Control method, device, medium and product for bank outlets

Abstract

This application provides a control method, device, medium, and product for bank branches, relating to the field of artificial intelligence technology. The method is applied to a quadruped robot management platform and includes: receiving multimodal data collected by the quadruped robot within the bank branch; generating control instructions based on the multimodal data and preset task configuration information; sending the control instructions to the quadruped robot, enabling the quadruped robot to execute target tasks according to the control instructions and obtain task execution results; receiving the task execution results; and generating corresponding early warning information when the task execution results indicate the presence of illegally parked vehicles, absent staff, malfunctioning self-service terminals, or failed delivery handover, and sending the early warning information to a preset notification terminal. This method enables integrated collaborative control of bank branches in scenarios such as customer service, security patrols, and internal management, improving service response speed, security automation levels, and reducing manual management costs.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a control method, device, medium and product for bank branches. Background Technology

[0002] With the rapid development of artificial intelligence, computer vision, and embodied intelligent robot technologies, bank branches are accelerating their transformation towards intelligence and automation.

[0003] In existing technologies, bank branch management systems generally adopt a modular architecture, with each functional module operating independently. For example, traditional monitoring systems can only passively record video data, relying on manual analysis to identify abnormal behavior in real time (such as illegal parking of armored trucks in the armored vehicle lane, absence of on-duty personnel, etc.). Guided tours are mostly provided by static signage or robots offering fixed-route explanations. At night or during non-business hours, manual patrols by security guards are essential.

[0004] Therefore, existing bank branches suffer from slow service response times, low levels of security automation, and high manual management costs. Summary of the Invention

[0005] This application provides a control method, device, medium, and product for bank branches, which enables integrated collaborative control of bank branches in scenarios such as customer service, security inspection, and internal management through a quadruped robot management platform that integrates quadruped robots with multimodal data. This improves service response speed, security automation level, and reduces manual management costs.

[0006] In a first aspect, this application provides a control method for bank branches, applied to a quadruped robot management platform, the method comprising:

[0007] It receives multimodal data collected by the quadruped robot within the bank branch. The multimodal data includes video images, infrared temperature measurement data, and voice interaction content.

[0008] Control commands are generated based on multimodal data and preset task configuration information. The preset task configuration information includes at least one of the following: armored vehicle lane inspection route, guide point list, material delivery address, and duty post location.

[0009] Control commands are sent to the quadruped robot so that the quadruped robot can execute the target task according to the control commands, obtain the task execution result, and send the task execution result to the quadruped robot management platform. The target task includes at least one of the following: vehicle illegal parking detection, intelligent navigation, material distribution, personnel on-duty verification, and self-service terminal equipment status inspection.

[0010] Receive task execution results;

[0011] When the task execution result indicates that there are illegally parked vehicles, absent duty personnel, abnormal self-service terminal equipment, or delivery handover failure, a corresponding warning message is generated and sent to a preset notification terminal. The preset notification terminal includes at least one of the following: administrator email, property management system, and mobile terminal.

[0012] Secondly, this application provides a control method for bank branches, applied to a quadruped robot, the method comprising:

[0013] Multimodal data is collected from the target area within the bank branch using multimodal sensors. The multimodal data includes video images, infrared temperature measurement data, and voice interaction content.

[0014] Multimodal data is sent to the quadruped robot management platform, which then generates control commands based on the multimodal data and preset task configuration information, and sends the control commands to the quadruped robot. The preset task configuration information includes at least one of the following: armored truck channel inspection route, guide point list, material delivery address, and duty post location.

[0015] Receive control commands sent by the quadruped robot management platform;

[0016] The system executes the target task according to the control command, obtains the task execution result, and sends the task execution result to the quadruped robot management platform. The target task includes at least one of the following: vehicle illegal parking detection, intelligent navigation, material delivery, personnel on-duty verification, and self-service terminal equipment status inspection.

[0017] Thirdly, this application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0018] The memory stores instructions that the computer executes;

[0019] The processor executes computer execution instructions stored in memory to implement a control method for bank branches that includes the contents of the first or second aspect of the invention.

[0020] Fourthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement a control method for bank branches that implements the contents of the first or second aspect of the invention.

[0021] Fifthly, this application provides a computer program product, including a computer program, which, when executed by a processor, is used to implement a control method for bank branches that is based on the first or second aspect of the invention.

[0022] This application provides a control method, device, medium, and product for bank branches, comprising: first, receiving multimodal data collected by a quadruped robot within the bank branch; then, generating control instructions based on the multimodal data and preset task configuration information; subsequently, sending the control instructions to the quadruped robot, enabling the quadruped robot to execute the target task according to the control instructions, obtaining the task execution result, and sending the task execution result to the quadruped robot management platform; then, receiving the task execution result; finally, when the task execution result indicates the presence of illegally parked vehicles, absent staff, malfunctioning self-service terminal equipment, or failed delivery handover, generating corresponding early warning information and sending the early warning information to a preset notification terminal. This achieves the following technical effects: by incorporating heterogeneous tasks such as vehicle illegal parking detection, intelligent navigation, material delivery, staff on-duty verification, and self-service terminal inspection into the same control framework, and with the quadruped robot management platform dynamically generating control instructions based on multimodal data and receiving task execution results for feedback, the problem of isolated operation and data fragmentation of traditional bank subsystems is solved, significantly improving overall operational collaboration efficiency. Quadruped robots autonomously patrol key areas such as armored vehicle lanes. Combined with video image recognition of illegally parked vehicles, the quadruped robot management platform automatically triggers vehicle relocation notifications. This transforms passive monitoring into proactive intervention, significantly shortening response time to illegal parking and greatly improving the automation level of bank security. By interpreting user intent through voice interaction, quadruped robots can dynamically guide users to corresponding service points, avoiding blind queuing or consulting lobby managers, reducing user waiting time, improving user experience, and freeing up manpower. In material distribution and personnel on-duty verification, multiple verification mechanisms such as dynamic passwords, employee badge recognition, and facial recognition are introduced to automatically determine whether handover is successful or whether a position is vacant, eliminating human manipulation issues such as proxy sign-ins and false deliveries, achieving digital traceability and audit tracking of key processes. Quadruped robots perform inspection tasks around the clock, covering areas inaccessible to wheeled robots such as staircases and ramps, replacing manual nighttime security patrols and effectively saving labor costs. Simultaneously, abnormal events automatically generate corresponding early warning information and send it to email, property management systems, or mobile applications, effectively reducing missed and false reports and improving management efficiency. Attached Figure Description

[0023] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0024] Figure 1 This is a schematic diagram of a system architecture for a control method for bank branches, provided as an embodiment of this application.

[0025] Figure 2 A flowchart illustrating a control method for bank branches provided in this application embodiment. Figure 1 .

[0026] Figure 3 A flowchart illustrating a control method for bank branches provided in this application embodiment. Figure 2 .

[0027] Figure 4 A flowchart illustrating a control method for bank branches provided in this application embodiment. Figure 3 .

[0028] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0029] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0030] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0031] In the embodiments of this application, the terms "first" and "second" are used to distinguish identical or similar items with substantially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that "first" and "second" do not necessarily imply difference. It should be noted that in the embodiments of this application, the words "exemplary" or "for example" are used to indicate that something is being used as an example, illustration, or description. Any embodiment or design scheme described as "exemplary" or "for example" in this application should not be construed as being better or more advantageous than other embodiments or design schemes. Specifically, the use of "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner. In the embodiments of this application, "at least one" refers to one or more, and "more than one" refers to two or more.

[0032] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in one or more embodiments of this specification are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of related data must comply with relevant laws, regulations and standards, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0033] It should be noted that the phrase "at...time" in the embodiments of this application can refer to the instant at which a certain situation occurs, or to a period of time after the occurrence of a certain situation; the embodiments of this application do not specifically limit this. Furthermore, the control method for bank branches provided in the embodiments of this application is merely an example; a control method for bank branches may include more or fewer elements.

[0034] It should be noted that the control methods, equipment, media and products for bank branches provided in this application can be used in the field of artificial intelligence technology, or in any field other than artificial intelligence technology. The application fields of the control methods, equipment, media and products for bank branches in this application are not limited.

[0035] With the rapid development of artificial intelligence, computer vision, and embodied intelligent robot technology, bank branches are moving towards intelligence and unmanned operation at an unprecedented speed.

[0036] Under the existing technological architecture, bank branch management systems generally adopt a modular design concept, with each functional module operating independently and data fragmented, making efficient collaboration difficult. Taking traditional monitoring systems as an example, they only have the function of passively recording video data. For real-time identification of abnormal behavior (such as illegal parking of armored trucks in lanes, verification of staff attendance, etc.), they still rely on manual review or on-site inspections, and are highly dependent on manual analysis, resulting in low efficiency and a high risk of oversight. Regarding guided tours, static signage or fixed-route guided robots are mostly used, lacking the ability to perceive and respond to users' real-time needs, and unable to provide flexible and personalized guidance based on actual user requirements. At night or outside of business hours, bank branch security mainly relies on manual security patrols, which suffers from blind spots, delayed response, and high labor costs. This not only consumes a large amount of manpower, but the effectiveness of patrols is also greatly affected by factors such as staff condition and experience.

[0037] It is evident that bank branches currently face numerous challenges in their actual operations. Service response speed is insufficient to meet the growing demands of users, the level of security automation needs improvement, and the cost of manual management remains high. These problems severely restrict the further development of bank branches.

[0038] Based on this, this application proposes a control method, device, medium, and product for bank branches, applicable to the field of artificial intelligence technology, aiming to solve the aforementioned technical problems of the prior art. By using a quadruped robot as a mobile sensing terminal, combined with a quadruped robot management platform supporting multimodal data fusion and intelligent decision-making, an integrated collaborative control system covering user service, security inspection, and internal management is constructed. This enables proactive perception of the bank branch environment, dynamic task scheduling, and closed-loop handling of abnormal events, thereby significantly improving service response speed, enhancing security automation levels, and effectively reducing manual management costs.

[0039] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0040] To better understand the solutions of the embodiments of this application, a system architecture involved in the embodiments of this application will be introduced below.

[0041] Please refer to the specific system architecture of this application. Figure 1 . Figure 1 This is a schematic diagram of a system architecture for a control method used in bank branches, provided as an embodiment of this application. It should be noted that... Figure 1 The examples shown are merely examples of system architectures that can be applied to the embodiments of this application, in order to help those skilled in the art understand the technical content of this application, but do not mean that the embodiments of this application cannot be used in other devices, systems, environments or scenarios.

[0042] like Figure 1 As shown, the system architecture includes a quadruped robot 110 and a quadruped robot management platform 120. The quadruped robot management platform 120 is communicatively connected to the quadruped robot 110. The quadruped robot 110, acting as a mobile sensing terminal, collects multimodal data from target areas within the bank branch using multimodal sensors and transmits this data to the quadruped robot management platform 120. The quadruped robot management platform 120 generates control commands based on the multimodal data and preset task configuration information to control the quadruped robot to perform tasks such as vehicle illegal parking detection, intelligent navigation, material delivery, personnel on-duty verification, and self-service terminal equipment status inspection. This enables integrated collaborative control of bank branches in scenarios such as customer service, security inspection, and internal management, improving service response speed, security automation levels, and reducing manual management costs.

[0043] Figure 2 A flowchart illustrating a control method for bank branches provided in this application embodiment. Figure 1 .like Figure 2 As shown, this method is applied to a quadruped robot management platform, including:

[0044] S201, Receive multimodal data collected by the quadruped robot within the bank branch.

[0045] In this embodiment, the execution entity of a control method for bank branches is a quadruped robot management platform. This quadruped robot management platform can be implemented using a data processing server or other devices with data processing capabilities, such as laptops, personal computers, or tablets. The data processing server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server, etc., without specific limitations.

[0046] In addition, multimodal data includes video images, infrared temperature measurement data, and voice interaction content.

[0047] Specifically, the quadruped robot management platform can be deployed on a bank's local server or cloud data center to communicate with one or more quadruped robots in real time.

[0048] Quadruped robots are embodied intelligent agents capable of all-terrain mobility, equipped with multimodal sensors and edge computing units. For example, they may be equipped with high-definition red-green-blue (RGB) cameras, infrared thermal imagers, microphone arrays, speakers, and edge computing modules.

[0049] A quadruped robot management platform may include a multimodal data receiving module, a task scheduling engine, an artificial intelligence (AI) analysis module, an early warning generation module, and a notification interface.

[0050] During the process of performing tasks in bank branches, the quadruped robot can collect and upload three types of multimodal data in real time to the quadruped robot management platform: video images, infrared temperature measurement data, and voice interaction content.

[0051] The video images can be captured by a high-definition RGB camera on top of the quadruped robot. The coverage area can include: armored vehicle lanes, VIP waiting areas, lobby pathways, and self-service areas. Self-service areas can be Automated Teller Machines (ATMs) or Cash Recycler Systems (CRS). The image data can carry metadata such as timestamps, location coordinates, lens focal length, and light intensity.

[0052] Infrared temperature measurement data can be acquired by an integrated infrared thermal imager. Key monitoring areas can include critical positions such as duty rooms, vault entrances, and night shift posts. The quadruped robot management platform can preset the human body heat source detection threshold to 35 to 37 degrees Celsius. Temperatures below 34 degrees Celsius for more than 5 minutes are considered as no one on duty.

[0053] Voice interaction can capture ambient speech using a microphone array (supporting beamforming and noise reduction). Users can choose to upload the original audio stream or a transcribed version of the original audio, such as: "Hello, Xiaozhi, how do I get to the cash counter?"

[0054] Quadruped robots can perform movement tasks within bank branches according to preset strategies or real-time commands. For example, they can autonomously move along the inspection route of the armored truck 30 minutes before its expected arrival; during peak business hours, they can proactively patrol the customer waiting area and monitor voice interaction content; and during off-peak hours at night, they can conduct infrared temperature measurement and video inspection of the self-service area.

[0055] S202. Generate control commands based on multimodal data and preset task configuration information.

[0056] In this embodiment of the application, the preset task configuration information includes at least one of the following: armored vehicle channel inspection route, guide point list, material delivery address, and duty post location.

[0057] Specifically, after receiving multimodal data, the quadruped robot management platform can dynamically generate control commands by combining preset task configuration information. This preset task configuration information can be stored in the quadruped robot management platform's database, and includes, but is not limited to, the armored vehicle's inspection route, a list of guide points, material delivery addresses, and the location of duty posts.

[0058] The armored truck inspection route refers to a closed-loop path consisting of multiple geofenced points. The list of guide points can be: ATM - Corporate Banking Counter - Cash Counter. The material delivery address can be: Cash Sorting Center - Safe Deposit Center. The duty station location can be: Lobby Manager - Night Shift.

[0059] The specific logic for generating control commands can be as follows: If a vehicle is detected lingering in the armored truck lane for more than 2 minutes and its license plate is not on the armored truck whitelist, a vehicle illegal parking detection command is generated. If the user's intent is identified as wanting to withdraw cash, a command to guide them to the nearest ATM is generated based on the navigation point list. If the current time is 09:00 AM and infrared data shows no heat source at the lobby manager's post, a personnel on-duty verification command is generated. Upon receiving a delivery key file request, navigation and handover verification commands are generated based on the material delivery address.

[0060] The generated control commands can include target coordinates, task type, execution parameters (such as verification method), timeout threshold, etc., and can be sent to the designated quadruped robot through an encrypted channel.

[0061] S203. Send control commands to the quadruped robot so that the quadruped robot can execute the target task according to the control commands, obtain the task execution result, and send the task execution result to the quadruped robot management platform.

[0062] In this embodiment of the application, the target task includes at least one of vehicle illegal parking detection, intelligent navigation, material delivery, personnel on-duty verification, and self-service terminal equipment status inspection.

[0063] Specifically, after receiving control commands, the quadruped robot can execute corresponding target tasks and send the structured task execution results back to the quadruped robot management platform. For example: for vehicle detection in armored truck lanes, it can upload images of the parked vehicles, license plate recognition results, and duration of stay. For intelligent navigation, it can upload confirmation signals of users arriving at their destination and service satisfaction ratings. For material delivery, it can upload video clips of the handover process and recipient identity verification results (success / failure). For personnel on-duty verification, it can upload infrared heat maps of the work area and facial recognition comparison results. For self-service terminal equipment status inspection, it can upload close-up images of ATM card readers and cash dispensers for the quadruped robot management platform to detect foreign objects or damage.

[0064] S204. Receive the task execution result.

[0065] Specifically, the quadruped robot management platform can receive the task execution results sent by the quadruped robot and automatically judge the task execution results according to preset rules.

[0066] S205. When the task execution result indicates that there are illegally parked vehicles, absent duty personnel, abnormal self-service terminal equipment, or delivery handover failure, generate corresponding early warning information and send the early warning information to the preset notification terminal.

[0067] In this embodiment of the application, the preset notification terminal includes at least one of the following: administrator email, property management system, and mobile terminal.

[0068] Specifically, if the license plate of the vehicle parked in the armored vehicle passage area is not in the armored vehicle white list and the parking time exceeds the preset time threshold, corresponding warning information can be generated. For example, if the vehicle illegal parking detection result shows "License plate: Beijing A12345, parking > 10 minutes", it is determined as an illegally parked vehicle, and corresponding warning information can be generated. If the personnel on-duty verification result shows "no heat source at the post and no face detected", it is determined that the duty personnel are absent. If the self-service terminal inspection image shows "the card reader slot is blocked by tape", it is determined that the self-service terminal device is abnormal; if the material distribution verification result is "the dynamic password is incorrect 3 times" or "the work permit cannot be recognized", it is determined that the distribution transfer fails.

[0069] Once it is determined that any task execution result is abnormal, the quadruped robot management platform can immediately generate structured warning information (including time, location, abnormal type, evidence attachment), and send it to at least one preset notification terminal through the notification interface. Among them, the preset notification terminals include but are not limited to the administrator's email, property management system, and mobile terminal, etc.

[0070] For the administrator's email, an HTML (HyperText Markup Language) email containing screenshots can be sent. For the property management system, a high-priority work order can be created through the API (Application Programming Interface). For the mobile terminal, a pop-up warning can be pushed to the security supervisor's mobile application (Application, APP), and one-key linkage with the vehicle barrier or access control is supported.

[0071] A control method for a bank branch provided in this embodiment is applied to a quadruped robot management platform. The method includes: First, receiving multi-modal data collected by the quadruped robot in the bank branch; then, generating a control instruction according to the multi-modal data and preset task configuration information; subsequently, sending the control instruction to the quadruped robot so that the quadruped robot executes the target task according to the control instruction, obtains the task execution result, and sends the task execution result to the quadruped robot management platform; then, receiving the task execution result; finally, when the task execution result indicates the existence of illegally parked vehicles, absent duty personnel, abnormal self-service terminal devices, or failed distribution transfers, generating corresponding warning information and sending the warning information to the preset notification terminal.

[0072] The following technical effects have been achieved: By integrating heterogeneous tasks such as vehicle illegal parking detection, intelligent navigation, material delivery, personnel on-duty verification, and self-service terminal inspection into a unified control framework, the quadruped robot management platform dynamically generates control commands based on multimodal data and receives feedback on task execution results. This solves the problems of isolated operation and data fragmentation among traditional bank subsystems (monitoring, access control, customer service, and inspection), significantly improving overall operational efficiency. Quadruped robots autonomously inspect key areas such as armored vehicle lanes, and combined with video image recognition of illegally parked vehicles, the quadruped robot management platform can automatically trigger vehicle relocation notifications. This transforms passive monitoring into proactive intervention, significantly shortening response time for illegal parking and greatly improving the bank's security automation level. By interpreting user intent through voice interaction, the quadruped robot can dynamically guide users to corresponding service points, avoiding blindly queuing or consulting lobby managers, reducing user waiting time, improving user experience, and freeing up manpower to focus on high-value business. In sensitive processes such as material distribution and personnel on-duty verification, multiple verification mechanisms, including dynamic passwords, employee badge recognition, and facial recognition, are introduced to automatically determine whether handover is successful or whether a position is vacant, eliminating human error such as proxy sign-ins and fraudulent deliveries, and achieving digital traceability and audit tracking of key processes. Quadruped robots perform inspection tasks around the clock, covering areas inaccessible to wheeled robots such as staircases and ramps, replacing manual nighttime security patrols and effectively saving labor costs. Simultaneously, abnormal events automatically generate corresponding warning information and send it to email, property management systems, or mobile applications, effectively reducing missed and false alarms and improving management efficiency. Pre-configured task information (such as inspection routes, guide points, and delivery addresses) can be dynamically updated remotely without modifying the quadruped robot's underlying code, allowing banks to quickly deploy customized strategies based on different branch layouts, business priorities, or security levels, demonstrating good scalability and adaptability.

[0073] In one possible implementation, control instructions are generated based on multimodal data and preset task configuration information, including: recognizing the scene status within the bank branch based on video images, including the illegal parking status of armored trucks, the operational status of self-service terminal equipment, or the activity status of personnel in public areas; parsing the user's service request intent based on voice interaction content; determining the on-duty status of personnel based on infrared temperature measurement data; and generating control instructions based on the scene status, service request intent, on-duty status of personnel, and preset task configuration information.

[0074] Specifically, after receiving real-time video streams or keyframe images captured by the high-definition camera mounted on the quadruped robot, the quadruped robot management platform can analyze the video images using a computer vision model deployed on the platform.

[0075] Specifically, for cases of illegal parking in armored vehicle lanes, a target detection model can be used to identify whether a vehicle exists within the lane area. If a vehicle is detected, the license plate number can be further extracted using a license plate recognition subnetwork and compared with a pre-stored whitelist of armored vehicles. If the vehicle is not on the whitelist and its stay exceeds a preset threshold (e.g., 2 minutes), it is determined to be illegally parked.

[0076] Regarding the operational status of self-service terminal equipment, image segmentation and anomaly detection can be performed on key parts of equipment such as ATMs / CRS (e.g., card reader slots, cash dispensers, and operation screens) to determine whether there are abnormal situations such as foreign object obstruction, physical damage, or prolonged inactivity.

[0077] To assess the activity status of people in public areas, human detection and behavior analysis models (such as the SlowFast motion recognition network) can be used to identify suspicious or dangerous behaviors such as prolonged loitering, gathering, or falling.

[0078] Secondly, after receiving the voice interaction content uploaded by the quadruped robot, the quadruped robot management platform can perform intent classification and slot extraction on the text. For example, if the user's voice is "Hello, Xiaozhi, how do I get to the cash counter?", it can be parsed as the intent type "intelligent navigation", and the slot information includes "target service: cash business".

[0079] If a user's voice message is "I want to complain about the counter staff from last time," it can be interpreted as an intent type "complaint / suggestion" and can trigger the reporting process.

[0080] Furthermore, after receiving the thermal image data collected by the quadruped robot's infrared thermal imager, the quadruped robot management platform can combine the preset job location coordinates (such as "Lobby Manager Post" and "Night Watch Post") and perform the following logical judgment: If no heat source within the human body temperature range (35 degrees Celsius to 37 degrees Celsius) is detected in the specific job area within the preset time period (such as 09:00 to 17:00 on weekdays), it is initially determined that "no one is on duty".

[0081] To further verify this, the quadruped robot management platform can issue instructions to the quadruped robot to go to the designated position to capture facial images and compare them with the employee identity database using a facial recognition model (such as ArcFace). If no match is found or no face is detected, it is ultimately confirmed that the staff member is absent from their post.

[0082] Finally, by combining the above multi-source information with the preset task configuration information, the quadruped robot management platform can dynamically generate control commands.

[0083] Specifically, the quadruped robot management platform can integrate identified scene states (such as illegal parking in the armored truck lane), service request intentions (such as guiding to the VIP room), and staff on-duty status (such as the lobby manager being absent) with preset task configuration information (including armored truck lane inspection routes, guide point lists, material delivery addresses, and duty station locations). For example, when an armored truck is detected to be approaching (timestamp from the bank's dispatch system) and there is an illegally parked vehicle in the lane, an instruction to initiate the illegal parking handling process is generated, including a secondary inspection plan and a gate linkage strategy. When a VIP user is detected to be visiting and the user manager is on duty, an instruction to guide the user to the VIP room is generated and the user manager is notified to prepare for reception is generated. When an abnormality is detected at a self-service terminal and there are no users currently queuing, an instruction to proceed to ATM-03 for equipment status verification is generated.

[0084] The control commands can be encapsulated in a structured data format (such as JSON), including task type, target coordinates, execution parameters, priority and timeout threshold, and sent to the corresponding quadruped robot through a secure communication channel.

[0085] This application's embodiments, by integrating multimodal data such as video, voice, and infrared temperature measurement, achieve accurate perception of the bank branch scene status, user intent, and staff on-duty status, significantly improving the accuracy of illegal parking identification, equipment anomaly detection, and identity verification. Based on multi-source information and preset task configurations, control commands are dynamically generated, enabling the quadruped robot to shift from passive execution to proactive service and intelligent inspection, greatly shortening the response time to abnormal events and improving task completion rate and user experience. Simultaneously, by automatically completing security inspections, on-duty verification, and equipment monitoring, manual intervention is reduced, enhancing the level of automation.

[0086] The above methods have enabled an end-to-end closed loop from multimodal perception to intelligent task generation, significantly improving the automated response capabilities and service accuracy of bank branches.

[0087] In one possible implementation, vehicle illegal parking detection includes: receiving a first image of the armored truck lane area sent by a quadruped robot during its initial inspection; identifying whether there are illegally parked vehicles in the armored truck lane area based on the first image; and if there are illegally parked vehicles, generating a vehicle relocation notification and sending it to a preset notification terminal.

[0088] Specifically, the quadruped robot management platform can receive the first image of the armored truck lane area captured and uploaded by the quadruped robot's onboard high-definition camera during its initial inspection as it travels along the preset inspection route. This first image can include metadata such as timestamps, geographic location information, and lighting conditions to ensure the completeness of the analysis information.

[0089] Next, the quadruped robot management platform can call a computer vision model to analyze the first image. First, it uses an object detection algorithm (such as YOLOv8) to identify whether a vehicle exists in the image. If a vehicle is detected, the license plate number is further extracted using a license plate recognition sub-model. Subsequently, the identified license plate is compared with the bank's pre-stored whitelist of armored trucks.

[0090] If a vehicle is not on the armored vehicle whitelist and its stay in the passage area exceeds a preset threshold (e.g., 2 minutes), it is determined to be an illegally parked vehicle. In this case, the quadruped robot management platform can automatically generate a structured vehicle relocation notification, including an image of the illegally parked vehicle, its license plate number, the location of the violation, the time, and a suggested action. This notification is then pushed to at least one preset notification terminal via a secure communication interface. These preset notification terminals include, but are not limited to, the branch administrator's email address, the property management system's work order module, or the security personnel's mobile app, to trigger manual intervention or an automated handling process.

[0091] This application embodiment utilizes a quadruped robot to automatically collect images of armored vehicle lanes and combines this with intelligent recognition algorithms to identify illegally parked vehicles, replacing manual inspections and significantly improving detection efficiency and accuracy. Once an illegal parking violation is detected, a vehicle relocation notification is immediately generated and pushed to the administrator's email, property management system, or mobile terminal, effectively ensuring the safe and smooth operation of armored vehicles, reducing security manpower costs, and improving the level of automated control in key areas of bank branches.

[0092] In one possible implementation, after generating a vehicle relocation notification and sending it to a preset notification terminal, the method further includes: receiving a second image of the armored vehicle lane area sent by the quadruped robot during a secondary inspection; confirming, based on the second image, whether the illegally parked vehicle has still not been moved; if it is confirmed that the illegally parked vehicle has still not been moved, adding the license plate information of the illegally parked vehicle to a preset no-parking list, and sending a gate control command to the entrance and exit gate system of the bank branch to prevent the illegally parked vehicle from entering the bank branch again.

[0093] Specifically, after generating a vehicle relocation notification and sending it to a preset notification terminal, the quadruped robot management platform can initiate a secondary verification process after the initial alarm and issue a secondary inspection command to the quadruped robot. Based on this secondary inspection command, the quadruped robot can travel along the armored truck lane inspection route again at a preset time interval (e.g., after 10 minutes), collect a second image of the armored truck lane area, and upload it to the quadruped robot management platform.

[0094] The quadruped robot management platform can compare the second image with the first image from the initial inspection. It then uses a target detection model to re-identify the vehicle's location and license plate information, and makes a judgment based on spatiotemporal consistency. If a vehicle with the same license plate remains in its original location (within the armored truck lane area), it is confirmed that the illegally parked vehicle has not been moved.

[0095] Once it is confirmed that an illegally parked vehicle has not been moved, the quadruped robot management platform can automatically add the vehicle's license plate information to a preset no-parking list and generate a gate control command. This command is then sent to the bank branch's entrance and exit gate system via a standard API interface. Upon receiving the command, the gate system updates the bank branch's restricted access list, refusing entry to vehicles with that license plate for a set period (e.g., 7 or 30 days). This forms a closed-loop mechanism of identification, alarm, verification, and blocking, effectively addressing illegal parking and ensuring the long-term smooth and secure operation of cash transport channels.

[0096] This application's embodiment uses a secondary inspection mechanism to verify illegally parked vehicles, avoiding misjudgments and improving the reliability of handling. Once it is confirmed that the vehicle has not been moved, its license plate is automatically added to a preset no-parking list, and the entrance and exit gate system is linked to implement access control, forming a closed-loop handling mechanism of identification, alarm, and verification. This effectively solves the problem of obstructing cash transport channels, ensures the safety of financial escort, reduces manual intervention, and significantly improves the level of security automation and intelligent management in bank branches.

[0097] In one possible implementation, the execution result corresponding to the material delivery includes a delivery handover status, which is determined by the quadruped robot management platform based on the verification result of any of the following methods: receiving a dynamic password uploaded by the quadruped robot and verifying its validity; receiving an image of the recipient's name tag uploaded by the quadruped robot and comparing it with pre-stored identity information through image recognition; or receiving a facial image uploaded by the quadruped robot and comparing it with pre-stored identity information through facial recognition. When the verification result indicates that the identity does not match, the password is invalid, or no valid verification information is received, the delivery handover status is determined to be a failure, and an early warning message is generated.

[0098] Specifically, the execution results of material delivery include the delivery handover status, which can be determined by the quadruped robot management platform based on the recipient's identity verification results.

[0099] Specifically, after arriving at the preset delivery address, the quadruped robot can prompt the recipient to verify their identity in any of the following ways: enter a dynamic password, show their work badge, or face the camera for facial recognition, and upload the corresponding data to the quadruped robot management platform.

[0100] After receiving the verification data, the quadruped robot management platform can perform the following processing: If it is a dynamic password, it will be compared with the current valid password pool (which is dynamically generated by the quadruped robot management platform according to time or task, and the validity period is usually 2 to 5 minutes) to verify the timeliness and correctness of the dynamic password.

[0101] If the image is an employee badge, the name, employee number, and other information are extracted using Optical Character Recognition (OCR) technology and matched with pre-stored records in the bank's employee identity database.

[0102] If it is a face image, the face recognition engine (such as the ArcFace algorithm) is called to compare the feature vector with the face template library of the authorized recipient in a 1:1 or 1:N ratio (N is an integer greater than or equal to 2). The similarity must exceed a preset threshold (such as 0.75) to pass.

[0103] When any verification method returns an identity mismatch, expired / incorrect password, or no valid verification information is received within the timeout period, the quadruped robot management platform determines that the delivery handover status has failed and automatically generates structured early warning information, including the type of goods, delivery time, destination address, reason for verification failure, and on-site image evidence, and pushes it to the administrator's email, property management system, or the mobile terminal of the relevant responsible person to ensure that the delivery process of highly sensitive goods (such as keys, cash packages, and important documents) is traceable, auditable, and strongly controlled.

[0104] This application embodiment employs multiple identity verification mechanisms, including dynamic passwords, employee badge recognition, and facial recognition, to ensure that highly sensitive materials are received only by authorized personnel, significantly improving delivery security and compliance. The quadruped robot management platform automatically determines the handover status, preventing human error. If verification fails, an immediate warning is generated and pushed to the management terminal, achieving full traceability and auditability. This effectively prevents issues such as fraudulent claims and incorrect handovers, ensuring the secure and controllable flow of materials within bank branches.

[0105] Figure 3 A flowchart illustrating a control method for bank branches provided in this application embodiment. Figure 2 .like Figure 3 As shown, this method is applied to a quadruped robot, including:

[0106] S301. Collect multimodal data of the target area within the bank branch through a multimodal sensor. The multimodal data includes video images, infrared temperature measurement data, and voice interaction content.

[0107] Specifically, quadruped robots can use their onboard multimodal sensors to collect multimodal data in real time as they autonomously move within target areas of bank branches.

[0108] The multimodal sensors include: a high-definition RGB camera (for capturing video images at 1080P and above resolution), an infrared thermal imager (for acquiring ambient temperature distribution and human body heat source information), and a microphone array (for picking up voice interaction content and supporting noise reduction and sound source localization). The acquired multimodal data can be simultaneously accompanied by timestamps, indoor positioning coordinates, and task context identifiers.

[0109] S302. Send the multimodal data to the quadruped robot management platform so that the quadruped robot management platform can generate control commands based on the multimodal data and preset task configuration information, and send the control commands to the quadruped robot.

[0110] Specifically, the preset task configuration information includes at least one of the following: armored vehicle lane inspection route, guide point list, material delivery address, and duty post location.

[0111] Quadruped robots can upload multimodal data to the quadruped robot management platform via secure communication links (such as 5G private networks or encrypted Wi-Fi 6). The management platform can then perform intelligent analysis based on this multimodal data and preset task configuration information (including at least one of the following: armored vehicle channel inspection routes, guide point lists, material delivery addresses, and duty station locations), and generate corresponding control commands. These control commands can be encrypted and then sent by the quadruped robot management platform to the corresponding quadruped robot.

[0112] S303: Receive control commands sent by the quadruped robot management platform.

[0113] Specifically, after receiving control commands, the quadruped robot can further analyze the task type, target position, execution parameters, and verification requirements, and then call the underlying motion control, navigation planning, and interaction modules to execute the corresponding target task.

[0114] S304. Execute the target task according to the control instructions, obtain the task execution result, and send the task execution result to the quadruped robot management platform.

[0115] Specifically, the target tasks include at least one of the following: vehicle illegal parking detection, intelligent navigation, material distribution, personnel on-duty verification, and self-service terminal equipment status inspection.

[0116] Quadruped robots can: 1) patrol along the armored truck lane and capture images of designated areas during vehicle parking violation detection; 2) guide users to target service points and provide voice explanations during intelligent navigation; 3) navigate to recipient locations during material delivery, wait for identity verification, and open the storage compartment after confirming the recipient's identity; 4) collect infrared data or facial images before stopping at designated posts during personnel on-duty verification; and 5) take close-up photos of designated parts of ATMs / CRS systems during self-service terminal equipment status inspections.

[0117] After the task is completed, the quadruped robot can package the structured task execution results (such as images, verification feedback, status flags, etc.) and send them back to the quadruped robot management platform, so that the quadruped robot management platform can make anomaly judgments and generate early warnings, thereby completing the closed-loop collaboration of perception, execution and feedback.

[0118] This application provides a control method for bank branches. By using a quadruped robot as a mobile sensing and execution terminal, it deeply integrates multimodal data from video, infrared, and voice to achieve comprehensive perception of the bank branch environment. Through collaboration with a quadruped robot management platform, the quadruped robot can flexibly perform diverse tasks such as illegal parking detection, intelligent navigation, and material delivery, and transmit results in real time, forming a closed-loop control. It fully leverages the strong obstacle-crossing ability of the quadruped structure, covering complex areas such as stairs and ramps, significantly improving inspection coverage and service response efficiency, reducing reliance on manual inspections and fixed equipment, and enhancing the intelligent management level of bank branches.

[0119] In one possible implementation, vehicle illegal parking detection includes: During the initial inspection, the robot travels along the armored truck lane inspection route according to control commands and captures a first image of the armored truck lane area using its onboard camera. This first image is then sent to the quadruped robot management platform, which identifies whether an illegally parked vehicle exists in the armored truck lane area based on the first image. If an illegally parked vehicle is found, a vehicle relocation notification is generated and sent to a preset notification terminal. During the second inspection, the robot again travels along the armored truck lane inspection route according to control commands and captures a second image of the armored truck lane area using its camera. This second image is sent to the quadruped robot management platform, which confirms whether the illegally parked vehicle has still not been moved. If the illegally parked vehicle is still not moved, the license plate information of the illegally parked vehicle is added to a preset no-parking list, and a control command is sent to the bank branch's entrance and exit gate system to prevent the illegally parked vehicle from re-entering the bank branch.

[0120] Specifically, vehicle illegal parking detection can be divided into two stages: initial inspection and secondary inspection. In the initial inspection stage, the quadruped robot receives control commands from the quadruped robot management platform. These commands may include the inspection route and task priority for the armored vehicle lane. Based on its built-in Simultaneous Localization and Mapping (SLAM) navigation system, the quadruped robot autonomously travels along a preset path to the armored vehicle lane area. Using a high-definition camera mounted on its top (e.g., with a resolution of at least 1920×1080 pixels and support for low-light enhancement), it captures images of the designated area within the lane, obtaining a first image. This first image may include a timestamp, location information, and lighting information. The location information includes geographic coordinates obtained from the Global Positioning System (GPS) in outdoor environments, or local coordinates obtained from an indoor positioning system in indoor environments. The quadruped robot can upload the first image to the quadruped robot management platform in real time via an encrypted communication link (such as 5G or Wi-Fi 6). The quadruped robot management platform can perform vehicle detection and license plate recognition based on the first image. If it determines that an unauthorized vehicle is illegally parked, it will generate a vehicle relocation notice and push it to the administrator's email, property management system, or security personnel's mobile terminal.

[0121] During the secondary inspection phase, the quadruped robot management platform can automatically trigger a review process after the initial alarm and issue a secondary inspection command to the quadruped robot. The quadruped robot then travels along the same inspection route again, acquiring and uploading a second image of the armored vehicle lane area under similar viewing angles and lighting conditions. The quadruped robot management platform can perform a spatiotemporal consistency comparison between the second image and the first image. If the vehicle with the same license plate is still in the same position and has not moved, it is confirmed that the illegally parked vehicle has not been moved. At this time, the quadruped robot management platform can add the license plate information to a preset no-parking list and send a gate control command to the bank branch's entrance and exit gate system via a standard API interface. The command includes the license plate number and the prohibition period (e.g., 7 days). After receiving the gate control command, the entrance and exit gate system can update the bank branch's no-entry database and automatically block the vehicle from entering the area within the validity period, thereby achieving fully automated processing from detection, alarm, review, and blocking.

[0122] This application embodiment employs a combined initial and secondary inspection method to perform dual verification of vehicles illegally parked in armored vehicle lanes, effectively avoiding false alarms. The quadruped robot management platform automatically identifies and compares license plates based on first and second images, and coordinates with the bank branch's entrance and exit gate system to implement dynamic access control, forming a closed loop of detection, alarm, and blocking. This significantly enhances the security capabilities of armored vehicle lanes, reduces manual intervention, ensures efficient, smooth, and safe financial escort operations, and improves the bank branch's intelligent management capabilities for external vehicles.

[0123] In one possible implementation, the material delivery includes: upon arrival at the target location, prompting the recipient to enter a dynamic password, show their work badge, or have their face captured by a camera via voice prompts or screen display; receiving the dynamic password, or capturing the recipient's work badge image or facial image via the onboard camera; and sending the dynamic password, work badge image, or facial image to the quadruped robot management platform so that the quadruped robot management platform can verify the delivery handover status.

[0124] Specifically, in the task of material delivery, the quadruped robot can first autonomously navigate to the preset material delivery address (such as a cash sorting center, safe area, or designated office workstation) based on the received control commands. After arriving at the target location, the quadruped robot can guide the recipient to choose one of the following methods for identity verification through voice broadcast (e.g., "Hello, please enter a dynamic password, or show your employee ID, or face the camera for facial verification") and / or display text and graphics prompts on the display screen on its body: enter a one-time dynamic password pre-issued by the quadruped robot management platform on the quadruped robot's interactive interface (the validity period is usually 2 to 5 minutes); place the employee ID in the camera's field of view for the quadruped robot to capture the image; or face the quadruped robot's front camera to complete the facial image capture.

[0125] Subsequently, the quadruped robot can perform preliminary processing on the collected data through its edge computing module (such as image denoising, face alignment, and OCR preprocessing), and upload the dynamic password, work badge image or face image along with meta-information such as task identifier (ID), timestamp, and geographical location to the quadruped robot management platform through an encrypted communication channel.

[0126] After receiving the aforementioned verification data, the quadruped robot management platform can perform identity verification. Specifically, for dynamic passwords, it verifies whether they match the current valid password pool and have not expired; for employee badge images, it extracts the name and employee number using OCR technology and compares them with the employee database; for facial images, it calls the facial recognition model and performs similarity matching with the authorized recipient template library.

[0127] Based on the verification results, the quadruped robot management platform can determine the status of the delivery handover (success or failure) and send back a confirmation command. If the verification is successful, the quadruped robot can automatically open the storage compartment to complete the delivery of materials; if the verification fails, the compartment door remains locked and an alarm process is triggered. The entire process achieves identity authentication, full traceability, and contactless safe handover of materials.

[0128] This application embodiment uses both voice and screen prompts to guide recipients to easily complete identity verification via dynamic password, employee ID, or facial recognition, improving user-friendliness and applicability. The quadruped robot uploads verification data to the quadruped robot management platform in real time, achieving strong remote authentication and ensuring that materials are received only by authorized personnel. The entire process requires no human supervision, supports zero-contact, traceable, and secure handover, effectively preventing fraudulent claims and misdeliveries, and significantly improving the automation, security, and compliance of internal bank material distribution.

[0129] Figure 4 A flowchart illustrating a control method for bank branches provided in this application embodiment. Figure 3 .like Figure 4 As shown, the method includes:

[0130] S401, the quadruped robot collects multimodal data of the target area within the bank branch through multimodal sensors.

[0131] S402, the quadruped robot sends multimodal data to the quadruped robot management platform.

[0132] S403, the quadruped robot management platform receives multimodal data collected by the quadruped robot within the bank branch.

[0133] The S404 quadruped robot management platform generates control commands based on multimodal data and preset task configuration information.

[0134] S405, the quadruped robot management platform sends control commands to the quadruped robot.

[0135] S406, the quadruped robot receives control commands sent by the quadruped robot management platform.

[0136] S407. The quadruped robot executes the target task according to the control instructions and obtains the task execution result.

[0137] S408, the quadruped robot sends the task execution results to the quadruped robot management platform.

[0138] S409, the quadruped robot management platform receives the task execution results.

[0139] S410. When the task execution result indicates that there are illegally parked vehicles, absent duty personnel, abnormal self-service terminal equipment, or delivery handover failure, the quadruped robot management platform generates corresponding early warning information and sends the early warning information to the preset notification terminal.

[0140] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 5 As shown, the electronic device includes at least one processor 510 and a memory 520. The electronic device also includes a communication component 530. The processor 510, memory 520, and communication component 530 are connected via a bus 540.

[0141] In the specific implementation process, at least one processor 510 executes computer execution instructions stored in memory 520, causing at least one processor 510 to execute a control method for bank branches as executed on the electronic device side as described above.

[0142] The specific implementation process of processor 510 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0143] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0144] The memory may include high-speed RAM, and may also include non-volatile storage (NVM), such as at least one disk storage.

[0145] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0146] The above description of the functions implemented by electronic devices and main control devices has introduced the solutions provided by the embodiments of the present invention. It is understood that, in order to implement the above functions, the electronic device or main control device includes hardware structures and / or software modules corresponding to the execution of each function. By combining the units and algorithm steps of the various examples described in the embodiments of the present invention, the embodiments of the present invention can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the technical solutions of the embodiments of the present invention.

[0147] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the control method for bank branches described above.

[0148] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0149] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in an electronic device or a host device.

[0150] This application also provides a computer program product, which includes a computer program stored in a readable storage medium. At least one processor of an electronic device can read the computer program from the readable storage medium, and the at least one processor executes the computer program to cause the electronic device to perform the solution provided in the above embodiments.

[0151] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disk, or optical disk.

[0152] The technical solutions of this application have been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it is readily understood by those skilled in the art that the scope of protection of this application is obviously not limited to these specific embodiments. The above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A control method for bank branches, characterized in that, Applications to quadruped robot management platforms include: The system receives multimodal data collected by a quadruped robot within a bank branch, including video images, infrared temperature measurement data, and voice interaction content. Control commands are generated based on the multimodal data and preset task configuration information, wherein the preset task configuration information includes at least one of the following: armored vehicle channel inspection route, guide point list, material delivery address, and duty post location; The control command is sent to the quadruped robot so that the quadruped robot can execute the target task according to the control command, obtain the task execution result, and send the task execution result to the quadruped robot management platform. The target task includes at least one of vehicle illegal parking detection, intelligent navigation, material distribution, personnel on-duty verification and self-service terminal equipment status inspection. Receive the task execution result; When the task execution result indicates the presence of illegally parked vehicles, absent duty personnel, malfunctioning self-service terminal equipment, or failed delivery handover, a corresponding warning message is generated and sent to a preset notification terminal, which includes at least one of the following: administrator email, property management system, and mobile terminal.

2. The method according to claim 1, characterized in that, The step of generating control commands based on the multimodal data and preset task configuration information includes: The scene status within the bank branch is identified based on the video images. The scene status includes the illegal parking status of armored trucks in the armored truck lane, the operational status of self-service terminal equipment, or the activity status of people in public areas. The intent of the service request issued by the user is analyzed based on the voice interaction content. The presence of personnel on duty is determined based on the infrared temperature measurement data. The control command is generated based on the scenario status, the service request intent, the personnel on-duty status, and the preset task configuration information.

3. The method according to claim 1, characterized in that, The vehicle illegal parking detection includes: Receive the first image of the armored truck passage area sent by the quadruped robot during its initial inspection; Based on the first image, identify whether there are illegally parked vehicles in the armored vehicle lane area; If there is an illegally parked vehicle, a vehicle relocation notice is generated and sent to the preset notification terminal.

4. The method according to claim 3, characterized in that, After generating the vehicle relocation notification and sending it to the preset notification terminal, the method further includes: Receive a second image of the cash transport vehicle passage area sent by the quadruped robot during the secondary inspection; Confirm whether the illegally parked vehicle has not been moved based on the second image; If it is confirmed that the illegally parked vehicle has not been moved, the license plate information of the illegally parked vehicle will be added to the preset no-parking list, and a gate control command will be sent to the entrance and exit gate system of the bank branch to prevent the illegally parked vehicle from entering the bank branch again.

5. The method according to claim 1, characterized in that, The execution result corresponding to the material delivery includes the delivery handover status, which is determined by the quadruped robot management platform based on the verification result of any of the following methods: Receive the dynamic password uploaded by the quadruped robot and verify its validity; The robot receives the image of the recipient's identification card uploaded by the quadruped robot and compares it with the pre-stored identity information through image recognition. Alternatively, it can receive facial images uploaded by the quadruped robot and compare them with pre-stored identity information through facial recognition; When the verification result indicates that the identity does not match, the password is invalid, or no valid verification information is received, the delivery handover status is determined to be failed, and the warning information is generated.

6. A control method for bank branches, characterized in that, Applications in quadruped robots include: Multimodal data of the target area within the bank branch is collected using multimodal sensors. The multimodal data includes video images, infrared temperature measurement data, and voice interaction content. The multimodal data is sent to the quadruped robot management platform, so that the quadruped robot management platform generates control commands based on the multimodal data and preset task configuration information, and sends the control commands to the quadruped robot. The preset task configuration information includes at least one of the following: armored truck channel inspection route, guide point list, material delivery address, and duty post location. Receive the control commands sent by the quadruped robot management platform; The target task is executed according to the control command, the task execution result is obtained, and the task execution result is sent to the quadruped robot management platform. The target task includes at least one of vehicle illegal parking detection, intelligent navigation, material distribution, personnel on-duty verification, and self-service terminal equipment status inspection.

7. The method according to claim 6, characterized in that, The vehicle illegal parking detection includes: During the initial inspection, the robot moves along the inspection route of the armored truck lane according to the control command and captures a first image of the armored truck lane area through the onboard camera. The first image is then sent to the quadruped robot management platform so that the quadruped robot management platform can identify whether there are illegally parked vehicles in the armored truck lane area based on the first image. If there are illegally parked vehicles, a vehicle relocation notice is generated and sent to the preset notification terminal. During the second inspection, the robot again proceeds along the inspection route of the armored vehicle lane according to the control command, and captures a second image of the armored vehicle lane area through the camera. The second image is then sent to the quadruped robot management platform, so that the quadruped robot management platform can confirm whether the illegally parked vehicle has not been moved based on the second image. If it is confirmed that the illegally parked vehicle has not been moved, the license plate information of the illegally parked vehicle is added to the preset no-parking list, and a control command is sent to the entrance and exit gate system of the bank branch to prevent the illegally parked vehicle from entering the bank branch again.

8. The method according to claim 6, characterized in that, The delivery of the supplies includes: Upon arrival at the target location, the recipient will be prompted by voice or screen to enter a dynamic password, show their employee ID, or face the camera for facial recognition. Receive the dynamic password, or capture the recipient's employee ID card image or facial image through the built-in camera; The dynamic password, employee badge image, or facial image is sent to the quadruped robot management platform so that the quadruped robot management platform can verify the delivery handover status.

9. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 5 or 6 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 5 or 6 to 8.

11. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 5 or 6 to 8.

Images