KR102114808B1 - An Autonomous Convenience Store in the type of Vending Machine of the Container with using Internet of Things, Bigdata, Deep Learning and Block Chain Technology - Google Patents

Abstract

The present invention relates to a concept of the construction and design of a vending machine type convenience store implemented in a container through a smartphone, a method of operating multiple convenience stores, and a platform on the cloud. In order to order goods at convenience stores, when customers order their desired items online using a smartphone as an operating service platform, a convenience store located near the customer is selected and connected, and the goods stored on the shelves in the convenience store's container are controlled by a control system. It is a sales method that adjusts, picks and packs on the shelf and delivers it to the customer through an exit window in front of the container. Here, the present invention relates to an automated facility inside a container for implementing a convenience store function described in the present invention. First, a robot facility for transporting goods is an X axis-Y axis on the upper part of the container and a design to move up and down and a lower part behind the container. When the goods are received and the goods are delivered to the lower part of the front of the container, the transport robot is located at the entrance where the goods enter and picks up the goods while moving toward the exit. After placing the goods on the aisle between the storage shelf and the shelf, the transport robot attached to the Y-rail bar moves to the desired pallet position, increases and decreases up and down, arrives at the pallet, and uses the RFID chip and reader attached to the ordered robot arm It is to pick up the goods and put them in a cart, take them to the exit, pack them, and provide them to the customer. In order to make this automation efficient and optimal, the spatial position of the stacked goods storage pallet in the container was coordinated with XYZ and the moving distance, cost and time of the transport robot passing through it were quantified, and this was set as the target function of optimization. As a core concept, we proposed an algorithm that bundles orders with similar routes through deep learning in a way that efficiently responds to multiple orders at the same time through deep learning. In addition, since it is a fully automatic unmanned convenience store, all payments are made with a smartphone, and we propose a blockchain-based dedicated virtual currency as a safer and more efficient transaction method.
As an effect on the present invention, first, a product shelf, a customer shopping space, a cash register, etc., which occupy the most space in an existing convenience store, are eliminated, and a customer searches for a product through a smartphone and makes an order payment, and only stores the product in a container. By automatically picking and packing in response to orders through automatic facilities to provide them to customers, it is possible to dramatically reduce rental and maintenance costs for maintaining existing stores. In particular, it can be implemented in a standardized container and simply installed to eliminate the cost of installing the store interior and other facilities.
Second, the convenience store can be quickly installed in a small space, so it can be easily placed in more areas, so that it can be placed in a location that is most economically accessible to the customer according to the customer's demand, and can be easily moved if necessary. In addition, it builds big data by accumulating all information about customers and products, and makes it easy to build an optimal and efficient logistics system by applying deep learning using history.
Third, by operating as a fully unmanned system, it is possible to reduce the labor cost burden of hiring a 24-hour management clerk.
Fourth, by making payment settlement through a blockchain-based secure currency, financial costs such as fees for small-scale small-scale product sales can be reduced with cash.
Finally, by creating these platforms with an open API (application association) Structure franchise operators and goods suppliers, including a number of projects to take advantage of this by utilizing 3 ^rd Party as a model to be developed. In particular, efficient delivery service may be possible through linkage with logistics companies. Therefore, through the rental of open platforms on the cloud and the provision of standardized containers, a virtuous cycle ecosystem of the business will be built to enable activation through more dissemination.

Description

An Autonomous Convenience Store in the type of Vending Machine of the Container with using Internet of Things, Bigdata, Deep Learning and Block, implemented as a container using the Internet of Things, big data, deep learning and blockchain technology. Chain Technology}

In the present invention, as a technology that replaces the function of a convenience store opened around the customer in the form of a fully unmanned vending machine around the customer, it is to implement a function that allows customers to easily purchase a variety of low-cost, small-sized items that are needed by customers anytime, anywhere. This is a container-based vending machine, which is an automated robot design and RFID-based sensor and communication technology that can automatically store, pick and pack goods in containers, and big data and deep learning-based logistics that optimize supply to meet real-time demands of customers. It includes technology and location-based app technology for efficient provision of purchase information using smartphones, and finally, blockchain-based secure money technology for payment.

Since vending machine-type convenience stores are planned to occupy small spaces in many places, they must be implemented and installed in standardized containers. Therefore, there is a need for an automated technology based on sensors and robots to efficiently and optimally perform the process of placing, storing, picking, and packing items in a small space. In addition, since most convenience store goods are low-cost and small-scale, there is a need for a technology for optimizing logistics that provides them appropriately so that they do not run out of stock or leave inventory according to real-time demand. This requires machine learning and deep learning technology that enables automatic learning to be performed quickly based on the purchase history as continuous purchase is performed.

In addition, unlike conventional vending machines, since the purchase of goods is made in a closed container, a web or app technology is required so that the goods can be easily provided through a smartphone. In particular, it is necessary to provide an app in which a service based on the location of the customer and the location of the vending machine is grafted.

Finally, as a technology for payment, a virtual cryptocurrency that can replace cash suitable for a large amount of micropayment is suitable. Blockchain technology is needed for this.

In the present invention, while implementing the functions of the existing convenience store to solve the problem.

Existing convenience stores should provide low cost and small quantity of various goods that customers need anytime, anywhere. To this end, customers are usually invited to visit and select and purchase items.

First, a large store space is required. In particular, most of the places where customers can easily access are expensive. Therefore, the cost of renting a store is the biggest burden on costs. If a vending machine-type container replaces a store, it is a convenience store with only a storage warehouse, which can drastically reduce management costs including rental and energy costs.

Second, a 24-hour clerk's residence is required for management and payment. The resulting labor cost is another great burden on costs. Therefore, it is possible to almost eliminate labor costs by automatically searching and selecting it through a smartphone.

Third, as it operates on a franchise basis, items meet the supply and demand of items according to their sales history. This can be implemented more efficiently through location-based big data and deep learning-based logistics optimization programs.

Lastly, purchase a variety of small items at low prices at convenience stores, most of them pay by credit card. In this case, the fees incurred can be a burden, but if you use the virtual security currency as cash, it can be greatly reduced. It will make a dedicated security currency using blockchain technology or link it with an existing security currency.

The present invention provides a solution for the four problems described above.

First, it is a concept of a container-based vending machine convenience store as a means to dramatically reduce the store rent by reducing the space occupied by existing convenience stores to the maximum. This is to efficiently store all the items stored in the existing store in the container. This is based on the list of items that have been ordered with a smartphone, and the picking machine or robot calculates the optimal route, picks it, packs it, waits for it, and provides it when the customer arrives. Here, the core technology is an algorithm technology that allows customers to find it by quickly and efficiently preparing and responding to various purchase requests. In addition, it is an automation technology through the design of a robot that will efficiently perform this. Here, the cost of the automation robot will be the largest part of the investment cost for initial container production and installation. However, by minimizing the cost through technology development and mass production, it will be possible to make the burden much less for the benefits such as rent and labor costs that are reduced through this.

Second, labor costs will be eliminated by eliminating the need for resident staff to operate and pay at existing convenience stores. This can be solved by allowing the customer to search for the product on the smartphone and select and pay for it, so that only the pickup is picked up in the field. This will be possible to minimize the waiting time in front of the vending machine so that there is no inconvenience by efficiently packing and waiting by estimating the location and arrival time of the customer.

Third, by installing and operating a vending machine convenience store in a number of regions in the form of a franchise, we are optimizing the supply and demand of goods based on sales history. By performing this through deep learning based on big data, it is possible to implement an optimized logistics system more efficiently. In addition, it will be possible to implement a more efficient logistics system by optimizing the out-of-stock and inventory management in real time based on the location of convenience stores and customers.

Lastly, as a way to dramatically reduce the burden of fees associated with card payments, a virtual security currency using blockchain is applied. Initially, existing cryptocurrencies such as Bitcoin can be used, but in the end, it will be possible to propose and build a virtual security currency for this. In addition, it is possible to install devices or functions that can use cash or card at the same time in the initial stage, but ultimately, it will ensure that all payments of goods are made through virtual security currency, so that efficient payments can be made. Furthermore, through the spread of dedicated virtual security currency, consumers and producers of one item will be able to make real-time payments for each item.

As described above, in the present invention

First, it eliminates product shelves, customer shopping spaces, and checkout counters, which occupy the most space in existing convenience stores, and allows customers to search for products and pay for orders through smartphones, keeping products in containers and responding to orders through automatic facilities. By efficiently picking and packing and providing it to customers, it is possible to dramatically reduce rental and maintenance costs for maintaining existing stores. In particular, it can be implemented in a standardized container and simply installed to eliminate the cost of installing the store interior and other facilities.

Second, since a convenience store can be quickly installed in a small space, it can be easily placed in a larger area, so that the convenience store can be placed in a location that is most economically accessible to the customer according to the customer's demand, and can be easily moved if necessary. In addition, it builds big data by accumulating all information about customers and products, and makes it easy to build an optimal and efficient logistics system that applies deep learning by utilizing history.

Third, by operating as a fully unmanned system, it is possible to reduce the labor cost burden of hiring a 24-hour management clerk.

Fourth, by making payment settlement through a blockchain-based secure currency, financial costs such as fees for small-scale small-scale product sales can be reduced with cash.

Finally, there is created such a platform with open API (application association) Structure franchise operators and goods suppliers, including a number of projects to take advantage of this by utilizing 3 ^rd Party as a model to be developed. In particular, efficient delivery service may be possible through linkage with logistics companies. Therefore, through the rental of open platforms on the cloud and the provision of standardized containers, a virtuous cycle ecosystem of the business will be built to enable activation through more dissemination.

1 is a conceptual diagram of a vending machine type unmanned convenience store operating concept in a container according to the present invention
Figure 2 is a vending machine-type unmanned convenience store container interior automation facility configuration according to the present invention
Figure 3 is an implementation of the automatic pick-up picking feature in the container
Figure 4 is a spatial XYZ location of the goods storage pallet in the container and how to respond to the order
5 is a conceptual diagram of quantification and optimization according to the transport path of the transport robot in the container
FIG. 6 is a concept of similar path bundling for responding to multiple orders at the same time.
7 is a conceptual diagram of deep learning for optimizing a common path of a transport robot for responding to multiple orders at the same time.
8 is a conceptual diagram of deep learning for vending machine-type convenience store sales in a container and a block chain-based P2P payment conceptual diagram.

First, with reference to the drawings, a preferred embodiment according to the present invention will be described below with reference to the operating platform service and operating method of a vending machine convenience store through a smartphone.

Figure 1 relates to the operation of the vending machine-type convenience store (100, 101, 102) according to the present invention, customers (200, 201, 202) operating on the cloud the desired product with a smartphone (300, 301, 302) on the cloud When ordering online with the service platform 400, a convenience store located near the customer is selected and connected. This convenience store 100 adjusts the goods 120 stored on the shelves in the container 110 to the control system 140 with the automation facility 130 and picks 150 from the shelves and packs 160 to order customers 170 ) (180).

2 shows a configuration screen for an automated facility for operation inside the container 110 described in the present invention from the top. First, the robot equipment 500 for transporting goods is designed to move up and down the X-axis 501-Y-axis 502 on the upper part of the container. The upper left block 600 of FIG. 2 indicates the entrance into which the goods are received at the lower rear of the container, and the lower right block 700 indicates the exit at which the goods are delivered to the lower front of the container. Initially, the transport robot is located at the entrance 600 where the goods enter and moves toward the exit 700 to pick the goods. Six shelves 510, 511, 512, 513, 514, and 515 for storing goods are displayed inside the container by the robot apparatus 500. The transport robot 500 is designed to pick and transport goods while moving through the four passages 520, 521, 522, and 523 between these shelves. First, the transport robot 500 is hung on the Y-rail bar 530 in the Y-axis 502 direction installed on the upper portion of the container to move in the Y-axis direction. The Y-rail bar 530 has four passage spaces between shelves along the X-axis along another pair of X-rail bars 540 and 550 installed at both ends in the X-axis direction 501 at the top of the container. It is designed to be located at (520, 521, 522, 523). For example, the transport robot 500 is hung on the Y-rail bar 530, and when the Y-rail bar is located in the second passage 521 along the X-rail bar, along the Y-ray bar between the two shelves 511 and 512. You can move to the desired position and move it up and down to the pallet of the desired shelf and pick up the goods for transportation.

In FIG. 3, a method for operating the transport robot 500 among automation facilities for operating inside the container 110 described in the present invention is demonstrated. After the Y-rail bar 530 moves on the X-rail bars 540 and 550 to be positioned on the passage 521 between the shelf 511 and the shelf 512, the transport robot moves to the desired pallet 800 position. (500) is to move up and down to decrease and move down to the moving Z-string (560). When the desired pallet 800 arrives, the RFID chip 820 attached to the robot recognizes the ordered item 810 and the RFID recognizer 590 attached to the robot moves the picker 570 in the form of a robot arm to pick up the cart 580 ). The items on the pallet of the opposite shelf 512 are returned to the transport robot 500. Moving to the pallet with the ordered items, all the ordered items are placed in a cart and taken to the exit 600 to be packed and provided to the customer.

FIG. 4 shows a method of picking and packing an item ordered in this way from a pallet through an order robot in the present invention. First, when the customers 200 and 201 place an order with the service server 400 through the smartphones 300 and 301, online orders 900 and 901 are prepared and delivered to the vending machine convenience store 100, and this order form 900, 901) the order contents are delivered to the convenience store operating system 140 to operate the order robot 500 to pick the goods from the pallet 800 stacked on the shelf in the container to the robot arm 570 through the RFID sensor 590. Pack it in the corresponding customer cart 580 to pack it. 4 shows an example. As shown in FIGS. 2 and 3, XYZ coordinates the spatial position of 300 pallets 800 on 6 shelves within a container, and subdivides them into 6 X-axis, 10 Y-axis, and 5 Z-axis coordinates. Doing. That is, the specific pallet 801 shown in FIG. 4 is a pallet located in the fourth row (Y4) of the first shelf (X1) and the second column (Z2) from the top (X1, Y4, Z2) or (1, 4,2). Therefore, 300 pallets are displayed from (1,1,1) to (6,10,5).

In FIG. 4, as an example, the order items of two specific customers 200 and 201 are displayed in red (R) and green (G), respectively, and the order details appearing in the list of the respective orders (900, 901) are displayed on the pallet drawing. It is colored. When a red customer and a green customer order the same item or items on the same pallet, the two colors are displayed on the pallet 802 together.

When ordering 4 items each, as described above, the transport robot 500 arrives while moving the 4 pallet positions indicated in the order in the XYZ direction, picks up the desired number of items, puts them in the customer's cart, and finally packs them ( 180, 181) to provide to customers 170, 171.

In order to respond to the customer's order in this way in the most efficient manner, it is necessary to minimize the moving distance of the transport robot 500 first. In addition, if multiple customers place orders at the same time, there is a need for a way to best handle them. To this end, the service operation server on the cloud applies a deep learning algorithm based on big data.

5, in the present invention, for the methodology for optimizing the operation of the transport robot 500, the movement path of the transport robot according to the position coordinates of the pallet in the container was quantified. In other words, a path for allowing the transport robot 500 to pick the ordered item as the passing of the pallets 801, 802, 803, and 804 marked in red containing the items in the order list 900 of the customer 200 is passed. It is defined as P(R) and a method for calculating the cost function is proposed. The path P(R) for corresponding to the order list 900 is composed of eight partial paths P(R)1, P(R)2, ..., P(R)8. The first partial path P(R)1 910 starts at the starting point of the entrance 600 coordinates (X1, Y1, Z1) and moves the Y-rail bar 530 to the pallet 801 containing the first ordered product. It is a path along which the transport robot 500 moves. At this time, the value of the cost function is Y4. The coordinates of this point are (1,4,1), where the transport robot is located at point Z2, two columns below, using the Z-string 560, picks up the ordered item, puts it in the cart, and goes up again. Therefore, in this case, since it moves up and down, the cost function value is Z4, which is Z2x2. If the cost function is calculated to go through all four pallets in this way, it becomes X6+Y24+Z26. Assuming that the pallet is a cube and each travel distance of XYZ is the same, the final cost function value is 56. Of course, considering the difference between the movement of the X-rail bar and the cost function in the Y-rail bar and Z-string, weighting is required for each. The travel time for these routes should be calculated in consideration of the travel distance and cost value considering the travel weight, and the time to find and load the goods on each pallet, and define the final cost function value. Therefore, it will be used as a variable for calculating the optimization of the cost function value F(P(R)) corresponding to each order 900.

As shown in FIG. 5, when a large number of orders 900 and 901 come in at the same time, it shows that it is more efficient to process them collectively. That is, an independent P(G) path corresponding to this order called P(G) may be configured for the list 901 of another order displayed in green. However, as shown in the example of FIG. 5, all of the green palettes 802, 805, 806, and 807 responding to the order 901 of the additional customer can be rough even with the path of P(R), so the movement to the Z axis Only the additional movement amount according to is added to the cost function, which can be much more efficient than the two separate orders.

6 shows a method in which the transport robot 500 automatically responds to multiple orders at the same time in the present invention. As a practical example of implementing this, it is an optimal way to respond when 9 orders (900, 901, ..., 908) are received for 10 minutes. A container 950 is shown that displays the pallets containing the items in the list of nine orders in different colors and displays them in spatial XYZ locations. At this time, it is preferable to make a path of the transport robot so that the transport robot 500 can respond by bundling orders having similar paths rather than moving the number 9 by individually responding to all nine orders. This is a clustering engine 960 is designed that bundles orders having a similar path with a plurality of order item inventory data and a corresponding spatial location of a storage pallet. This is to design artificial intelligence engines through deep learning based on big data.

In this practical example, nine orders are grouped by three with similar paths, and the transport paths for groups of three orders are set to P1, P2, and P3, respectively. That is, P1 combines the red order 900, the blue order 902, and the violet order 906 to move the items in the list of orders to pallets where they are stored and pick up the items and put them in a cart. The spatial location within the containers of these pallets is conceptually illustrated (951). The other two bundle transport paths P2 and P3 are similarly constructed (953).

The function of minimizing the cost function F(P) for the bundle paths P1, P2, and P3 made in the present invention and satisfying the time T(P) it takes to perform below the target value is the core, deep data-based deep learning This is where algorithms are needed.

7 shows an optimization algorithm of the transport robot 500 to be implemented in the present invention. First, based on the spatial data 950 about the location of the pallet where a plurality of orders and the corresponding ordered goods are stored at the same time, a similar transport order P1, P2, P3 (961) is created and input to the input layer ( 971), and passes through a number of hidden layers 972 and 973 to make the output layers F(P) and T(P) 974, so that they learn to reach the desired values. Through this, the integrated cost function F(P) =F(P1)+F(P2)+F(P3) is minimized, and T(P)=T(P1)+T(P2)+T(P3) is fixed for a certain time. In order to adjust to the following, a solution for optimizing the position of the pallet of the corresponding product is obtained, and the position of the pallet for initial storage is adjusted. In addition, these solutions can be used to prevent out-of-stock and minimize inventory.

8 shows a function related to block-based P2P payment in the present invention. Functionally, in the case of an existing convenience store, if a customer buys multiple items and pays them to the convenience store with a credit card, the convenience store owner collects them, pays the fee to the credit card company, receives it, and pays the fee through the online payment to the goods suppliers and transfers it. do. Therefore, through this, it takes at least 2-3% of the fee and 1 to 2 weeks or more for the price to be paid from the customer to the supplier. In the present invention, payment is made immediately and within a maximum of 10 minutes through a P2P transaction 280 with the blockchain virtual currency 281.

This minimizes the transaction cost by allowing a small amount of low-cost goods to be settled immediately, providing the effect of purchasing goods directly from the market with cash. The virtual currency used here can be used as a dedicated currency based on the existing Bitcoin or Ethereum platform or as a dedicated virtual currency for this vending machine convenience store.

Claims (6)

In the vending machine type fully unmanned system provided in the container,
A control system 140 that receives online orders from customers through an operating service platform 400 on the cloud and controls the automated equipment 150 to pick the goods based on the online orders; And
Automated equipment 150;
Including,
The automation facility 150,
A plurality of shelves provided inside the container to store the article, the article being placed on a pallet;
An X-rail bar in the X-axis direction and a Y-rail bar in the Y-axis direction installed on the upper container;
Moving along the X-rail bar and Y-rail bar, and using the Z-string that can be stretched or reduced in the Z direction, moves to the pallet of the desired shelf through the passage between the shelves, and then picks and transports the goods. A transport robot 500;
Including,
The control system 140,
The spatial location of a plurality of pallets raised on a plurality of shelves in the container is coordinated with XYZ coordinates, and a path for picking all of a plurality of items ordered online using the coordinates is defined as P(R). , The P(R) includes the sum of the number of pallets that the transport robot goes through in each direction to pick up the plurality of items-the direction is the X-axis direction, the Y-axis direction, and the Z-axis direction. ,
In addition, the control system 140,
When a plurality of online orders are received for a predetermined time, similar movement paths among the transport robot movement paths corresponding to each online order on the coordinates are defined as one bundle transport path, and the transport robot is controlled to control the bundle transport. It characterized in that to pick up the goods while moving along the path,
The control system 140 defines a plurality of bundle transport paths for the received plurality of online orders, and picks items along the cost function F(P) and the bundle transport path for each bundle transport path. Define the time T(P),
The deep learning algorithm is performed to minimize the F(P) and make the T(P) less than or equal to a target value.
The bundled transport path includes: a clustering engine, wherein the clustering engine is an artificial intelligence engine using deep learning based on big data-to determine the spatial location of a plurality of order item list data and a storage pallet corresponding to the plurality of order item lists. Characterized by grouping online orders with similar movement paths on a basis,
Vending machine type fully unmanned system.
According to claim 1,
The transport robot,
A picker for picking items as having a robot arm shape;
A cart into which an item picked up by the picker is placed; And
RFID recognizer for recognizing the RFID chip attached to the article;
It includes,
The transport robot, after approaching a pallet of a desired shelf, recognizes an ordered item-an RFID chip is attached to the item-with the RFID recognizer, picks the item using the picker and puts it in a cart. And also
The transport robot, characterized in that for picking the goods while moving in the direction of the exit from which the goods located at the bottom of the container from the entrance to the goods located above the container is received,
Vending machine fully unmanned system.
According to claim 2,
The vending machine type fully unmanned system,
Characterized in that, by the operation service platform 400 on the cloud, it is selected as the location closest to the customer who placed the online order,
Vending machine fully unmanned system.
delete delete delete

Images