An automatic product selection and replenishment method for an intelligent cabin, an intelligent cabin, and a body robot

Abstract

The application provides an automatic product selection and replenishment method of an intelligent cabin, an intelligent cabin and a robot with a body, the system comprising an intelligent cabin and a plurality of robots with bodies, the automatic product selection and replenishment method comprising: acquiring inventory data of each shelf, sales data of each sales window and a current working mode of the intelligent cabin; determining a corresponding product selection strategy according to the working mode, and generating a replenishment plan according to the inventory data, the sales data and the product selection strategy, the replenishment plan containing a plurality of replenishment tasks; determining the priority of each replenishment task according to the working mode, and scheduling the corresponding robot with a body to execute the replenishment tasks in turn according to the replenishment plan and the priority of the replenishment tasks; in the process of executing the replenishment tasks by the robot with a body, controlling a display device to display replenishment progress information corresponding to the replenishment tasks, and updating the inventory data after each replenishment task is completed, so that the system collaboration and inventory turnover efficiency in the unmanned retail scene can be improved.

Description

Technical Field

[0001] This disclosure relates to the fields of embodied smart retail and spatial smart technology, and in particular to an automatic product selection and replenishment method for a smart cabin, a smart cabin, and an embodied robot. Background Technology

[0002] With the integration of embodied intelligent technology and offline commercial spaces, embodied robots are increasingly being used in retail picking, interactive displays, and traffic-driving performances. Meanwhile, intelligent retail pods (or simply intelligent pods) are gradually becoming important carriers for unmanned retail and commercial displays in shopping districts, cultural and tourism scenic spots, and other similar settings due to their ease of deployment and high mobility. In actual operation, intelligent retail pods are typically equipped with display devices for product display. Achieving consistent linkage between displayed content and inventory status, as well as the coordinated adaptation of replenishment operations and operational models, are key issues for ensuring the continuous and stable operation of unmanned retail.

[0003] However, in related technologies, the display system and inventory management system of existing smart retail kiosks are independent of each other, often requiring manual replenishment. As a result, it is difficult to combine real-time sales data and product richness constraints to dynamically optimize product selection, leading to both out-of-stock hot-selling products and stockpiled slow-moving products. In addition, the display content on the display device is relatively independent, ultimately resulting in a break in the information link between display and inventory, and a conflict between replenishment and operation mode scheduling, which reduces the system synergy and inventory turnover efficiency in unmanned retail scenarios. Summary of the Invention

[0004] The main objective of this disclosure is to propose an automatic product selection and replenishment method for an intelligent cabin, an intelligent cabin, and an embodied robot, which can improve system coordination and inventory turnover efficiency in unmanned retail scenarios.

[0005] To achieve the above objectives, a first aspect of this disclosure proposes an automatic product selection and replenishment method for an intelligent cabin, applied to an automatic product selection and replenishment system. The automatic product selection and replenishment system includes an intelligent cabin and multiple embodied robots. The intelligent cabin includes a cabin body that encloses an internal space. Several sales windows are provided on the walls of the cabin body, and a display device is provided on the walls of the cabin body. Several shelves are also provided inside the cabin body, and the shelves correspond to the positions of the sales windows. The automatic product selection and replenishment method includes: Obtain inventory data for each shelf, sales data for each sales window, and the current operating mode of the smart cabin; The corresponding product selection strategy is determined according to the working mode, and a replenishment plan is generated based on the inventory data, the sales data and the product selection strategy. The replenishment plan contains several replenishment tasks. The priority of each replenishment task is determined according to the working mode, and the corresponding embodied robot is scheduled to execute the replenishment task in sequence according to the replenishment plan and the priority of the replenishment task. During the process of the robot performing the replenishment task, the display device is controlled to display the replenishment progress information corresponding to the replenishment task, and the inventory data is updated after each replenishment task is completed.

[0006] In some embodiments, determining the corresponding product selection strategy based on the working mode and generating a replenishment plan based on the inventory data, the sales data, and the product selection strategy includes: The selection weight of each product category is determined based on the aforementioned working mode; The replenishment plan is generated by optimizing the solution based on the inventory data, sales data, and product selection weights, with the objective function being to maximize the product weighted gross profit multiplied by the demand forecast, and the constraints being that the product category richness is not lower than the preset richness lower bound and the inventory of each shelf does not exceed the corresponding arc capacity.

[0007] In some embodiments, the working mode includes at least one of a performance mode, a hybrid mode, and a retail mode; The step of determining the selection weight of each product category according to the working mode includes: When the working mode is the performance mode or the hybrid mode, increase the selection weight of traffic-driving product categories; When the working mode is the retail mode, increase the selection weight of high-turnover product categories.

[0008] In some embodiments, determining the priority of each replenishment task according to the working mode, and scheduling the corresponding embodied robot to execute the replenishment task sequentially according to the replenishment plan and the priority of the replenishment task, includes: Select at least one of the multiple locations within the cabin space as the target location, wherein the target location is the cabin location for reserving space for manual replenishment operations; In the remaining multiple cabin locations other than the target location, the priority of each replenishment task is determined according to the working mode, and the corresponding embodied robot is scheduled to execute the replenishment task in sequence according to the replenishment plan and the priority of the replenishment task.

[0009] In some embodiments, at least one of the multiple sales windows can be selected as a reserved sales window, wherein the reserved sales window is a sales window reserved for manual sales operations. When any customer arrives at the reserved sales window, the display device displays a manual service prompt in the corresponding display area, so that the sales staff can provide manual sales service to the customer according to the manual service prompt. After the manual sales service is completed, obtain the manual sales confirmation information and update the inventory data of the corresponding product based on the manual sales confirmation information.

[0010] In some embodiments, the cabin is provided with multiple arc segments, each shelf is respectively disposed in the corresponding arc segment, and each embodied robot is respectively disposed corresponding to each arc segment; The step of determining the priority of each replenishment task according to the working mode, and scheduling the corresponding embodied robot to execute the replenishment task sequentially according to the replenishment plan and the priority of the replenishment task, includes: When the working mode is retail mode, the priority of the replenishment task is determined to be lower than the priority of the order picking task, and the embodied robot corresponding to the arc segment where the target shelf is located is scheduled to perform the replenishment task. When the working mode is performance mode or hybrid mode, the priority of the replenishment task is determined to be lower than that of the performance task, and the embodied robot executing the replenishment task is controlled to perform the replenishment task across arc segments.

[0011] In some embodiments, the automatic product selection and replenishment method further includes: During the execution of the replenishment task, the order service level agreement status of each sales window is monitored in real time; When the Service Level Agreement (SLA) of any of the sales windows is triggered, the currently executing replenishment task is paused or canceled, and the avatar robot executing the replenishment task is switched to execute the order picking task of the corresponding sales window.

[0012] In some embodiments, the automatic product selection and replenishment method further includes: Establish a binding relationship between the products displayed on the display device and the shelves and the sales window. The binding relationship includes at least the product identifier, the corresponding sales window identifier, the corresponding shelf identifier, and the corresponding available inventory quantity. Before the display device displays the highlight content of the target product, the available inventory quantity of the target product is verified to be greater than zero according to the binding relationship; When the available inventory quantity is greater than zero, the highlight display content is played through the display device; when the available inventory quantity is zero, the display device is controlled to cancel the display of the highlight display content and switch to the display of alternative products with a available inventory quantity greater than zero under the same sales window.

[0013] In some embodiments, during the process of the android performing the replenishment task, controlling the display device to display replenishment progress information corresponding to the replenishment task includes: Determine the target arc segment where the target shelf for which the embodied robot will perform the replenishment task; In the display area corresponding to the target arc segment on the display device, a replenishment progress overlay is rendered, and the replenishment progress overlay includes at least a replenishment status indicator and product information that is being replenished; After the replenishment task is completed, the replenishment progress overlay is cleared.

[0014] In some embodiments, an outdoor customer area is provided around the perimeter of the cabin; The automatic product selection and replenishment method also includes: Collect interaction intent information of visitors in the outdoor customer area, including voice interaction intent and / or gesture interaction intent; Based on the visitor's circumferential position within the outdoor customer area, determine the target sales window that is closest to the visitor from each of the sales windows; The corresponding interactive response result is generated based on the interactive intent information, and the interactive response result is routed to the product interface of the target sales window.

[0015] In some embodiments, the automatic product selection and replenishment method further includes: Obtain the display skin switching command of the smart cabin, the display skin switching command being triggered by the switching of the working mode; According to the display skin switching command, the display device is controlled to switch from the current display skin to the target display skin, and the switching process adopts a gradual transition method; The display skin includes a performance-type display skin corresponding to the performance mode and a retail-type display skin corresponding to the retail mode.

[0016] To achieve the above objectives, a second aspect of this disclosure provides a smart cabin, comprising: The cabin is enclosed to form an interior space. Several sales windows are opened on the walls of the cabin, and display devices are installed on the walls of the cabin. Several shelves are installed inside the cabin, and the shelves are positioned corresponding to the sales window; The control unit is used to acquire inventory data of each shelf, sales data of each sales window, and the current working mode of the smart cabin; determine the corresponding product selection strategy according to the working mode, and generate a replenishment plan based on the inventory data, sales data, and product selection strategy, wherein the replenishment plan contains several replenishment tasks; determine the priority of each replenishment task according to the working mode, and schedule the corresponding embodied robot to execute the replenishment task sequentially according to the replenishment plan and the priority of the replenishment task; during the execution of the replenishment task by the embodied robot, control the display device to display the replenishment progress information corresponding to the replenishment task, and update the inventory data after each replenishment task is completed.

[0017] To achieve the above objectives, a third aspect of the present disclosure provides an automated product selection and replenishment system for a smart cabin, comprising a smart cabin and multiple embodied robots; The smart cabin includes: a cabin body that encloses an interior space, a number of sales windows on the walls of the cabin body, a display device on the walls of the cabin body, a number of shelves inside the cabin body that correspond to the positions of the sales windows, and a control unit. The control unit is used to acquire inventory data of each shelf, sales data of each sales window, and the current working mode of the smart cabin; determine the corresponding product selection strategy according to the working mode, and generate a replenishment plan based on the inventory data, sales data, and product selection strategy, wherein the replenishment plan contains several replenishment tasks; determine the priority of each replenishment task according to the working mode, and schedule the corresponding embodied robot to execute the replenishment task sequentially according to the replenishment plan and the priority of the replenishment task; during the execution of the replenishment task by the embodied robot, control the display device to display the replenishment progress information corresponding to the replenishment task, and update the inventory data after each replenishment task is completed.

[0018] To achieve the above objectives, a fourth aspect of the present disclosure provides an electronic device, the electronic device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the automatic product selection and replenishment method for the smart cabin described in the first aspect embodiment.

[0019] To achieve the above objectives, a fifth aspect of the present disclosure provides a storage medium, which is a computer-readable storage medium storing a computer program that, when executed by a processor, implements the automatic product selection and replenishment method for the smart cabin described in the first aspect embodiment.

[0020] The beneficial effects of the embodiments disclosed herein include: By synchronously acquiring inventory data from each shelf, sales data from each sales window, and the current operating mode of the smart warehouse, the three previously separate data categories of inventory, sales, and operational status are incorporated into a unified management system. Then, based on the current operating mode, corresponding product selection strategies are matched, and replenishment plans with multiple replenishment tasks are dynamically generated by combining real-time inventory and sales data. This replaces the experience-based decision-making of traditional manual replenishment, and can dynamically optimize the product selection structure based on real-time operational data to match market demand, effectively alleviating the contradiction between stockouts of best-selling products and stockpiling of slow-moving products, directly improving inventory turnover efficiency from the product selection level. Subsequently, based on the operating mode, each... The system prioritizes replenishment tasks and schedules embodied robots to perform them sequentially. This adapts the pace and resource usage of replenishment operations to the current operating model, avoiding scheduling conflicts between replenishment operations and daily operations, and improving the overall operational synergy of the system. Finally, during the replenishment process, the system simultaneously controls the display device to show the corresponding replenishment progress information and automatically updates the inventory data after replenishment is completed. This achieves real-time linkage between the displayed content and the inventory status, ensuring the real-time accuracy of the inventory data and ultimately comprehensively improving the system synergy and inventory turnover efficiency in unmanned retail scenarios. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the automatic product selection and replenishment system for the smart cabin provided in this embodiment of the disclosure; Figure 2 This is a schematic diagram of the spatial division of the smart cabin provided in an embodiment of this disclosure; Figure 3 This is a schematic diagram of the planar area division of the smart cabin provided in an embodiment of this disclosure; Figure 4 This is a schematic diagram of the multi-layer partitioning of the smart cabin provided in the embodiments of this disclosure.

[0022] Figure 5 This is a flowchart illustrating the automatic product selection and replenishment method for the smart cabin provided in this embodiment of the disclosure; Figure 6 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this disclosure. Detailed Implementation

[0023] The accompanying drawings in the embodiments clearly and completely describe the technical solutions in the embodiments of this disclosure. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0024] It is understood that in the specific embodiments of this disclosure, which involve retrieving the working mode of the intelligent cabin, the operation data of the embodied robot, related control commands, sensor data and related data, when the above embodiments of this disclosure are applied to specific products or technologies, permission or consent can be obtained from the subject, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards.

[0025] Furthermore, when the embodiments of this disclosure require access to the working mode of the intelligent cabin, the operational data of the avatar robot, related control commands, sensor data, and related data, separate permission or consent to the working mode of the intelligent cabin, the operational data of the avatar robot, related control commands, sensor data, and related data can be obtained through pop-up windows or redirection to a confirmation page. After clearly obtaining separate permission or consent to the working mode of the intelligent cabin, the operational data of the avatar robot, related control commands, sensor data, and related data, the necessary working mode of the intelligent cabin, the operational data of the avatar robot, related control commands, sensor data, and related data for the normal operation of the embodiments of this disclosure are then obtained.

[0026] In this disclosure, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0027] Please see Figures 1 to 4 , Figure 1 This is a schematic diagram of the automatic product selection and replenishment system for the smart cabin provided in this embodiment of the disclosure. Figure 2 This is a schematic diagram of the spatial division of the smart cabin provided in an embodiment of this disclosure. Figure 3 This is a schematic diagram of the planar area division of the intelligent cabin provided in an embodiment of this disclosure. Figure 4 This is a schematic diagram of the multi-layer partitioning of the smart cabin provided in the embodiments of this disclosure.

[0028] The automated product selection and replenishment system includes a smart cabin 100 and multiple embodied robots 200. The smart cabin 100 includes a cabin body 110, which encloses an interior space 120. Several sales windows 130 and at least one entrance / exit 140 for the embodied robots 200 are located on the walls of the cabin body 110. Display devices are installed on the walls of the cabin body 110, including a transparent grid screen 151 on the walls and / or a circular display screen 152 on the top of the cabin body 110. Several shelves 160 are located within the interior space 120, corresponding to the sales windows 130. An external customer area 173 is located around the cabin body 110, where customers shop and watch performances without needing to enter the interior space 120. An apron performance area 170 (R-apron) is also provided on the outer wall of the cabin 110. The apron performance area 170 is located between the cabin 110 and the external customer area 173, and is used by the embodied robot 200 to perform performances outside the cabin. Furthermore, the external customer area 173 can also be equipped with multiple sensing sensors, including a top-down vision sensor, a surround-view vision sensor, a window queue depth acquisition sensor, and a POS / order event stream acquisition module, to collect real-time pedestrian flow perception data and transaction data in the external customer area 173. The intelligent cabin 100 also includes a control unit (not shown in the figure), which is communicatively connected to each module or device to execute the automatic product selection and replenishment method in this embodiment.

[0029] It should be noted that in the field of embodied intelligent retail and commercial display technology, the display system and inventory management system of existing intelligent retail pods are independent of each other, often requiring manual replenishment. Therefore, replenishment plans are difficult to dynamically optimize by combining real-time sales data and product richness constraints, resulting in both out-of-stock best-selling items and stockpiled slow-moving goods. Furthermore, the display content on the display devices is relatively independent, ultimately leading to a break in the information link between display and inventory, and conflicts between replenishment and operational scheduling, reducing system synergy and inventory turnover efficiency in unmanned retail scenarios. Meanwhile, digital humans, display devices, and physical embodied robots belong to different systems, with separate content production and asynchronous virtual-real elements; parallax and occlusion problems exist between display devices and physical embodied robots, lacking systematic solutions; and the performance-driven product promotion and inventory strategies are not linked, easily resulting in situations where products are displayed but not in stock. This disclosure addresses the above-mentioned technical deficiencies by providing an automatic product selection and replenishment method for intelligent pods, an intelligent pod, and an embodied robot, which can comprehensively improve system synergy and inventory turnover efficiency in unmanned retail scenarios. A detailed description follows.

[0030] Furthermore, the cabin body 110 of the intelligent cabin 100 can be constructed from a glass curtain wall; therefore, both the wall and the cabin body 110 can be referred to as the curtain wall layer, which is a transparent structure. The display device may include a transparent grid screen 151 disposed on the wall of the cabin body 110 and / or a ring-shaped display screen 152 disposed on the top of the cabin body 110. The transparent grid screen 151 is used to dynamically display product information, live operation or performance of the embodied robot, temperature information, and other content on the wall. In this embodiment, the transparent grille screen 151 can be installed behind the glass curtain wall, that is, behind the curtain wall layer. Customers can observe the display content of the transparent grille screen 151 through the curtain wall layer on the surface of the cabin 110. The area of ​​the transparent grille screen 151 can be set according to actual needs. For example, the transparent grille screen 151 can be installed in most or all areas behind the curtain wall layer, or the transparent grille screen 151 can be installed in a part of the area visible to customers. This embodiment does not impose specific limitations on this. The circular display screen 152 is used to present immersive visual information such as digital human performance content on the top of the cabin 110, and can also display temperature information. The two together constitute the information display system of the cabin wall.

[0031] Furthermore, an apron performance area 170 (R-apron) is set up around the cabin 110 in the external space. The apron performance area 170 is divided into multiple apron sectors 171. An angle mapping relationship is established between each entrance 140 and an apron sector 171, so that after the embodied robot 200 moves out of the cabin space 120 through the entrance 140, it can directly enter the corresponding apron sector 171 to perform the outboard performance operation.

[0032] It should be noted that the embodied robot 200 can be any embodied intelligent robot, which can be a humanoid robot with capabilities such as autonomous movement, item picking, human-computer interaction, and performance actions. In addition, this embodiment can also set up various different types of embodied robots 200 to meet different needs, such as meeting the special needs of retail or performance; this embodiment does not impose specific limitations on this.

[0033] Furthermore, the embodied robot 200 may be equipped with a receiving module and a work control module. The receiving module is used to receive data from the smart cabin 100 in the system, such as receiving replenishment tasks or picking tasks sent by the smart cabin 100. Then, the work control module is used to perform corresponding operations based on the replenishment tasks or picking tasks. This disclosure embodiment does not impose specific limitations on this.

[0034] In some embodiments of this disclosure, the intelligent cabin 100 further includes a spatial coding module (not shown in the figure). This module is used to establish and maintain a concentric annular spatial model of the cabin space 120. This concentric annular spatial model includes at least: a hub layer L0 located at the geometric center of the cabin 110; an arc-shaped working layer L1 surrounding the hub layer L0; a window ring layer L2 surrounding the arc-shaped working layer L1; and a curtain wall layer L3 located on the walls of the cabin 110. These four layers are arranged sequentially from the inside out, forming a multi-layered spatial topology model. The outer skirt area R-apron can also serve as a fifth layer. Wherein: The hub layer L0 is located at the geometric center of the cabin 110 and is the core traffic area of ​​the cabin space 120. It is mainly used for the embodied robot 200 to perform radial traversal and turning maneuvers within the cabin space 120. The arc segment operation layer L1 surrounds the hub layer L0 and is divided into multiple arc segments in the circumferential direction. For example, the arc segment operation layer L1 in the figure can be divided into 8 arc segments at equal angles, with each arc segment corresponding to an angle range. The window ring layer L2 is located on the periphery of the arc segment operation layer L1, and sales windows 130 are opened in the circumferential direction corresponding to the positions of each arc segment. The curtain wall layer L3 is the wall structure of the cabin 110 and is made of transparent glass curtain wall, such as LOW-E glass. Transparent grid screens 151 are set on the curtain wall layer L3 in areas other than where sales windows 130 and entrances / exits 140 are located.

[0035] Furthermore, a one-to-one four-element binding relationship is established between each arc segment and an embodied robot 200, a sales window 130, and a shelf 160. Specifically, for the i-th arc segment ARC... i Its binding relationships include: embodied robot R i Window W i Shelf area S i and the angle interval θ i Among them, the embodied robot R i The embodied robot 200 corresponding to this arc segment is responsible for performing tasks such as picking, replenishing, and performing in-cabin shows within this arc segment; Window W i Sales window 130 corresponds to this arc segment, through which customers place orders and pick up their goods; shelf area S i Shelf 160 corresponds to this arc segment and stores the corresponding product inventory; angle range θ i This refers to the angular range of the arc segment in the circumferential direction. This quaternary binding relationship allows any order from a sales window 130 to be uniquely mapped to the corresponding embodied robot 200 and shelf 160, realizing a precise matching link from order, window, arc segment, embodied robot to shelf. This fundamentally solves the technical problems of response delay and high error rate caused by the mismatch between window and operation in related technologies.

[0036] It should be noted that, through the five-layer structure division of the concentric circular annular space model and the establishment of the four-element binding relationship, the cabin space is transformed from a traditional single open space into a structured space with clear functional zoning and binding relationships. This provides a clear spatial topology basis for the collaborative operation of multiple androids and effectively avoids the problems of multi-robot path conflict and resource contention.

[0037] The intelligent cabin 100 in this embodiment adopts a dual-channel operation architecture, including a physical sales channel (channel A) and a visual information channel (channel B). Channel A corresponds to the sales window 130 of the window ring layer L2, which is used to realize the physical delivery of goods from the shelf 160 to the customer. When a customer places an order at a certain sales window 130, the corresponding embodied robot 200 and shelf 160 are determined according to the four-element binding relationship. The embodied robot 200 completes the entire process from picking, packing to delivery within the corresponding arc segment. The customer picks up the goods at the sales window 130. The customer does not need to enter the cabin space 120 during the entire process, realizing a fully automated and safe retail mode without human intervention.

[0038] Channel B corresponds to the display devices on the curtain wall layer L3, including a transparent grid screen 151 and a top circular display screen 152. The transparent grid screen 151 dynamically displays product information, live footage of the android's operations, or live performances on the wall, allowing customers outside the cabin to see the android 200's picking process in real time, enhancing transparency and trust. The top circular display screen 152 presents digital human performance content, brand promotional content, or immersive visual content on the top of the cabin, providing customers outside with a long-narrative visual experience. The display content in Channel B is synchronized with the operational tasks in Channel A. For example, when the android 200 in a certain arc is performing picking operations, the corresponding transparent grid screen 151 simultaneously plays the picking footage; when the system enters performance mode, the transparent grid screen 151 plays the performance footage, and the circular display screen 152 plays the digital human performance content. This dual-channel architecture, which separates physical sales from visual displays while maintaining synergy, integrates transparent displays with pickup windows. Customers can see the actual condition of the goods inside the cabin and the operation process of the embodied robot through the transparent grid screen 151, achieving an immersive shopping experience where what you see is what you buy.

[0039] It should be noted that the embodiments disclosed herein use a dual-channel operating architecture to functionally separate the physical sales channel and the visual information channel to avoid mutual interference, while coordinating the content to enhance the consistency of the experience. This solves the technical problems of separating the viewing experience from the sales experience and separating the transparent display from the pickup window in related technologies, thereby improving customer experience and purchase conversion rate.

[0040] At least one entrance / exit 140 is provided on the curtain wall layer L3 of the cabin 110. The entrance / exit 140 can be implemented by an electric airtight door or other doors, which are normally locked and only open when the embodied robot 200 needs to enter or exit the cabin. An apron performance area 170 (R-apron) is arranged around the cabin 110 in the external space. The apron performance area 170 is divided into multiple apron sectors 171, each corresponding to a certain angle range. Furthermore, a human-machine isolation zone 172 can be provided, and beyond the human-machine isolation zone 172 is the external customer area 173 where users can stay and watch. In addition, each entrance / exit 140 can also establish an angle mapping relationship with an apron sector 171. The opening angle of the entrance / exit 140 on the curtain wall layer L3 is consistent with or adjacent to the angle range of the corresponding apron sector 171, ensuring that the embodied robot 200 directly enters the corresponding apron sector 171 after exiting from the entrance / exit 140.

[0041] Furthermore, the shape of the cabin 110 can be set according to actual needs. For example, it can be a sphere, an ellipsoid, or an ellipse. In the example shown in the figure, the cabin 110 is a sphere, but this is not intended to limit the embodiments of this disclosure. In addition, it can also be set as a cuboid or other irregular cube.

[0042] It should be noted that the spherical, ellipsoidal, or elliptical cabin 110 shape possesses excellent structural mechanical properties, and its rounded shape reduces safety hazards in public settings. Furthermore, this shape naturally adapts to concentric circular spatial layouts, allowing for the even distribution of sales windows 130, entrances / exits 140, and shelves 160 along the circumference, maximizing the use of the cabin space 120 and supporting parallel operation of multiple windows. Simultaneously, the continuous curved surface of the cabin 110's outer wall can completely cover the display device, forming a wraparound display interface. Regardless of the direction from which customers approach, a complete visual display effect can be obtained, significantly enhancing customer attraction and interactive value. In addition, the rounded outline of the cabin 110 naturally guides the flow of customers outside the cabin to disperse and circle around it, and the surrounding apron performance area 170 can accommodate the viewing needs of tourists from various directions, effectively avoiding crowd congestion. It combines spatial functionality, scenic landmark status, and operational adaptability, making it particularly suitable for unmanned retail and interactive exhibition needs in complex crowd scenarios such as commercial districts and cultural tourism scenic spots.

[0043] In some embodiments of this disclosure, each arc segment is provided with a corresponding angle range. An arc segment anchor point mapping relationship is established between the angle range and the pixel display area on the transparent grid screen 151. The arc segment anchor point table records the angle range of each arc segment and its corresponding pixel display area on the transparent grid screen 151. Based on the unified master timecode, the system performs spatiotemporal alignment on the digital human image, the merchandise display layer, and the embodied robot live layer. Spatiotemporal alignment includes network latency compensation and parallax offset compensation calibrated according to the curvature of the dome curtain wall. Network latency compensation refers to the displacement correction on the time axis of the telemetry video frame of the embodied robot caused by network transmission delay, so that the timestamp of the video frame is aligned with the master timecode. Parallax offset compensation refers to the calculation of the pixel offset of the virtual overlay layer on the screen according to the radial position of the embodied robot in the arc segment and the dome curvature parameter of the transparent grid screen, so that the virtual overlay layer seen by the customer outside the cabin is consistent with the projection position of the physical embodied robot on the curtain wall, eliminating the visual deviation caused by the dome curvature.

[0044] In some embodiments of this disclosure, the transparent grille screen 151 is used to carry a product information layer, a live background layer of the embodied robot cabin or its outer skirt area, and an interactive overlay layer. The top circular display screen 152 is used to carry digital human narrative theater content with a duration of not less than 30 seconds. The content of the two screens is synchronized through a master timecode, forming a dual-screen layered collaborative display. The product information layer is used to display product information such as product images, prices, and text descriptions. The live background layer of the embodied robot refers to a transparent background video layer obtained by removing the background from the real-time video of the embodied robot's operation or performance. This layer is overlaid on the corresponding arc-shaped display area of ​​the transparent grille screen 151, allowing customers outside the cabin to see the real-time dynamics of the embodied robot 200 inside the cabin. The interactive overlay layer is used to display visual elements such as prompts and guiding arrows related to customer interaction. The digital human narrative theater content refers to a narrative brand story or performance content presented by a virtual digital human character and played through the top circular display screen 152.

[0045] In some embodiments of this disclosure, the system establishes an SKU display inventory binding table. Each record in this binding table includes at least a product identifier, a corresponding sales window identifier, a corresponding shelf identifier, a corresponding available inventory quantity, and a highlight display priority weight. The product identifier uniquely identifies a product category or individual item; the sales window identifier identifies which sales window 130 the product is delivered in; the shelf identifier identifies which shelf 160 the product is stored on; the available inventory quantity refers to the current inventory quantity available for sale; and the highlight display priority weight is used to determine display priority when multiple products compete for display resources. In performance mode, this priority weight is higher than in retail mode. This binding table structurally records and maintains the association between the products displayed on the display device and the shelves 160 and sales windows 130. Before triggering a product highlight display or performance-driven display, the system verifies whether the available inventory quantity of the target product is greater than zero based on the binding relationship. Only when the available inventory quantity is greater than zero is the highlight display content of the product allowed to be played on the transparent grid screen 151 or the top circular display screen 152. When the available inventory drops to zero, the system automatically removes the highlight display of the product and switches to an alternative product with a greater available inventory in the same sales window 130. At the same time, it sends a notification of the change in the target of traffic acquisition to the upper-level scheduling system.

[0046] In some embodiments of this disclosure, the system collects the voice interaction intentions and / or gesture interaction intentions of customers outside the cabin. Voice interaction intentions are collected via the microphone array in the customer area 173 and analyzed using speech recognition technology. Gesture interaction intentions are collected via a surround-view vision sensor and analyzed using gesture recognition technology. Based on the customer's circumferential angular coordinates within the horizontal plane of the customer area 173, the system selects the arc segment with the smallest angular distance from the concentric circle layout. The sales window 130 bound to this arc segment is used as the target window for interaction routing, and the interaction response result is routed to the product selection interface of that window. For example, when a customer points to a product and says "I want this," the system determines the sales window directly in front of the customer based on their circumferential position, adds the corresponding product to the shopping cart of that window, or displays the product details page.

[0047] In some embodiments of this disclosure, the product selection engine receives multi-source input data, including inventory telemetry data for each shelf 160 zone, sales data for each sales window, foot traffic index, conversion index, and the current operating mode. Inventory telemetry data refers to the real-time inventory quantity monitored by weight sensors or photoelectric sensors installed on the shelves 160. Sales data refers to the sales records of each sales window 130 within a preset time window. The product selection engine uses maximizing the product of weighted gross profit and demand forecast as its objective function. Weighted gross profit refers to the gross profit margin of the product multiplied by the product selection weight determined according to the operating mode. Demand forecast refers to the future demand quantity predicted based on historical sales data and external factors. Constraints of the product selection engine include a category richness not lower than a preset lower bound and an inventory quantity in each shelf 160 zone not exceeding the corresponding arc capacity. Category richness refers to the number of different product categories within each shelf 160 zone. The lower bound ensures product diversity and avoids excessive concentration on a few best-selling categories. Arc capacity refers to the maximum physical capacity of each shelf 160 within each arc. The product selection engine optimizes and solves the above constraints to generate a replenishment plan that includes several replenishment tasks.

[0048] In some embodiments of this disclosure, the replenishment plan includes several replenishment tasks, which are written into a replenishment task queue. The system sets the priority of replenishment tasks based on the current operating mode and order service level agreement (SLA) pressure. In retail mode, replenishment tasks have a lower priority than order picking tasks, and replenishment tasks must give way to order picking tasks. In performance mode or hybrid mode, replenishment tasks have a lower priority than performance tasks, but the embodied robot 200 is allowed to perform replenishment tasks across arcs. The replenishment path follows the traffic rules of circumferential movement priority and radial crossing reservation of the hub layer L0. Cross-arc replenishment means that the embodied robot 200 can leave its bound arc and go to the shelf 160 of other arcs to perform replenishment tasks. When the order service level agreement of any sales window 130 alarms, the system immediately suspends or cancels the currently executing replenishment task and switches the embodied robot 200 performing the replenishment task to perform the order picking task of that window, ensuring the absolute priority of order response efficiency.

[0049] In some embodiments of this disclosure, when the embodied robot 200 performs a replenishment task, the system determines the target arc segment where the target shelf for the replenishment task is located. A replenishment progress overlay is rendered on the transparent grid screen 151 within the display area corresponding to the target arc segment. The replenishment progress overlay includes at least a replenishment status indicator and information about the products being replenished. This overlay is rendered only within the display area of ​​the transparent grid screen 151 corresponding to the arc segment. The overlay is automatically cleared upon completion of replenishment. This design allows customers outside the cabin to see the replenishment dynamics inside the cabin in real time, enhancing operational transparency.

[0050] In some embodiments of this disclosure, the system receives a job mode switching command output by the upper-level scheduling system, switching between a performance-style display skin and a retail-style display skin between performance mode and retail mode, with the switching process employing a gradual transition of 5 to 15 seconds. The performance-style display skin refers to the display interface style that matches the performance mode, including a theater-style product display, a digital human narrative background, and highlight displays for lead-generation products. The retail-style display skin refers to the display interface style that matches the retail mode, including a simple product queue, order status, and queuing information. Gradual transition means that the displayed content smoothly changes from one skin style to another, rather than switching abruptly, thereby avoiding visual abruptness and abrupt changes in user experience caused by abrupt cuts.

[0051] Please see Figure 5 , Figure 5 This is a flowchart illustrating the automatic product selection and replenishment method for a smart cabin provided in this embodiment. This automatic product selection and replenishment method for a smart cabin can be applied to the automatic product selection and replenishment system in the above embodiments, and is jointly executed by the smart cabin and the embodied robot. The automatic product selection and replenishment method for the smart cabin includes steps S501 to S504: Step S501: Obtain inventory data for each shelf, sales data for each sales window, and the current operating mode of the smart cabin. Step S502: Determine the corresponding product selection strategy based on the working mode, and generate a replenishment plan based on inventory data, sales data and product selection strategy. The replenishment plan contains several replenishment tasks. Step S503: Determine the priority of each replenishment task according to the working mode, and schedule the corresponding embodied robot to execute the replenishment task in sequence according to the replenishment plan and the priority of the replenishment task. In step S504, during the process of the embodied robot performing replenishment tasks, the control display device displays the replenishment progress information corresponding to the replenishment tasks, and updates the inventory data after each replenishment task is completed.

[0052] Regarding step S501 above, inventory data refers to the quantity of goods currently stored on each shelf. This data reflects the availability of various goods within the smart warehouse and is the basis for determining whether replenishment is needed and how much to replenish. Inventory data can be collected in real time through various methods. For example, weight sensors can be installed on the shelves to detect increases or decreases in the quantity of goods through weight changes; photoelectric sensors can be installed on the shelves to detect the presence or absence of goods through light occlusion; or visual sensors can be used to periodically photograph the shelves and analyze the quantity of goods using image recognition algorithms. These collection methods can be selected and combined according to actual deployment costs and accuracy requirements. The real-time nature and accuracy of inventory data directly determine the effectiveness of replenishment decisions. If the inventory data is lagging or has excessive deviation, it may lead to replenishment being too early or too late. Replenishment too early wastes resources, while replenishment too late causes stockouts and affects sales.

[0053] Sales data refers to the sales records of goods sold at each sales window within a preset time window, including information such as the category, quantity, and transaction time of the goods sold. Sales data reflects the actual market demand and is an important basis for determining which goods need to be restocked and how much to restock. Sales data can be collected in real time through the transaction system of the sales window. Whenever a customer completes a purchase, the product information, quantity, and transaction time of that transaction are recorded and summarized in the sales database. Sales data not only reflects historical sales, but can also be used to predict future sales trends through methods such as time series analysis, providing data support for the formulation of product selection strategies. For example, if the sales of a certain category of goods suddenly increase in the past hour, it may indicate that the market demand for that category of goods is rising, and the product selection engine should prioritize replenishing inventory for that category of goods when generating a replenishment plan.

[0054] The operating mode refers to the current operational status of the smart cabin, which can include at least one of performance mode, hybrid mode, and retail mode. The operating mode determines the current operational focus of the smart cabin: performance mode focuses on attracting customers with performances, retail mode focuses on merchandise sales, and hybrid mode balances both. The operating mode can be obtained by reading the operating mode flag within the control unit. This flag is dynamically set by the upper-level scheduling system based on real-time data such as passenger flow density and conversion rate outside the cabin. The operating mode is the core basis for formulating product selection strategies; the priority of product categories required by the smart cabin differs significantly under different operating modes.

[0055] The control unit synchronously acquires three types of data—inventory data, sales data, and operating mode data—and incorporates them into a unified data processing flow. This means that replenishment decisions no longer rely on a single type of data in isolation, but rather integrate information from three dimensions: supply status, market demand, and operational strategies. This provides a complete data foundation for subsequent product selection optimization and replenishment scheduling.

[0056] Regarding step S502 above, the product selection strategy refers to the set of rules for product selection and replenishment decisions determined based on the current operating mode. The core content of the product selection strategy includes the selection weights of each product category and the setting of constraints. The selection weight is a numerical coefficient used to quantify the importance of different product categories in the current operating scenario; a higher weight indicates that the product category is more worthy of priority replenishment and key support under the current operating mode. The product selection strategy is dynamically adjusted according to changes in the operating mode, enabling replenishment decisions to adapt to different operational objectives.

[0057] When the work mode is performance mode or hybrid mode, the selection weight of traffic-driving products is increased. Traffic-driving products refer to those with high visual appeal, suitable for display and promotion during performances, such as beautifully packaged cultural and creative products and visually striking specialty items. In performance mode, the core operational goal of the smart cabin is to attract customer traffic by guiding the audience's attention to specific products through the performances of embodied robots and the displays of digital humans, thereby achieving the effect of attracting traffic. Therefore, inventory assurance for traffic-driving products has a higher priority in performance mode. Increasing their selection weight makes the product selection engine more inclined to arrange restocking for these products, ensuring that there are goods to display and sell during performance-driven traffic generation.

[0058] When operating in retail mode, the selection weight of high-turnover product categories is increased. High-turnover products refer to goods with high sales volume and fast turnover, such as best-selling beverages, snacks, and other daily consumer goods. In retail mode, the core operational goal of the smart cabin is to respond quickly to order demands and maximize sales. Therefore, ensuring sufficient inventory for high-turnover products has a higher priority. Increasing their selection weight allows the product selection engine to prioritize ensuring adequate inventory for these best-selling products, avoiding lost sales opportunities due to stockouts.

[0059] A replenishment plan is a replenishment operation plan generated based on inventory gaps, sales trends, and product selection strategies. A replenishment plan comprises several replenishment tasks, each of which must include at least the target shelf, target product, and replenishment quantity. The generation process of a replenishment plan is essentially a constrained optimization process.

[0060] Specifically, the product selection engine uses the product of weighted gross profit and demand forecast as its objective function. Weighted gross profit refers to the product's gross profit margin multiplied by the product selection weight determined based on the work pattern, reflecting the product's contribution to profit after deducting costs. Demand forecast is based on historical sales data and uses time series analysis to predict future demand. The design of this objective function reflects the core business objective of replenishment decisions: prioritizing products that are both profitable and easy to sell.

[0061] Meanwhile, the product selection engine is also subject to two types of constraints. The first constraint is that the category richness must not be lower than a preset lower bound, meaning the number of different product categories within a shelf section cannot be less than a pre-set minimum requirement. This constraint exists to ensure product diversity and prevent the product selection engine from excessively focusing on a few high-margin, high-turnover best-selling categories to maximize profits, which would limit customer choice and degrade the customer experience. The second constraint is that the inventory level of each shelf section cannot exceed the capacity of the corresponding arc segment, i.e., the maximum physical capacity of the shelves within each arc segment. This constraint exists because the shelf space in each arc segment is limited, and the replenishment plan cannot exceed the physical carrying capacity of the shelves. The product selection engine optimizes under these constraints, outputting a replenishment plan that satisfies the constraints and maximizes the objective function value.

[0062] It should be noted that the embodiments of this disclosure model the product selection problem as a constrained optimization problem, replacing traditional manual experience-based decision-making with a data-driven approach. This achieves automation, scientification, and optimization of product selection and replenishment. At the same time, it ensures product diversity through category richness constraints and physical feasibility through arc capacity constraints, directly improving inventory turnover efficiency and operational effectiveness from the product selection level.

[0063] Regarding step S503 above, the priority of replenishment tasks refers to the execution order of each replenishment task in the embodied robot's scheduling queue. The higher the priority, the earlier the replenishment task is executed. The specific value of the priority can be set according to the rule that the smaller the value, the higher the priority. The priority of replenishment tasks is not fixed, but dynamically determined according to the current working mode, so that the rhythm of replenishment operations and resource consumption can adapt to the current operational strategy.

[0064] Furthermore, when the operating mode is retail mode, replenishment tasks have a lower priority than order picking tasks. Order picking refers to responding to customer orders, retrieving goods from the shelves, and delivering them to the sales window. In retail mode, the core operational activity of the smart warehouse is to quickly respond to orders and ensure customers receive their purchased goods promptly. Therefore, order picking tasks should have the highest resource priority, while replenishment tasks are subordinate in retail mode, and their execution should not interfere with the normal operation of order picking. When both replenishment and order picking tasks are waiting for scheduling, the system prioritizes order picking tasks, and replenishment tasks are executed after order picking is completed or during idle periods.

[0065] Furthermore, in the retail model, replenishment tasks are limited to the corresponding embodied robot in the arc where the target shelf is located, and cross-arc replenishment is not performed. This is because cross-arc replenishment requires the embodied robot to leave its bound arc, which may increase the risk of traffic conflicts when crossing the hub layer or passing through other arcs. Such additional traffic burden should be avoided as much as possible in the retail model.

[0066] When the operating mode is performance mode or hybrid mode, the priority of replenishment tasks is lower than that of performance tasks. Performance tasks refer to tasks performed by avatar robots inside or outside the cabin, such as dancing, facial expressions, and interactive performances to attract customers. In performance mode, the core operation of the smart cabin is to attract customers through performances; therefore, performance tasks should enjoy the highest resource priority. Replenishment tasks are also subordinate in performance mode, and their execution should not interfere with the normal performance. However, at the same time, in performance mode or hybrid mode, avatar robots performing replenishment tasks are allowed to perform replenishment tasks across arcs. This is because in performance mode, some avatar robots are in performance mode, while others may be idle. Allowing idle avatar robots to perform replenishment across arcs can utilize dispersed idle resources to complete the replenishment needs of each arc. Furthermore, the cross-arc replenishment path must follow the smart cabin's circumferential movement priority and the radial crossing reservation traffic rules of the hub layer, improving replenishment efficiency while ensuring safety.

[0067] During the scheduling and execution of replenishment tasks, the control unit queues each replenishment task in the replenishment plan according to its priority from highest to lowest, forming a replenishment task queue. When the embodied robot is idle, the control unit retrieves the highest priority replenishment task from the queue, determines the replenishment location based on the target shelf information in the task, and dispatches the corresponding embodied robot to perform the replenishment. After the replenishment task is completed, it is removed from the queue, and the next task is scheduled, until all replenishment tasks are completed.

[0068] It should be noted that the embodiments disclosed herein achieve coordinated adaptation between replenishment operations and operating modes by linking the priority of replenishment tasks with the working mode. In the retail mode, replenishment is ensured to give way to order picking, while in the performance mode, replenishment is allowed to be carried out using performance gaps and idle embodied robot resources. This ensures that the resources of core operational activities are prioritized, while making full use of idle resources to complete replenishment, avoiding resource competition between replenishment and core operational activities, and improving the overall operational coordination of the system.

[0069] Regarding step S504 above, replenishment progress information refers to visual information reflecting the execution status of the replenishment task, including at least a replenishment in progress status indicator and information on the products being replenished. The replenishment in progress status indicator is used to inform customers outside the cabin that a replenishment operation is currently underway for that segment, and the information on the products being replenished informs customers which specific products are being added to the shelves. The display of replenishment progress information makes the replenishment operation, which was originally hidden inside the cabin, transparent and visible, enhancing customers' perception and trust in the smart cabin operation.

[0070] During the display of replenishment progress information, when the embodied robot begins executing a replenishment task, the control unit first determines the target arc segment where the target shelf for that replenishment task is located. Since each shelf is situated within a corresponding arc segment, and each arc segment has an anchor point mapping relationship with a display area on the transparent grid screen, the control unit can accurately determine the display area position corresponding to the replenishment task. Then, the control unit overlays and renders a replenishment progress overlay within the transparent grid screen display area corresponding to the target arc segment. This replenishment progress overlay is a semi-transparent visual layer that does not completely obscure the underlying display content; instead, it adds a layer of information on top of the underlying content. The content of the replenishment progress overlay includes at least a replenishment status indicator, such as a text prompt or dynamic icon indicating that replenishment is in progress, and information about the products being replenished, such as product names and images. When the replenishment task is completed, the control unit automatically clears the replenishment progress overlay, and the transparent grid screen of that arc segment resumes its normal product display content.

[0071] After the replenishment task is completed, the control unit also needs to update the inventory data. Specifically, the control unit adds the replenishment quantity corresponding to this replenishment task to the inventory record of the corresponding product on the corresponding shelf, so that the inventory data is consistent with the actual inventory status. The updated inventory data will serve as the input for the next round of replenishment decisions, forming a data-driven closed loop. In addition, the updated inventory data will also be synchronized to the SKU display inventory binding table, so that subsequent highlight display verification can be based on the latest inventory status, avoiding the display of sold-out products.

[0072] It should be noted that, in this embodiment of the present disclosure, by simultaneously controlling the display device to show the corresponding replenishment progress information during the process of the robot performing the replenishment task, and automatically updating the inventory data after the replenishment is completed, the visualization of the replenishment operation and the real-time linkage of inventory data are realized, so that customers outside the cabin can intuitively perceive the replenishment dynamics inside the cabin, and the real-time accuracy of the inventory data is guaranteed.

[0073] In summary, this embodiment of the present disclosure, through the automatic product selection and replenishment method of the intelligent cabin in steps S501 to S504, simultaneously acquires inventory data from each shelf, sales data from each sales window, and the current operating mode of the intelligent cabin. This integrates the previously fragmented data of inventory, sales, and operational status into a unified management system. Then, it matches corresponding product selection strategies based on the current operating mode and dynamically generates replenishment plans containing multiple replenishment tasks by combining real-time inventory and sales data. This replaces the experience-based decision-making of traditional manual replenishment, dynamically optimizing the product selection structure based on real-time operational data to match market demand. This effectively alleviates the contradiction between stockouts of best-selling products and stockpiles of slow-moving products, directly addressing the issue from the product selection perspective. First, improve inventory turnover efficiency; then, determine the execution priority of each replenishment task according to the working mode, and schedule the embodied robot to execute the replenishment task in sequence, so that the rhythm of replenishment operation and resource consumption are adapted to the current operating mode. From the scheduling level, the scheduling conflict between replenishment operation and daily operation is avoided, and the overall system operation coordination is improved; finally, during the process of the embodied robot executing the replenishment task, the display device is synchronously controlled to display the corresponding replenishment progress information, and the inventory data is automatically updated after the replenishment is completed. This realizes real-time linkage between the displayed content and the inventory status, ensures the real-time accuracy of the inventory data, and ultimately comprehensively improves the system coordination and inventory turnover efficiency in the unmanned retail scenario.

[0074] Furthermore, in some embodiments, step S503 determines the priority of each replenishment task according to the working mode, and schedules the corresponding embodied robot to execute the replenishment task in sequence according to the replenishment plan and the priority of the replenishment task, specifically including two sub-steps: position reservation and partition scheduling.

[0075] Specifically, before executing a replenishment task scheduling, at least one location within the warehouse space is selected as the target location from multiple locations within the warehouse. This target location is a warehouse location reserved for manual replenishment operations. The selection of the target location can be dynamic. The system dynamically determines the target location to be reserved based on the target product type of each replenishment task in the current replenishment plan and the real-time workload of each arc segment. For example, when the replenishment plan includes replenishment tasks for non-standard specifications, the system selects a shelf location or adjacent area within the arc segment containing the non-standard specifications as the target location; when the replenishment plan does not include non-standard specifications, the system can select a shelf location within an idle arc segment or a designated area in the hub layer as the target location to accommodate potential manual intervention needs during the replenishment process.

[0076] In addition, target locations can be selected using a static preset method, where specific shelf locations or space areas are reserved as fixed manual replenishment operation areas during the cabin design phase, or by receiving user calibration information and selecting one or more corresponding locations from multiple cabin locations based on user needs. The dynamic selection method offers greater flexibility, allowing for dynamic adjustment of reserved locations based on actual replenishment needs, adapting to different replenishment scenarios; the static preset method provides determinism, facilitating the training of manual replenishment personnel and the standardization of operating procedures. The two methods can be selected and combined according to actual operational needs, for example, simultaneously setting up fixed and dynamic reserved locations within the cabin to balance determinism and flexibility.

[0077] After selecting the target location, the system excludes it from the automatic replenishment scheduling range. That is, among the remaining locations within the warehouse besides the target location, the priority of each replenishment task is determined based on the current working mode, and the corresponding embodied robot is scheduled to execute the replenishment task sequentially according to the replenishment plan and its priority. In other words, the target location is not involved in the automatic replenishment scheduling of the embodied robot, and the embodied robot will not use the target location as a replenishment target or a stopping point in its travel path when performing replenishment tasks. This exclusion mechanism ensures that the manual replenishment operation space remains unobstructed and available during the replenishment process, avoiding spatial conflicts or operational interference between the embodied robot and manual replenishment personnel in the same location.

[0078] Furthermore, during the replenishment process performed by the embodied robot, when the target product for any replenishment task is a non-standard specification item, the system marks the replenishment task as a mobile replenishment task and displays replenishment prompts at the target location. This allows human replenishment personnel to place the non-standard specification item at the target location based on the replenishment prompts. Non-standard specification items refer to goods that are too large, too heavy, or irregularly shaped, making them unsuitable for automatic grasping and handling by the embodied robot, such as large plush toys or oddly packaged souvenirs. The replenishment prompts are displayed on a transparent grid screen within the corresponding display area at the target location. These prompts include at least the name, quantity, and storage instructions of the item to be replenished, guiding human replenishment personnel to quickly complete the replenishment operation. After replenishment is completed at the target location, the system obtains human replenishment confirmation information, updates the available inventory quantity of the corresponding item in the SKU display inventory binding table based on the confirmation information, and restores the target location to an available state. Throughout the entire process of the mobile replenishment task, except when the target location is occupied, the system still schedules other idle robots to continue to perform the remaining replenishment tasks in the replenishment plan according to the determined priority, so that mobile replenishment operations and automatic replenishment operations can be carried out in parallel without interfering with each other.

[0079] It should be noted that, in this embodiment, by pre-selecting at least one target location from multiple positions within the warehouse as a reserved space for manual replenishment operations before scheduling and executing replenishment tasks, and excluding the target location from the automatic replenishment scheduling range, a safe and unobstructed operating area is preserved for manual replenishment while ensuring that the embodied robot's automatic replenishment operations in the remaining area are not interfered with. This zoned scheduling mechanism achieves spatial isolation and parallel collaboration between automatic and manual replenishment, leveraging the automation advantages of the embodied robot while covering the replenishment needs of non-standard specification goods that are difficult for the embodied robot to handle. It avoids interruptions to the entire replenishment process due to manual replenishment intervention, improving the scenario adaptability and overall replenishment efficiency of the replenishment system.

[0080] Furthermore, during the replenishment task scheduling process, the system can select at least one of the multiple sales windows on the cabin wall as a reserved sales window. This reserved sales window is a sales window reserved for manual sales operations. The selection of reserved sales windows can be dynamic. The system dynamically determines which sales windows need to be reserved based on the current order flow, queue length, and customer interaction intent data of each sales window. For example, when the order service level agreement of a certain sales window is under continuous high pressure and the queue length exceeds a preset threshold, the system selects that sales window as a reserved sales window to alleviate the service pressure of that window through manual intervention. Similarly, when the arc segment corresponding to a certain sales window contains non-standard specifications or special goods requiring manual assistance, the system can also select that sales window as a reserved sales window. The selection of reserved sales windows can also be statically preset, that is, designating one or more sales windows as fixed manual sales service windows during the cabin design stage, or receiving user calibration information and selecting one or more corresponding windows as reserved sales windows according to user needs. The static preset method facilitates the job setting of sales staff and customer identification, while the dynamic selection method can flexibly respond to the service pressure of different windows based on real-time operational data. The two methods can be selected and combined according to actual operational needs.

[0081] Furthermore, during the replenishment process, when any customer arrives at a designated sales window, the system displays a human service prompt on the corresponding display area of ​​the window via a transparent grid screen. This prompt includes at least a "Service in Progress" status indicator, the current queue order, and the estimated waiting time, guiding customers to wait orderly for human assistance. Human staff provide services based on the prompts, including but not limited to product recommendations, on-site delivery of non-standard items, response to special requests, and payment assistance. Displaying the human service prompt in the corresponding area ensures customers can clearly identify the window as a human service window, preventing confusion between automated and human-operated vending machines and improving the customer experience.

[0082] After a sales transaction is completed, the salesperson enters the sales confirmation information through a handheld terminal or the confirmation device built into the window. Upon receiving the confirmation information, the system updates the available inventory quantity of the corresponding product in the SKU display inventory binding table based on the product category and quantity recorded in the sales confirmation information, ensuring the real-time accuracy of inventory data.

[0083] Furthermore, a locational correspondence is established between the reserved sales window and the aforementioned target location, ensuring that they belong to the same arc segment in space. Through this design, after a customer places an order manually at the reserved sales window, the corresponding robot can pick the item from the shelf within that arc segment and deliver it to the reserved sales window, where a human salesperson completes the final delivery service, achieving a collaborative division of labor between human sales and robot picking. Once the human sales service is completed, the system automatically de-reserves the sales window and restores it to automatic sales mode, allowing the system to flexibly switch between human and automatic sales modes according to operational needs.

[0084] It should be noted that this embodiment of the disclosure reserves a dedicated service window for manual sales by selecting at least one of multiple sales windows as a manual sales window. The display device shows manual service prompts in the corresponding display area, guiding customers to receive manual service in an orderly manner. Inventory data is automatically updated after the manual sale is completed, achieving real-time synchronization of inventory data in manual sales scenarios. This design leverages the efficiency advantages of automated sales by the embodied robot while also covering service scenarios requiring human intervention, such as the sale of special goods and special customer needs. This enhances the scenario adaptability and service flexibility of the intelligent cabin sales system, while also providing customers with more personalized shopping options.

[0085] In some embodiments, the automatic product selection and replenishment method may also include a service level agreement (SLA) preemption step, where SLA refers to the promised time from order placement to delivery completion, such as picking, packing and delivery within 90 seconds.

[0086] During replenishment tasks, the control unit continuously monitors the service level agreement (SLA) status of orders at each sales window. The SLA status refers to the ratio of the elapsed time from order placement to the promised time under the SLA. This ratio reflects the urgency of order timeouts. When this ratio exceeds a preset alarm threshold, such as 80%, it means the order is about to time out and requires immediate response. At this point, the control unit pauses or cancels the currently executing replenishment task and switches the embodied robot performing the replenishment task to perform the order picking task for that window. After a replenishment task is paused, its current progress is saved and it resumes execution after the order picking task is completed. After a replenishment task is canceled, it is re-queued for subsequent scheduling. This preemptive mechanism ensures the absolute priority of order response efficiency in retail scenarios; replenishment tasks cannot be performed at the expense of order delivery.

[0087] It should be noted that the embodiments disclosed herein achieve dynamic resource allocation of replenishment tasks and order picking tasks through the service level agreement preemption mechanism, thereby ensuring the core retail indicator of order delivery timeliness.

[0088] In some embodiments, the automatic product selection and replenishment method may further include an SKU display inventory linkage step. In this embodiment, SKU represents a specific product category or individual item. The system establishes an SKU display inventory binding table, which records the association between the products displayed on the display device and the shelves and sales windows. Each record in the binding table includes at least a product identifier, a corresponding sales window identifier, a corresponding shelf identifier, and a corresponding available inventory quantity. The product identifier uniquely identifies a product category, the sales window identifier identifies which sales window the product is delivered at, the shelf identifier identifies which shelf the product is stored on, and the available inventory quantity refers to the current inventory quantity available for sale. This binding relationship establishes a traceable link between the displayed content, inventory status, and delivery window, transforming them from isolated information silos.

[0089] Before displaying the highlight content of the target product on the display device, the system verifies whether the available inventory quantity of the target product is greater than zero based on the binding relationship. Highlighted content refers to the prominent display of the product, such as presenting product images and promotional information on a transparent grid screen using magnification, highlighting, or dynamic effects. Only when the available inventory quantity is greater than zero is the highlight content of the product allowed to be played on the transparent grid screen or the top circular display screen, ensuring that the displayed product is in stock and available for sale. When the available inventory quantity drops to zero, the system automatically cancels the highlight display of the product and switches to an alternative product with a greater than zero available inventory quantity in the same sales window, while simultaneously sending a notification of the change in the target product to the upper-level scheduling system.

[0090] It should be noted that, through the SKU display inventory binding table and the available inventory verification before highlight display, the display inventory that is displayed on the screen is the same as the available inventory, which effectively avoids the embarrassing situation of displaying out-of-stock items and improves customer experience and purchase conversion rate.

[0091] In some embodiments, the automated product selection and replenishment method may further include an interactive traffic guidance step. The control unit collects interaction intent information from visitors in the outdoor customer area. This interaction intent information includes voice interaction intent and / or gesture interaction intent. Voice interaction intent is obtained by collecting customers' voice commands through a microphone array in the outdoor customer area and parsing them using speech recognition technology. Gesture interaction intent is obtained by collecting customers' hand gestures through a surround-view vision sensor and parsing them using gesture recognition technology. Based on the visitor's circumferential position within the outdoor customer area, the control unit determines the target sales window closest to the visitor from among the various sales windows.

[0092] Specifically, the system uses visual sensors to determine the visitor's circumferential angular coordinates in the horizontal plane. It then selects the arc with the smallest angular distance from the concentric circle layout's arc segments, using the sales window associated with that arc as the target window. This selection logic is based on the geometric characteristics of the concentric circle layout: the customer's gaze and ease of operation are aligned with the sales window corresponding to the arc segment they are facing. Routing the interaction result to the nearest window maximizes the convenience for the customer to complete subsequent purchase operations. The control unit generates the corresponding interaction response based on the interaction intent information and routes the response to the product interface of the target sales window. For example, when a customer says "I want this" and gestures towards a product, the system recognizes the product and pushes its information or purchase interface to the screen of the sales window closest to the customer.

[0093] It should be noted that, through the interactive traffic redirection method, the voice or gesture interaction of customers outside the cabin is accurately routed to the nearest sales window, realizing the conversion link of interaction into sales and improving the customer's purchase conversion rate.

[0094] In some embodiments, the automatic product selection and replenishment method may further include a display skin switching step. The control unit acquires a display skin switching command from the smart cabin, triggered by a change in the operating mode. When the operating mode switches from performance mode to retail mode or vice versa, the system automatically generates a display skin switching command. The control unit, based on the command, controls the display device to switch from the current display skin to the target display skin, with the switching process employing a gradual transition. Here, the display skin refers to the overall visual style and interface layout of the display device, including color schemes, font styles, background patterns, and interface element layouts. Display skins include performance-type display skins corresponding to performance mode and retail-type display skins corresponding to retail mode.

[0095] The performance-style display skin includes theatrical product display styles, digital human narrative backgrounds, and highlighted product displays, with an overall style leaning towards immersion and entertainment. The retail-style display skin includes simple product queues, order status, and queuing information, with an overall style leaning towards efficiency and clarity. Gradient transition refers to a smooth, gradual change of display content from one skin style to another, rather than an instantaneous switch. The duration can range from 5 to 15 seconds, depending on the specific needs. Gradient transitions can be implemented through a gradual change in transparency, fading out the current skin's content while simultaneously fading in the target skin's content; or through animation transformation, smoothly transforming the current skin's interface elements into the corresponding elements of the target skin.

[0096] It should be noted that the embodiments of this disclosure achieve a smooth transition between performance-type skins and retail-type skins by displaying a synchronous and gradual switching between skins and operating modes, avoiding visual abruptness and experience breaks caused by hard cuts, and ensuring a consistent visual experience for customers during the switching process between different operating modes.

[0097] In some embodiments, the automated product selection and replenishment method may further include a mobile replenishment coordination step. The system receives replenishment confirmation for non-standard specification goods uploaded by the mobile replenishment terminal. The non-standard specification goods are placed into designated shelf areas by on-site mobile replenishment personnel. The SKU binding management module updates the available inventory quantity based on the replenishment confirmation and unlocks the out-of-stock lock of the corresponding SKU. Non-standard specification goods refer to products that are too large, too heavy, or irregularly shaped, and are not suitable for automated replenishment by the robot. The mobile replenishment terminal can be a handheld barcode scanner or a mobile application. After the on-site replenishment personnel have completed the manual shelving of non-standard specification goods, they scan the product barcode or enter product information through the mobile replenishment terminal. After receiving the replenishment confirmation, the system automatically updates the inventory data.

[0098] It should be noted that the embodiments disclosed herein achieve an organic combination of automatic replenishment by embodied robots and manual mobile replenishment through mobile replenishment collaboration. This not only leverages the automation advantages of embodied robots but also covers the special commodity replenishment needs that are difficult for embodied robots to handle.

[0099] In some embodiments, the automatic product selection and replenishment method may further include a spatiotemporal context reporting step. The control unit writes the SKU display inventory binding table, replenishment task queue status, and current display skin type into the SpatiotemporalContextV2 cabin-level extended metadata, which is then read by the upper-level elastic scheduling system to achieve consistency between performance-driven traffic generation, sales conversion, and inventory assurance. SpatiotemporalContextV2 is a unified data structure that includes core identifiers, four-dimensional information, multi-sensory data, an activity object list, and metadata. The metadata stores auxiliary information such as data source and data quality score. In this embodiment, the SKU display inventory binding table, replenishment task queue status, and current display skin type are written into the metadata field. Through this design, the display inventory linkage status and replenishment scheduling status of the smart cabin can be perceived and recorded in real time by the upper-level management system, providing a standardized data foundation for multi-cabin collaborative scheduling, remote monitoring, and data analysis.

[0100] In some embodiments, the automatic product selection and replenishment method may further include an autonomous operation and maintenance step during non-working hours. Specifically, the control unit initiates an autonomous operation and maintenance process when the smart cabin is in a non-operational state, based on a preset non-working hour schedule or by receiving a non-working hour instruction from the upper-level scheduling system. The autonomous operation and maintenance process includes at least one of an automatic inventory task, an automatic charging task, and an environmental inspection task.

[0101] The automatic inventory task involves scheduling all or some robots to conduct a one-by-one inventory check of goods in each shelf section. The embodied robot verifies the actual inventory quantity against the inventory data recorded by the system by visual recognition or reading data from shelf sensors. When the inventory data deviation exceeds a preset tolerance threshold, the inventory record is automatically corrected and an inventory discrepancy report is generated. The automatic charging task involves controlling the embodied robots to return sequentially to the charging piles located in the hub layer or designated arc section of the cabin for automatic charging during non-working hours. This ensures that the embodied robots have sufficient power before the start of the next working period. The charging priority is sorted according to the remaining power of each robot from low to high, with robots with low power being charged first. The environmental inspection task refers to the embodied robot conducting patrol inspections in various arc sections and apron performance areas within the cabin. It collects environmental data such as temperature, humidity, smoke, and abnormal sounds through onboard environmental sensors, and detects the status of equipment, shelf integrity, and entrance / exit door status through visual sensors. When an abnormality is detected, such as abnormal temperature, smoke alarm, equipment failure, or door abnormality, the system automatically generates an abnormal event report, including the event type, location, time of occurrence, and relevant sensor data, and reports it to the upper management system in real time.

[0102] Furthermore, the execution sequence and resource allocation of automatic inventory counting, automatic charging, and environmental inspection tasks are coordinated by the control unit based on the total duration of non-working periods, ensuring that all necessary maintenance tasks are completed before the end of the non-working period. Through this design, the intelligent cabin can autonomously complete maintenance tasks such as inventory counting, robot charging, and environmental inspection during non-working periods without manual on-site operation. This achieves a fully automated closed loop from product selection, replenishment, and shelving to inventory counting, charging, and anomaly reporting, significantly reducing manual maintenance costs and improving the reliability and continuity of system operation.

[0103] Please see Figure 6 , Figure 6 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 601 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this disclosure. The memory 602 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 602 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 602 and is called and executed by the processor 601 to execute the automatic product selection and replenishment method of the smart cabin according to the embodiments of this disclosure. The input / output interface 603 is used to implement information input and output; The communication interface 604 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 605 transmits information between various components of the device (e.g., processor 601, memory 602, input / output interface 603, and communication interface 604); The processor 601, memory 602, input / output interface 603, and communication interface 604 are connected to each other within the device via bus 605.

[0104] This disclosure also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described automatic product selection and replenishment method for the smart cabin.

[0105] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0106] The embodiments described in this disclosure are for the purpose of more clearly illustrating the technical solutions of this disclosure and do not constitute a limitation on the technical solutions provided by this disclosure. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by this disclosure are also applicable to similar technical problems.

[0107] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this disclosure, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0108] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0109] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0110] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in this disclosure and the foregoing drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such orders can be interchanged where appropriate so that embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion, for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0111] It should be understood that in this disclosure, "at least one item" means one or more, and "more than one" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0112] In the embodiments provided in this disclosure, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and the division of the units described above is only a logical functional division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or units may be electrical, mechanical, or other forms.

[0113] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0114] Furthermore, the functional units in the various embodiments of this disclosure can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0115] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0116] The preferred embodiments of the present disclosure have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present disclosure. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and spirit of the present disclosure shall be within the scope of the claims of the present disclosure.

Claims

1. A method for automatic product selection and replenishment in an intelligent warehouse, characterized in that, The system is applied to an automated product selection and replenishment system, which includes an intelligent cabin and multiple embodied robots. The intelligent cabin includes a cabin body that encloses an interior space. Several sales windows are opened on the walls of the cabin body, and display devices are installed on the walls of the cabin body. Several shelves are installed inside the cabin body, and the shelves correspond to the positions of the sales windows. The automatic product selection and replenishment method includes: Obtain inventory data for each shelf, sales data for each sales window, and the current operating mode of the smart cabin; The corresponding product selection strategy is determined according to the working mode, and a replenishment plan is generated based on the inventory data, the sales data and the product selection strategy. The replenishment plan contains several replenishment tasks. The priority of each replenishment task is determined according to the working mode, and the corresponding embodied robot is scheduled to execute the replenishment task in sequence according to the replenishment plan and the priority of the replenishment task. During the process of the robot performing the replenishment task, the display device is controlled to display the replenishment progress information corresponding to the replenishment task, and the inventory data is updated after each replenishment task is completed.

2. The automatic product selection and replenishment method for the intelligent cabin according to claim 1, characterized in that, The step of determining the corresponding product selection strategy based on the working mode, and generating a replenishment plan based on the inventory data, the sales data, and the product selection strategy, includes: The selection weight of each product category is determined based on the aforementioned working mode; The replenishment plan is generated by optimizing the solution based on the inventory data, sales data, and product selection weights, with the objective function being to maximize the product weighted gross profit multiplied by the demand forecast, and the constraints being that the product category richness is not lower than the preset richness lower bound and the inventory of each shelf does not exceed the corresponding arc capacity.

3. The automatic product selection and replenishment method for the intelligent cabin according to claim 2, characterized in that, The working mode includes at least one of performance mode, hybrid mode and retail mode; The step of determining the selection weight of each product category according to the working mode includes: When the working mode is the performance mode or the hybrid mode, increase the selection weight of traffic-driving product categories; When the working mode is the retail mode, increase the selection weight of high-turnover product categories.

4. The automatic product selection and replenishment method for the intelligent cabin according to claim 1, characterized in that, The step of determining the priority of each replenishment task according to the working mode, and scheduling the corresponding embodied robot to execute the replenishment task sequentially according to the replenishment plan and the priority of the replenishment task, includes: Select at least one of the multiple locations within the cabin space as the target location, wherein the target location is the cabin location for reserving space for manual replenishment operations; In the remaining multiple cabin locations other than the target location, the priority of each replenishment task is determined according to the working mode, and the corresponding embodied robot is scheduled to execute the replenishment task in sequence according to the replenishment plan and the priority of the replenishment task.

5. The automatic product selection and replenishment method for the intelligent cabin according to claim 1 or 4, characterized in that, The cabin is provided with multiple arc segments, each of the shelves is respectively set in the corresponding arc segment, and each of the embodied robots is respectively set in relation to each of the arc segments; The step of determining the priority of each replenishment task according to the working mode, and scheduling the corresponding embodied robot to execute the replenishment task sequentially according to the replenishment plan and the priority of the replenishment task, includes: When the working mode is retail mode, the priority of the replenishment task is determined to be lower than the priority of the order picking task, and the embodied robot corresponding to the arc segment where the target shelf is located is scheduled to perform the replenishment task. When the working mode is performance mode or hybrid mode, the priority of the replenishment task is determined to be lower than that of the performance task, and the embodied robot executing the replenishment task is controlled to perform the replenishment task across arc segments.

6. The automatic product selection and replenishment method for the intelligent cabin according to claim 5, characterized in that, The automatic product selection and replenishment method also includes: During the execution of the replenishment task, the order service level agreement status of each sales window is monitored in real time; When the Service Level Agreement (SLA) of any of the sales windows is triggered, the currently executing replenishment task is paused or canceled, and the avatar robot executing the replenishment task is switched to execute the order picking task of the corresponding sales window.

7. The automatic product selection and replenishment method for the intelligent cabin according to claim 1, characterized in that, The automatic product selection and replenishment method also includes: Establish a binding relationship between the products displayed on the display device and the shelves and the sales window. The binding relationship includes at least the product identifier, the corresponding sales window identifier, the corresponding shelf identifier, and the corresponding available inventory quantity. Before the display device displays the highlight content of the target product, the available inventory quantity of the target product is verified to be greater than zero according to the binding relationship; When the available inventory quantity is greater than zero, the highlight display content is played through the display device; when the available inventory quantity is zero, the display device is controlled to cancel the display of the highlight display content and switch to the display of alternative products with a available inventory quantity greater than zero under the same sales window.

8. The automatic product selection and replenishment method for the intelligent cabin according to claim 1, characterized in that, During the process of the robot performing the replenishment task, controlling the display device to display replenishment progress information corresponding to the replenishment task includes: Determine the target arc segment where the target shelf for which the embodied robot will perform the replenishment task; In the display area corresponding to the target arc segment on the display device, a replenishment progress overlay is rendered, and the replenishment progress overlay includes at least a replenishment status indicator and product information that is being replenished; After the replenishment task is completed, the replenishment progress overlay is cleared.

9. The automatic product selection and replenishment method for the intelligent cabin according to claim 1, characterized in that, An outdoor customer area is provided around the perimeter of the cabin; The automatic product selection and replenishment method also includes: Collect interaction intent information of visitors in the outdoor customer area, including voice interaction intent and / or gesture interaction intent; Based on the visitor's circumferential position within the outdoor customer area, determine the target sales window that is closest to the visitor from each of the sales windows; The corresponding interactive response result is generated based on the interactive intent information, and the interactive response result is routed to the product interface of the target sales window.

10. A smart cabin, characterized in that, include: The cabin is enclosed to form an interior space. Several sales windows are opened on the walls of the cabin, and display devices are installed on the walls of the cabin. Several shelves are installed inside the cabin, and the shelves are positioned corresponding to the sales window; The control unit is used to acquire inventory data of each shelf, sales data of each sales window, and the current working mode of the smart cabin. The system determines a corresponding product selection strategy based on the working mode, and generates a replenishment plan based on the inventory data, sales data, and product selection strategy. The replenishment plan contains several replenishment tasks. The system determines the priority of each replenishment task based on the working mode, and schedules corresponding embodied robots to execute the replenishment tasks sequentially based on the replenishment plan and the priority of the replenishment tasks. During the execution of the replenishment tasks by the embodied robots, the system controls the display device to display the replenishment progress information corresponding to the replenishment tasks, and updates the inventory data after each replenishment task is completed.

11. A body-worn robot, characterized in that, include A receiving module is used to receive replenishment tasks sent by the smart cabin, wherein the replenishment tasks are generated by the automatic product selection and replenishment of the smart cabin as described in claims 1 to 9. The work control module is used to perform corresponding operations based on the replenishment task.

Images