payment method registered in the mobile application (10 ) of user.

• 1) Captura as imagens das prateleiras com as câmeras (9) antes do usuário abrir a porta;
• 2) Captura as imagens das prateleiras com as câmeras (9) depois do usuário fechar a porta; e
• 3) Através da diferença entre as quantidades detectadas por inteligência artificial (I.A.), são detectados os produtos consumidos.

[0051] The Computer Vision System (50) detects the quantities consumed, through the following steps:

• 1) Captures the images of the shelves with the cameras (9) before the user opens the door;
• 2) Captures the images of the shelves with the cameras (9) after the user closes the door; and
• 3) Through the difference between the quantities detected by artificial intelligence (AI), the products consumed are detected.

[0052] At the end of the charge, the Billing System (51) sends a communication to the user's mobile application (10) informing the items charged, the total amount of the purchase made and the payment confirmation.

[0053] The present invention also provides a Sales System, for use with the intelligent point of sale (1) comprising:
at least one smart sales point (1), equipped with a refrigerated chamber (2), a door (3), a magnetic lock (4), shelves (5), indicator lights (6), a computer (7) , a QR code (8), cameras (9), a power supply (40), and a control board (41);
a server (12) equipped with a program to process the information from the images sent by the smart points of sale (1) and account for the withdrawal of items by the user and send the cell phone (10) the charge for the items removed from the smart point of sale (1); and
a means of communication with the internet (11) to make the connection between the server (12), the smart points of sale (1) and the user's cell phone (10), where this connection between the point of sale (1) and the cell phone (10) is made through the internet (11) where the point of sale (1) is connected to the internet (11) by a means of communication and the user's cell phone (10) is connected to the internet by mobile protocol, such as 3G, 4G or 5G or similar, or over a public or private network of the place where the point of sale (1) is installed.

• - passo (30): ligar o ponto de venda (1) inicializando o sistema, as imagens das prateleiras (5) são armazenadas para mensuração dos itens constantes do ponto de venda (1) e as variáveis de controle do software do ponto de venda (1) executado no computador interno (7) são inicializadas, o software do computador interno (7) fecha o fecho magnético (4) da porta (3) e acende a luz vermelha das luzes de indicação (6) informando que o ponto de venda (1) está com a porta (3) fechada;
• - passo (31): quando um usuário aproxima seu celular (10) do ponto de venda (1) e lê o código QR o aplicativo instalado no celular (10) se conecta ao software do computador (7) do ponto de venda (1) e este verifica se o usuário está cadastrado e habilitado para usar o ponto de venda (1);
• - passo (32) : caso o usuário não esteja cadastrado e/ou habilitado, o software fica aguardando uma nova leitura do código QR;
• - passo (33): caso o usuário esteja habilitado, o software do computador interno (7) aciona a placa de controle (41) que abre o fecho magnético (4) da porta (3) acendendo a luz verde das luzes de indicação (6) do ponto de venda (1) permitindo ao usuário abrir a porta (3);
• - passo (34) : enquanto o usuário pega os itens do ponto de venda (1) o sistema fica em loop verificando o status da porta;
• - passo (35) : caso a porta (3) esteja aberta, o sistema fica aguardando;
• - passo (36) : quando o usuário termina de retirar os itens e fecha a porta (3) as câmeras (9) capturam as imagens das prateleiras (5) do ponto de venda (1) e estas imagens são tratadas e analisadas pelo Sistema de Visão Computacional (50) implementado através de inteligência artificial, que pode estar instalado no computador (7) interno ao ponto de venda (1) quando se tem apenas um ponto de venda ou instalado no servidor (12) quando há mais de um ponto de venda (1) sendo monitorados e ligados em rede, o Sistema de Visão Computacional (50) identifica quais e quantos itens foram retirados do ponto de venda (1) pelo usuário; e
• - passo 37: enviar ao servidor (12) através da internet (11) a identificação e quantidade de itens retirados pelo usuário e o Sistema de Cobrança (51) efetua a cobrança no meio de pagamento cadastrado no aplicativo do celular (10) do usuário e enviar ao usuário uma mensagem indicando os itens cobrados e o valor total cobrado.

[0054] The point of sale (1) and the Sales System implement a sales method containing the following steps:

• - step (30): turn on the point of sale (1) by initializing the system, the images of the shelves (5) are stored to measure the items contained in the point of sale (1) and the control variables of the point of sale software (1) running on the internal computer (7) are initialized, the software of the internal computer (7) closes the magnetic latch (4) of the door (3) and turns on the red light of the indicator lights (6) informing that the sale (1) has the door (3) closed;
• - step (31): when a user approaches his cell phone (10) to the point of sale (1) and reads the QR code, the application installed on the cell phone (10) connects to the computer software (7) of the point of sale (1) ) and this checks if the user is registered and enabled to use the point of sale (1);
• - step (32): if the user is not registered and/or enabled, the software is waiting for a new reading of the QR code;
• - step (33): if the user is enabled, the software of the internal computer (7) activates the control board (41) that opens the magnetic lock (4) of the door (3) turning on the green light of the indicator lights ( 6) from the point of sale (1) allowing the user to open the door (3);
• - step (34) : while the user picks up the items from the point of sale (1) the system is in a loop checking the status of the door;
• - step (35) : if the door (3) is open, the system is waiting;
• - step (36) : when the user finishes removing the items and closes the door (3) the cameras (9) capture the images of the shelves (5) of the point of sale (1) and these images are treated and analyzed by the System of Computer Vision (50) implemented through artificial intelligence, which can be installed on the computer (7) internal to the point of sale (1) when there is only one point of sale or installed on the server (12) when there is more than one point of sale (1) being monitored and networked, the Computer Vision System (50) identifies which and how many items were removed from the point of sale (1) by the user; and
• - step 37: send to the server (12) via the internet (11) the identification and quantity of items withdrawn by the user and the Billing System (51) charges the payment method registered in the user's mobile application (10) and send the user a message indicating the items charged and the total amount charged.

Claims (3)

Vending machine characterized by comprising:
a refrigerated chamber (2);
a door (3);
a magnetic lock (4);
shelves (5);
indication lights (6);
a computer (7);
a QR code (8);
cameras (9);
a power supply (40); and
a control board (41)
Where
the refrigerated chamber (2) of the smart point of sale (1) is closed by a door (3) equipped with a magnetic lock (4) activated by the control board (41) and the refrigerated chamber (2) has indicator lights ( 6) that signal the user if the door (3) is closed, red light on, or if the door (3) is open, green light on;
the door (3) of the point of sale (1) is equipped with a QR code (8) that allows the identification and connection between the point of sale (1) and the user's cell phone (10), and when the user makes the QR code (8) with your cell phone (10), the cell phone (10) communicates with the point of sale (1) and after user recognition, the point of sale (1) opens the door (3) allowing the user removes products from inside the point of sale (1);
on each level of shelves (5) of the point of sale (1) there are cameras (6) that record the images of the items arranged on each shelf, and the images are processed by an internal computer (7) that processes the images through the system of computer vision (50) based on artificial intelligence to be transmitted to the server (12), where the images are recorded when the point of sale (1) is turned on in order to account for the items initially included in the point of sale and when the user closes the door (3); and
a power supply (40) that maintains the operation of the computer (7), the cameras (9), the lock (4), and the control board (41). System, for use with the vending machine (1) of claim 1, characterized in that it comprises:
at least one smart sales point (1), equipped with a refrigerated chamber (2), a door (3), a magnetic lock (4), shelves (5), indicator lights (6), a computer (7) , a QR code (8), cameras (9), a power supply (40), and a control board (41);
a server (12) equipped with a program to process the information from the images sent by the smart points of sale (1) and account for the withdrawal of items by the user and send the cell phone (10) the charge for the items removed from the smart point of sale (1); and
a means of communication with the internet (11) to make the connection between the server (12), the smart points of sale (1) and the user's cell phone (10), where this connection between the point of sale (1) and the cell phone (10) is made through the internet (11) where the point of sale (1) is connected to the internet (11) by a means of communication and the user's cell phone (10) is connected to the internet by mobile protocol, such as 3G, 4G or 5G or similar, or over a public or private network of the place where the point of sale (1) is installed. Vending method, for use with the vending machine (1) of claim 1 and the system of claim 2, characterized in that it contains the following steps:

• - step (30): turn on the point of sale (1) by initializing the system, the images of the shelves (5) are stored to measure the items contained in the point of sale (1) and the control variables of the point of sale software (1) running on the internal computer (7) are initialized, the software of the internal computer (7) closes the magnetic latch (4) of the door (3) and turns on the red light of the indicator lights (6) informing that the sale (1) has the door (3) closed;
• - step (31): when a user approaches his cell phone (10) to the point of sale (1) and reads the QR code, the application installed on the cell phone (10) connects to the computer software (7) of the point of sale (1) ) and this checks if the user is registered and enabled to use the point of sale (1);
• - step (32): if the user is not registered and/or enabled, the software is waiting for a new reading of the QR code;
• - step (33): if the user is enabled, the software of the internal computer (7) activates the control board (41) that opens the magnetic lock (4) of the door (3) turning on the green light of the indicator lights ( 6) from the point of sale (1) allowing the user to open the door (3);
• - step (34) : while the user picks up the items from the point of sale (1) the system is in a loop checking the status of the door;
• - step (35) : if the door (3) is open, the system is waiting;
• - step (36) : when the user finishes removing the items and closes the door (3) the cameras (9) capture the images of the shelves (5) of the point of sale (1) and these images are treated and analyzed by the System of Computer Vision (50) implemented through artificial intelligence, which can be installed on the computer (7) internal to the point of sale (1) when there is only one point of sale or installed on the server (12) when there is more than one point of sale (1) being monitored and networked, the Computer Vision System (50) identifies which and how many items were removed from the point of sale (1) by the user; and
• - step 37: send to the server (12) via the internet (11) the identification and quantity of items withdrawn by the user and the Billing System (51) charges the payment method registered in the user's mobile application (10) and send the user a message indicating the items charged and the total amount charged.

Images