Storage location determination system and method thereof

Abstract

The invention relates to a storage location system and a method thereof, which enables the determination of the most effective storage locations by deciding the locations of the blocks (products) in order to ensure operational efficiency in warehouses, which is one of the components of supply chain management.

Description

[0001] STORAGE LOCATION DETERMINATION SYSTEM AND METHOD THEREOF

[0002] Technical Area of the Invention

[0003] The invention relates to a storage location system and a method thereof, which enables the determination of the most effective storage locations by deciding the locations of the blocks (products) in order to ensure operational efficiency in warehouses, which is one of the components of supply chain management.

[0004] State of the Art

[0005] Different equipment such as forklifts (pallet trucks), robots, drones (unmanned aerial vehicles) are used for the placement and transportation of products in warehouses. In this process, the transportation system is maintained by managing the equipment through a controller. At the planning stage, the capability of some equipment allows operating only with block locations below a certain height. For example, some equipment, such as forklifts, cannot operate in block locations above the height limit.

[0006] One of the most innovative approaches to solving the problem of block placement is the use of drones as a block transport strategy. For this reason, it is important to include drones in planning to increase operational efficiency, primarily based on height capability. In this case, real-time demands need to be monitored and supply chain equipment needs to be managed effectively.

[0007] Current studies presented in the literature have mainly focused on 2D models and ignored the 3D flight movement of drones. The energy restriction of drones is an important component that needs to be included in the process. In addition, the techniques in the literature do not have a universal perspective and ignore the resource constraints. Moreover, finding the best solution can only be accomplished by checking all possible scenarios for such complex problems, including drone path planning and block placement assignment. For this reason, the block placement solutions presented in the literature do not fully adopt modeling, real-time monitoring, bidirectional data and control flows in accordance with the constraints of industrial equipment. In the state of the art, there is no 3D data-centered supply chain management that includes cargo drones by maximizing the use of storage space. For this reason, it is important to develop an online genetic algorithm-based method in order to solve the problem of block placement in warehouse management, which is one of the critical components of supply chain management. The problem of block placement in storage management is solved by using Digital Supply Chain Twin, which provides real-time monitoring and management capabilities, and by offering a data-based warehouse digital twin architecture.

[0008] Brief Description and Objects of the Invention

[0009] The invention relates to a storage location system, which enables the determination of the most effective storage locations by deciding the locations of the blocks (products) in order to ensure operational efficiency in warehouses, which is one of the components of supply chain management.

[0010] One object of the invention is to enable the provision of digital twins corresponding to digital / virtual representations of assets in the physical warehouse environment so that virtual copies can be worked on.

[0011] Another object of the invention is to ensure the effective use of resources by incorporating the advantages of the use of drones in warehouse management.

[0012] Another object of the invention is to enable the determination of effective storage locations with less energy consumption in the virtual warehouse thanks to the bidirectional data communication between the physical warehouse and the virtual warehouse.

[0013] Another object of the invention is to enable cost-effective warehouse management by performing analyzes on virtual twins with the data collected from the physical warehouse.

[0014] Another object of the invention is to enable real-time, fast and effective detection of the storage location problem with minimum energy consumption. Another object of the invention is to use a digital supply chain twin that provides realtime monitoring and management capabilities, and to provide a data-based warehouse digital twin architecture.

[0015] Description of the Drawings

[0016] Fig 1. is the schematic view of the system of the present invention.

[0017] Fig 2. is the schematic view of the system of the present invention.

[0018] Fig 3. is the schematic view of the method of the present invention.

[0019] Description of the References in Drawings

[0020] 1 . Physical warehouse

[0021] 2. Warehouse elements

[0022] 3. Access point

[0023] 4. Logical flow table

[0024] 5. Lidar camera

[0025] 6. Cloud tier

[0026] 7. Controller

[0027] 8. Product demand unit

[0028] 9. Digital Twin

[0029] 10. Optimization Module

[0030] 11 . Product placement module

[0031] 12. Genetic algorithm module

[0032] 13. Performance evaluation module

[0033] 14. Optimal result module

[0034] 1001. Creating the digital / virtual twin of the warehouse elements (2) in the physical warehouse (1) by the controller (7) over the data obtained through the lidar camera (5) and transmitted to the cloud tier (6) via the access point (3)

[0035] 1002. Collecting the product demands to be placed / shipped in the physical warehouse through the product demand unit (8) 1003. Collecting the current status statistics of the warehouse elements (2) from the logical flow tables (4) of the access points (3) in the physical warehouse, performing the status update of the digital twins (9) by the controller (7) and transmitting the relevant data to the optimization module (10)

[0036] 1004. Creating a 3D model of the physical warehouses (1) in the product / block placement module (11) with the data from the digital twins (9) and calculating the energy consumption of each digital twin (9) and transmitting it to the genetic algorithm module (12)

[0037] 1005. Selecting the parents in the genetic algorithm module (12) to run the genetic algorithm on the product placement module (11 )

[0038] 1006. By using the parents selected in the genetic algorithm module (12), separating the mother and father chromosomes from two randomly determined points, obtaining the first and last part of the mother chromosome and the middle part of the father chromosome and producing new generations

[0039] 1007. Performing mutation at a randomly determined point on the new generations produced in the genetic algorithm module (12) running on the controller (7)

[0040] 1008. Updating the new generations if there are parts that do not comply with the optimization formula constraints and updating the energy metrics in the genetic algorithm module (12)

[0041] 1009. Executing the survival phase by mixing the new generations with the parent pool in the genetic algorithm module (12), keeping the best one in the pool according to the energy metrics and deleting the others from the pool (1009),

[0042] 1010. Running the genetic algorithm module (12) in a certain number of iterations, then sorting the results in the performance evaluation module (13) and determining the best result,

[0043] 1011. Sharing the best solution with the physical warehouse according to the energy metric of the genetic pool by the optimal result module (14), 1012. Transmitting the optimal solution produced in the cloud tier (6) to the warehouse elements (2) through access points (3) via the OpenFlow protocol and performing the product / block placement

[0044] Detailed Description of the Invention

[0045] The invention enables the determination of the most effective storage locations by deciding the locations of the blocks (products) in order to ensure operational efficiency in warehouses, which is one of the components of supply chain management. For this purpose, it transfers the equipment in the physical warehouse to the cloud via access points through cameras. A virtual twin of the physical store is created based on the data obtained. Current status statistics of the warehouse equipment are collected from the OpenFlow tables of the access points in the physical warehouse and the status of the virtual twins is updated periodically. Thus, instead of directly interfering with the physical warehouse by monitoring the resources and incoming demands in real time, 3D virtual twin (Digital Twin) creation technology is used by performing tests on the virtual copy of the digital warehouse. Effective use of resources is ensured by including the advantages of using drones in warehouse management. Thanks to the bidirectional data communication between the physical warehouse and the virtual warehouse, effective storage locations are determined with less energy consumption in the virtual warehouse.

[0046] The invention provides cost-effective warehouse management by performing analyzes on virtual twins with the data collected from the physical warehouse, and also prevents the resource problem by performing the analyzes in the cloud. Thus, block management in warehouses can be carried out autonomously in supply chain management.

[0047] The storage location determination system comprises a physical warehouse (1 ), warehouse elements (2), an access point (3), a logical flow table (4), a lidar camera (5), a cloud tier (6), a controller (7), a product demand unit (8), a digital twin (9), an optimization module (10), a product placement module (11 ), a genetic algorithm module (12), a performance evaluation module (13) and an optimal result module (14). A physical warehouse (1) is the physical environment where the products are stored and distributed as well as where supply chain activities such as inventory management are managed.

[0048] The warehouse elements (2) are assets / auxiliary elements such as shelves, forklifts (pallet trucks) and drones to manage the activities of the physical warehouse (1 ).

[0049] The access point (3) is a device which is wired with other network devices and provides wireless access to all warehouse elements in the warehouse. In the preferred embodiment of the invention, Wi-Fi wireless connection is used as the access point.

[0050] The logical flow table (4) collects status statistics within the access point (3) and periodically transmits these statistics to the digital twin (9) tier via the OpenFlow protocol.

[0051] The lidar camera (5) is a device which facilitates digital twin modeling by monitoring the warehouse elements (2) and ensures that it is updated by periodically transmitting the digital twin model data to the digital twin (9) tier.

[0052] The cloud tier (6) enables the acquisition and processing of lidar camera (5) data through the access point (3).

[0053] The controller (7) is located in the cloud tier (6) and is the center where all modules are implemented with software services. The controller (7) is the core of the softwarebased network and is a server. It is located between the network devices at one end of the network and the storage location determination system / service. Any communication between the storage location determination system / service and the network devices passes through the controller (7) and manages the flow control. The product demand unit (8), the digital twin (9) and the optimization module (10) operate within the controller (7). The controller (7) periodically retrieves the data statistics contained in the logical flow tables (4) of the access points (3) located in the physical topology to the cloud tier (6). Thus, the current status statistics of the warehouse elements (2) and the status update of their digital twins (9) are carried out periodically by the controller (7) and the relevant data is transmitted to the optimization module (10). The product demand unit (8) is the financial unit that manages customers' product demands. The product demands to be placed / shipped in / from the physical warehouse are collected through the product demand unit (8) which periodically manages customer cargo demands.

[0054] The digital twin (9) corresponds to the digital representations of the assets in the physical warehouse (1) environment. Periodic status updates are performed with the data received from the physical tier.

[0055] The optimization module (10) is the structure that performs the product placement method based on the genetic algorithm. In the optimization module (10), the product / block placement module (11 ), the genetic algorithm module (12), the performance evaluation module (13) and the optimal result module (14) modules operate, respectively.

[0056] The Product / Block placement module (11) is the structure in which the physical warehouses (1 ) are modeled in 3D and records the periodically calculated energy metric of the warehouse elements. Thus, it reveals a mathematical model of the most optimal locations where products can be placed. After the input data from the digital twins (9) are obtained, the product placement model is created with the data, the energy consumption of each digital twin (9) is calculated and transmitted to the genetic algorithm module (12).

[0057] The genetic algorithm module (12) is the module that provides an energy-efficient solution by performing parent selection, recombination, mutation and survivor selection, respectively, on the product / block placement module (11 ) via the genetic algorithm. Parent chromosome selections are performed on the product placement module (11). Mother and father chromosomes are separated from two randomly determined points by using the selected parents in the product placement module (11). The first and last part of the mother's chromosome and the middle part of the father's chromosome are obtained and new generations are produced (recombination). After this stage, the new generations produced are mutated at a randomly determined point. If there are parts of the new generations that do not comply with the optimization formula constraints, they are updated and the energy metrics are updated. Afterwards, all solutions are ranked according to energy metrics by mixing the new generations and the parent pool. The best ones stay in the pool, while the others are deleted from the pool. The best solution in the pool is sent to the performance evaluation module (13).

[0058] The performance evaluation module (13) mathematically calculates the product placement performance obtained by the optimization module (10) and evaluates the effectiveness of the model by taking into account the energy metric and certain constraints. If the energy metric does not fall below a certain threshold value in the performance evaluation module (13) or if a certain number of iterations is not reached, the new generation production continues by creating a new pool until the values in the performance evaluation module (13) are obtained by repeating the parent selection, new generation production, mutation and survival stages. In the performance evaluation module (13), if the energy metric falls below a certain threshold value or if a certain number of iterations is reached, the value is transmitted to the optimal result module (14). The number of iterations in the preferred embodiment of the invention is 1000.

[0059] The optimal result module (14) is the structure in which the optimal results are shared with the physical warehouse according to the notifications from the performance evaluation module (13). The optimal result module (14) shares the best solution according to the energy metric of the genetic pool with the physical warehouse (1) and the warehouse elements (2) over access points (3) via the OpenFlow protocol and optimal product / block placement is ensured in the physical warehouse. The optimal result module (14) is a software service. The OpenFlow protocol is an interface. Optimal results are transmitted from the controller (7) to the access points (3) and from the access points (3) to the warehouse elements (2). The warehouse elements (2) receive commands via the OpenFlow protocol.

[0060] The storage location determination method consists of the following process steps: (1001) Creating the digital / virtual twin of the warehouse elements (2) in the physical warehouse (1) by the controller (7) over the data obtained through the lidar camera (5) and transmitted to the cloud tier (6) via the access point (3),

[0061] (1002) Collecting the product demands to be placed / shipped in the physical warehouse through the product demand unit (8), (1003) Collecting the current status statistics of the warehouse elements (2) from the logical flow tables (4) of the access points (3) in the physical warehouse, performing the status update of the digital twins (9) by the controller (7) and transmitting the relevant data to the optimization module (10),

[0062] (1004) Creating a 3D model of the physical warehouses (1) in the product / block placement module (11) with the data from the digital twins (9) and calculating the energy consumption of each digital twin (9) and transmitting it to the genetic algorithm module (12),

[0063] (1005) Selecting the parents (parent selection) in the genetic algorithm module (12) to run the genetic algorithm on the product placement module (11 ),

[0064] (1006) By using the parents selected in the genetic algorithm module (12), separating the mother and father chromosomes from two randomly determined points, obtaining the first and last part of the mother chromosome and the middle part of the father chromosome and producing new generations,

[0065] (1007) Performing mutation at a randomly determined point on the new generations produced in the genetic algorithm module (12) running on the controller (7),

[0066] (1008) Updating the new generations if there are parts that do not comply with the optimization formula constraints and updating the energy metrics in the genetic algorithm module (12),

[0067] (1009) Executing the survival phase (natural selection) by mixing the new generations with the parent pool in the genetic algorithm module (12), keeping the best one in the pool according to the energy metrics and deleting the others from the pool,

[0068] (1010) Running the genetic algorithm module (12) in a certain number of iterations, then sorting the results in the performance evaluation module (13) and determining the best result,

[0069] (1011) Sharing the best solution with the physical warehouse according to the energy metric of the genetic pool by the optimal result module (14), (1012) Transmitting the optimal solution produced in the cloud tier (6) to the warehouse elements (2) through access points (3) via the OpenFlow protocol and performing the product / block placement.

[0070] In the process step (1010) of running the genetic algorithm module (12) in a certain number of iterations, then sorting the results in the performance evaluation module (13) and determining the best result, a certain number of iterations is 1000.

[0071] In the preferred embodiment of the invention, firstly, digital / virtual twin modeling of all warehouse elements (2) (drone, block, forklift, etc.) in the physical warehouse (1 ) is performed through a lidar camera (5). This data is transmitted to the controller (7) in the cloud tier (6) via the access points (3) that provide communication in the warehouse and a digital / virtual twin (9) of the warehouse is created. On the other hand, product demands to be placed / shipped in the physical warehouse are collected through the product demand unit (8) which periodically manages customer cargo demands. Data statistics are kept in the logical flow tables (4) of the access points (3) in the physical topology. These statistics are periodically retrieved into the cloud tier (6) by the controller (7). Thus, the current status statistics of the warehouse elements (2) and the status update of their digital twins (9) are performed periodically and the relevant data is transmitted to the optimization module (10).

[0072] In the optimization module (10), the product / block placement module (11 ), the genetic algorithm module (12), the performance evaluation module (13) and the optimal result modules (14) operate, respectively. After the input data from the digital twins (9) are obtained, the product / block placement module (11 ) is created with the data, the energy consumption of each digital twin (9) is calculated and transmitted to the genetic algorithm module (12). Parent chromosome selections (parent selection) are performed on the product placement module (11). By using the parents selected in the genetic algorithm module (12), the mother and father chromosomes are separated from two randomly determined points. The first and last part of the mother's chromosome and the middle part of the father's chromosome are obtained and new generations are produced (recombination). After this stage, the new generations produced are mutated at a randomly determined point. If there are parts of the new generations that do not comply with the optimization formula constraints, they are updated and the energy metrics are updated. Afterwards, all solutions are ranked according to energy metrics by mixing the new generations and the parent pool. The best ones stay in the pool, while the others are deleted from the pool. The best solution in the pool is sent to the performance evaluation module (13). If the energy metric does not fall below a certain threshold value in the performance evaluation module (13) or if a certain number of iterations is not reached, the new generation production continues by creating a new pool until the values in the performance evaluation module (13) are obtained by repeating the parent selection, new generation production, mutation and survival stages. In the performance evaluation module (13), if the energy metric falls below a certain threshold value or if a certain number of iterations is reached, the value is transmitted to the optimal result module (14). In this module, the best solution according to the energy metric of the genetic pool is shared with the physical warehouse and the warehouse elements (2) over access points (3) via the OpenFlow protocol and optimal product / block placement is ensured in the physical warehouse.

[0073] The Digital Warehouse Twin environment for the Block Placement Problem is defined as follows, considering the 3D coordinate plane expressed by (x,y,z) and its coordinates:

[0074] Equation-1 represents the set of blocks. Each block of is expressed with and is located at x, y, z in the given 3D coordinates.

[0075] Equation-2 shows the 3D distance from the starting position in block to the order position for customer’s order. It is also defined as the flight distance of the drone that stores or receives the order.

[0076] If the Drone flight speed expressed as the flight time from the starting position to the destination position is calculated based on the

[0077] Euclidean distance as follows:

[0078] Equation-3

[0079] Accordingly, the energy consumption of drones for transportation purposes is calculated as follows: and respectively represent the hover power of the drone, the power required when the drone is moving at full speed, and the power consumed by the system while the drone is charging. indicates the recharge time of the drone. The hover power is calculated as follows:

[0080] Equation-5 drone mass refers to gravity represents the area of the rotating disk of a rotor, indicates the number of propellers of the drone, and P indicates the density of air.

[0081] Similarly, the transition power is expressed as:

[0082] Equation-6 and represents the power level when the drone is moving at full speed and when stopped at a stationary position, respectively. is the maximum speed of the drone, and is the speed of the drone that transmits the customer's order from the block to the zero position.

[0083] Assigning means that the cargo of the customer is located in the block

[0084] Accordingly, the optimization formula that defines the Block Placement Problem is expressed as follows:

[0085] Equation-10 is the number of blocks in a warehouse, and is the number of customer orders that need to be delivered. Each block has a capacity and it is expressed by

[0086] Each customer is indicated with , and each order demand is indicated with . Each drone has a travel cost denoted by for client-based demands from the block According to the first constraint defined in the optimization problem given above, the total demand cannot exceed the capacity of the block. In addition, according to the second constraint, each customer order must be served by exactly one block in a warehouse. Finally, according to the third constraint, each drone can fly for a maximum of 30 minutes due to battery capacity.

Claims

CLAIMS1 . A storage location determination system that allows the storage locations to be determined by deciding the locations of the products, characterized in that it comprises:- at least one physical warehouse (1) where the products are stored and distributed,- at least one warehouse element (2) which manages the activities of the physical warehouse (1 ),- at least one access point (3) that provides wireless access to the warehouse elements (2) in the physical warehouse (1),- at least one logical flow table (4) that collects status statistics within the access point (3) and periodically transmits these statistics to the digital twin (9) tier via the OpenFlow protocol,- at least one lidar camera (5) that monitors the warehouse elements (2) and periodically transmits the digital twin model data to the digital twin (9) tier, allowing it to be updated,- at least one cloud tier (6) that enables the reception and processing of the lidar camera (5) data via access point (3),- at least one controller (7), which is a server that periodically updates the current status statistics of the warehouse elements (2) and the status of the digital twins (9) by periodically retrieving the data statistics contained in the logical flow tables (4) of the access points (3) located in the physical topology to the cloud tier (6) and transmits this data to the optimization module (10),- at least one product demand unit (8) that collects product demands to be placed / shipped in / from the physical warehouse (1 ),- at least one digital twin (9) corresponding to digital representations of assets in the physical warehouse (1) environment,- the optimization module (10) which operates the product placement module (11), the genetic algorithm module (12), the performance evaluation module (13) and the optimal result module (14),- at least one product placement module (11) that models the physical warehouses (1) in three dimensions, calculates the energy consumption of each digital twin (9) by creating a product placement model with thedata after obtaining the input data from the digital twins (9), records the periodically calculated energy metric of the warehouse elements (2) and transmits it to the genetic algorithm module (12),- at least one genetic algorithm module (12) which performs parent chromosome selections on the product placement module (11 ), separates the mother and father chromosomes from two randomly determined points using the parents selected in the product placement module (11), obtains the first and last part of the mother chromosome and the middle part of the father chromosome and produces new generations, performs mutations at a randomly determined point on the new generations produced, updates the parts of the new generations that do not comply with the optimization formula constraints, and updates the energy metrics, ranks all solutions according to energy metrics by mixing new generations and the parent pool, and sends the best solution found to the solution performance evaluation module (13),- at least one performance evaluation module (13) which mathematically calculates the product placement performance obtained by the optimization module (10), evaluates the effectiveness of the model by taking into consideration the energy metric and constraints, and transmits the value to the optimal result module (14) if the energy metric falls below a certain threshold value or a certain number of iterations is reached,- at least one optimal result module (14) which ensures optimal product placement in the physical warehouse (1) by sharing the best solution from the performance evaluation module (13) with the physical warehouse (1) and the warehouse elements (2) over access points (3) via the Open Flow protocol,2. A method of determining the storage location, characterized by the process steps of:- (1001 ) Creating the digital / virtual twin of the warehouse elements (2) in the physical warehouse (1) by the controller (7) over the data obtained through the lidar camera (5) and transmitted to the cloud tier (6) via the access point (3),- (1002) Collecting the product demands to be placed / shipped in the physical warehouse through the product demand unit (8),- (1003) Collecting the current status statistics of the warehouse elements (2) from the logical flow tables (4) of the access points (3) in the physical warehouse, performing the status update of the digital twins (9) by the controller (7) and transmitting the relevant data to the optimization module (10),- (1004) Creating a 3D model of the physical warehouses (1) in the product / block placement module (11 ) with the data from the digital twins (9) and calculating the energy consumption of each digital twin (9) and transmitting it to the genetic algorithm module (12),- (1005) Selecting the parents in the genetic algorithm module (12) to run the genetic algorithm on the product placement module (11 ),- (1006) By using the parents selected in the genetic algorithm module (12), separating the mother and father chromosomes from two randomly determined points, obtaining the first and last part of the mother chromosome and the middle part of the father chromosome and producing new generations,(1007) Performing mutation at a randomly determined point on the new generations produced in the genetic algorithm module (12) running on the controller (7),- (1008) Updating the new generations if there are parts that do not comply with the optimization formula constraints and updating the energy metrics in the genetic algorithm module (12),Executing the survival phase by mixing the new generations with the parent pool in the genetic algorithm module (12), keeping the best one in the pool according to the energy metrics and deleting the others from the pool (1009),(1010) Running the genetic algorithm module (12) in a certain number of iterations, then sorting the results in the performance evaluation module (13) and determining the best result,- (1011 ) Sharing the best solution with the physical warehouse according to the energy metric of the genetic pool by the optimal result module (14),(1012) Transmitting the optimal solution produced in the cloud tier (6) to the warehouse elements (2) through access points (3) via the OpenFlow protocol and performing the product / block placement.

3. A storage location determination system according to Claim 1 , characterized by comprising a warehouse element (2) which can be a shelf and / or a forklift and / or a drone.

4. A storage location determination system according to Claim 2, characterized in that the number of iterations is 1000 in the process step (1010) of running the genetic algorithm module (12) in a certain number of iterations, then sorting the results in the performance evaluation module (13) and determining the best result,5. A storage location determination system according to Claim 2, characterized in that in the process step (1004) of creating a 3D model of the physical warehouses (1) in the product placement module (11) with the data from the digital twins (9), calculating the energy consumption of each digital twin (9) and transmitting it to the genetic algorithm module (12), the energy consumption is calculated with the Equation-4:

6. A storage location determination system according to Claim 2, characterized in that in the process step (1008) of updating the new generations if there are parts that do not comply with the optimization formula constraints and updating the energy metrics in the genetic algorithm module (12), the optimization formula is calculated with the Equation-7:

7. A storage location determination system according to Claim 2, characterized in that in the process step (1008) of updating the new generations if there are parts that do not comply with the optimization formula constraints and updating the energy metrics in the genetic algorithm module (12), the optimization formula constraints are calculated with Equation-8