Asset inspection method and device, electronic equipment and storage medium

Abstract

The invention discloses an asset inspection method and device, electronic equipment and a storage medium, and relates to the technical field of artificial intelligence or other related fields, and the method comprises the steps: determining an inspection path based on an inspection task and an environment map, and controlling target equipment to move in a target region according to the inspection path, the target equipment is asset inspection equipment provided with N sensors, and N is a positive integer; performing information sensing on the target area through a sensor on the target equipment in the moving process to obtain multi-modal sensing data; performing fusion analysis on the multi-modal sensing data to obtain an analysis result, the analysis result including asset identity information and an operation state evaluation result; and packaging the multi-modal sensing data and the analysis result into an asset checking result, and uploading the asset checking result to a cloud platform for storage. Through the method and the device, the technical problem of inaccurate inspection data in fixed asset management of financial institutions in related technologies is solved.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology or other related fields. Specifically, it relates to an asset inspection method and apparatus, electronic equipment, and storage medium. Background Technology

[0002] In the wave of digital transformation in financial institutions' asset management, with the rapid development of technologies such as cloud computing, the Internet of Things, and artificial intelligence, the automation level of fixed asset management systems in large financial institutions has been significantly improved. They have transitioned from traditional paper-and-pen records and manual inventory to a preliminary automation stage based on RFID (Radio Frequency Identification) and QR code recognition, which has simplified asset management processes to some extent. However, the core challenge of inaccurate inspection data remains in the fixed asset management of financial institutions, mainly manifested in the following aspects:

[0003] First, traditional systems relying on manual operation are prone to subjective errors during data collection, such as obstruction or damage to barcodes or QR codes, leading to tag reading failures, or oversights and misreadings during manual recording. Second, assets are widely distributed and diverse in form, especially for unlabeled assets and concealed equipment (such as circuit boards inside data centers or non-standard items in warehouses), existing technologies struggle to achieve comprehensive and thorough identification and tracking. Third, financial institutions' assets are often high-value and highly sensitive, and their operational status (such as the heat generation of electrical equipment or the operating indicator lights of servers) is crucial for asset management, but traditional methods struggle to monitor these subtle changes in real time and accurately, resulting in delayed risk warnings and untimely anomaly handling.

[0004] In summary, existing technologies suffer from inaccurate inspection and inventory data for fixed asset management in financial institutions, which hinders effective asset management.

[0005] There is currently no effective solution to the above problems. Summary of the Invention

[0006] The main purpose of this application is to provide an asset inspection method and device, electronic equipment, and storage medium to at least solve the technical problem of inaccurate inspection data in the fixed asset management of financial institutions in related technologies.

[0007] To achieve the above objectives, according to one aspect of this application, an asset inspection method is provided. The method includes: determining an inspection path based on an inspection task and an environmental map, and controlling a target device to move within a target area along the inspection path, wherein the target device is an asset inspection device equipped with N sensors, where N is a positive integer; during the movement, sensing information about the target area through the sensors on the target device to obtain multimodal sensing data, wherein the multimodal sensing data includes: equipment information, spatial location information, and operational indicator data of each asset device within the target area; performing fusion analysis on the multimodal sensing data to obtain analysis results, wherein the analysis results include: asset identity information and operational status assessment results; packaging the multimodal sensing data and the analysis results into an asset inventory result, and uploading the asset inventory result to a cloud platform for storage.

[0008] To achieve the above objectives, according to another aspect of this application, an asset inspection system is also provided for executing the asset inspection method described in any one of the above claims. The asset inspection system includes: a cloud platform for generating and distributing inspection tasks, and receiving, storing, and analyzing asset inventory results; at least one edge computing node deployed locally and communicatively connected to the cloud platform for requesting inspection from the cloud platform and sending a local environmental map, wherein "local" refers to the asset inspection site; and at least one asset inspection device communicatively connected to the cloud platform and the edge computing node for executing the inspection task within a target area. The asset inspection device includes: an autonomous navigation module for movement control based on the inspection task and the environmental map; a multimodal perception module for collecting multimodal perception data; and a data processing module for generating asset inventory results based on the multimodal perception data and uploading them to the cloud platform.

[0009] To achieve the above objectives, according to another aspect of this application, an asset inspection device is also provided, used to execute the asset inspection method described in any one of the above claims, or as the execution terminal of the asset inspection system described in any one of the above claims. The asset inspection device includes: a device body; a mobile chassis disposed at the bottom of the device body for autonomously moving the device body; a multimodal sensing module disposed on the device body, including a radio frequency identification reader, a visual sensor, a lidar, and an infrared thermal imager for collecting multimodal sensing data; a communication module for exchanging data with external systems, receiving inspection tasks, and uploading asset inventory data; and a control module disposed within the device body, electrically connected to the multimodal sensing module and the mobile chassis, for controlling device movement, sensing data, and data communication.

[0010] To achieve the above objectives, according to another aspect of this application, an asset inspection device is also provided. The device includes: a determining unit, configured to determine an inspection path based on an inspection task and an environmental map, and control a target device to move within a target area along the inspection path, wherein the target device is an asset inspection device equipped with N sensors, where N is a positive integer; a sensing unit, configured to perceive information about the target area through sensors on the target device during movement, obtaining multimodal sensing data, wherein the multimodal sensing data includes: equipment information, spatial location information, and operational indicator data of each asset device within the target area; an analysis unit, configured to perform fusion analysis on the multimodal sensing data to obtain analysis results, wherein the analysis results include: asset identity information and operational status assessment results; and an uploading unit, configured to package the multimodal sensing data and the analysis results into an asset inventory result, and upload the asset inventory result to a cloud platform for storage.

[0011] To achieve the above objectives, according to another aspect of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the asset inspection method described in any one of the above claims.

[0012] To achieve the above objectives, according to another aspect of this application, an electronic device is also provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the asset inspection method described in any one of the above claims.

[0013] To achieve the above objectives, according to another aspect of this application, a computer program product is also provided, including computer instructions, wherein when the computer instructions are executed by a processor, they implement the steps of the asset inspection method described in any one of the above claims.

[0014] This invention proposes an asset inspection method. First, an inspection path is determined based on the inspection task and an environmental map. Then, the target equipment is controlled to move within the target area according to the inspection path. The target equipment is an asset inspection device equipped with N sensors, where N is a positive integer. During the movement, the sensors on the target equipment perceive information about the target area, obtaining multimodal perception data. The multimodal perception data includes equipment information, spatial location information, and operational indicator data for each asset device within the target area. Then, the multimodal perception data is fused and analyzed to obtain analysis results, which include asset identity information and operational status assessment results. Finally, the multimodal perception data and analysis results are packaged into an asset inventory result and uploaded to a cloud platform for storage.

[0015] This invention employs a method that integrates artificial intelligence and Internet of Things technologies. By equipping intelligent inspection equipment with multimodal perception modules and autonomous decision-making algorithms, it achieves the goal of automatically and accurately collecting comprehensive information on the internal assets of financial institutions such as banks. This realizes the technical effect of intelligent asset management, greatly improves the efficiency and accuracy of fixed asset inspections, and solves the technical problem of inaccurate inspection data in the fixed asset management of financial institutions in related technologies.

[0016] Specifically, the asset inspection method proposed in this invention first relies on intelligent control of inspection equipment to autonomously navigate to the target area based on inspection task planning and environmental map. This process depends on SLAM (Simultaneous Localization and Mapping) technology using lidar and visual sensors to ensure accurate equipment positioning and path planning. Subsequently, the multi-sensor array on the inspection equipment works collaboratively to collect comprehensive information on each asset within the target area. This not only acquires the physical identification and location information of the equipment but also monitors its operating status, such as temperature, indicator light status, and other operational indicators, achieving deep perception of asset information. After data collection, the inspection equipment performs preliminary analysis of the multimodal perception data and then uses a fusion algorithm to extract and compare features from data from different sensors to obtain the asset's identity information and operational status assessment results. This fusion analysis process, combined with machine learning models and a business rule engine, can effectively identify whether there are any anomalies in the assets. Finally, the multimodal perception data and fusion analysis results are packaged into asset inventory results and uploaded to a cloud platform for storage and further analysis. The cloud platform manages the entire lifecycle of the assets and provides risk warnings, realizing intelligent and remote asset management. Attached Figure Description

[0017] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0018] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing an asset inspection method is shown.

[0019] Figure 2 This is a flowchart of an optional asset inspection method according to an embodiment of the present invention;

[0020] Figure 3 This is a schematic diagram of an optional asset inspection system according to an embodiment of the present invention;

[0021] Figure 4 This is a schematic diagram of an optional asset inspection device according to an embodiment of the present invention;

[0022] Figure 5 This is a schematic diagram of an optional asset inspection device according to an embodiment of the present invention;

[0023] Figure 6 This is a structural block diagram of an electronic device for performing an asset inspection method according to an embodiment of the present invention. Detailed Implementation

[0024] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0025] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention 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.

[0026] To facilitate understanding of the present invention by those skilled in the art, some terms or nouns involved in the various embodiments of the present invention are explained below:

[0027] Multimodal perception refers to the simultaneous acquisition of data from multiple types of sensors and the fusion of this data across different modalities to obtain more comprehensive and accurate environmental information. In this invention, multimodal perception is used to fuse data acquired from different sensors, thereby improving the accuracy of asset identification.

[0028] Autonomous navigation refers to the ability of a robot or unmanned device to automatically plan a path and move based on predetermined tasks and environmental information without direct human intervention. In this invention, autonomous navigation is achieved through the SLAM algorithm. SLAM allows the robot to simultaneously build a map of the environment and locate its own position in an unknown environment, ensuring that the robot can move efficiently and intelligently in the complex environment of a bank to conduct asset inspections.

[0029] Edge computing is a computing architecture that processes data at the network edge near the data source, rather than transmitting data to a remote data center or cloud for processing. This can significantly reduce data transmission latency, save network bandwidth, and improve the real-time performance and privacy protection of data processing.

[0030] A cloud platform refers to an internet infrastructure that provides computing resources, storage capacity, and software services, enabling large-scale data processing and analysis remotely. The cloud platform of this invention is responsible for receiving inspection data from edge devices, performing in-depth analysis, model training and updates, and centralized storage and management of the data.

[0031] A digital twin is a virtual model used to reflect the state of a physical object or system in real time. In this invention, a digital twin model of the banking environment can be constructed using 3D modeling and spatial analysis techniques. This maps the location, state, and attribute information of actual assets into the virtual environment, facilitating an intuitive display and management of the bank's fixed asset layout and supporting space utilization analysis and compliance checks.

[0032] ROS, or Robot Operating System, is not actually an operating system, but rather an open-source robot development framework. ROS provides a rich set of middleware and tool libraries for sensor management, motion control, AI algorithm integration, and more. The robot terminal of this invention runs ROS2, supporting coordinated processing of multi-sensor data and intelligent robot behavior.

[0033] It should be noted that the asset inspection method and device in this application can be used in the field of artificial intelligence technology for intelligent and automated inspection and status monitoring of fixed assets of financial institutions, and can also be used in any field other than artificial intelligence technology for intelligent and automated inspection and status monitoring of fixed assets of financial institutions. This application does not limit the application field of the asset inspection method and device.

[0034] It should be noted that all relevant information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, and displayed data) involved in this application are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, processing, transmission, provision, disclosure, use, and handling of such data comply with the laws, regulations, and standards of the relevant regions, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding access points are provided for users to choose to authorize or refuse access. For example, this system has interfaces with relevant users or organizations. Before obtaining relevant information, a request to obtain the information needs to be sent to the aforementioned user or organization through the interface, and the relevant information is obtained only after receiving consent from the aforementioned user or organization.

[0035] The information collection (e.g., user voice, video, and text collection) and analysis operations involved in this application have provided users with corresponding operation entry points during execution, allowing users to choose to agree to or reject the automated decision results; if the user chooses to reject, the process will proceed to the expert decision-making process.

[0036] The following embodiments of the present invention can be applied to various systems / applications / equipment that require intelligent management of fixed assets and real-time monitoring of environmental conditions, enabling an all-weather, high-precision, and intelligent asset management solution. The present invention uses an embodied intelligent robot equipped with a multimodal perception module to automatically collect asset and environmental information. Then, through edge computing and a cloud analytics platform, the acquired data is fused, analyzed, and used for intelligent decision-making. This allows for better detection of subtle changes and abnormal states in assets, while simultaneously providing immediate verification and early warning for suspected problematic assets.

[0037] This invention also utilizes 3D spatial modeling technology to accurately label the geographical location and spatial layout of assets, and performs deep semantic enhancement on asset data. This makes the asset management process not only intuitive and visual but also enables space utilization analysis and compliance checks. Combined with an autonomous inspection planning mechanism, this invention can dynamically adjust the inspection frequency and route based on the importance of the asset and its historical anomaly rate, ensuring the efficiency and accuracy of asset management.

[0038] The present invention will now be described in detail with reference to various embodiments.

[0039] Example 1

[0040] According to an embodiment of the present invention, an embodiment of an asset inspection method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0041] The asset inspection method provided in Embodiment 1 of the present invention can be executed on a mobile terminal, computer terminal or similar computing device. Figure 1 A hardware block diagram of a computer terminal (or mobile device) for implementing an asset inspection method is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0042] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0043] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the asset inspection method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned asset inspection method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0044] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0045] The display can be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0046] Under the above operating environment, the present invention provides, as follows: Figure 2 The asset inspection method shown is implemented by an asset inspection system or a bank fixed asset management system based on an embodied intelligent robot. Combining artificial intelligence and Internet of Things technologies, it is used in fixed asset management scenarios of banks and financial institutions. In particular, it addresses the problem of inaccurate inspection data by using multimodal perception fusion and autonomous navigation control. Specifically, it involves determining the inspection path, collecting multimodal perception data, fusing and analyzing the data, and generating and uploading the asset inventory results to achieve the goal of comprehensive, accurate, and real-time asset management and risk warning.

[0047] The embodiments of the present invention will now be described in detail with reference to each specific step.

[0048] Figure 2 This is a flowchart of an optional asset inspection method according to an embodiment of the present invention, such as... Figure 2 As shown, the method includes the following steps:

[0049] Step S201: Determine the inspection path based on the inspection task and the environmental map, and control the target equipment to move within the target area according to the inspection path. The target equipment is an asset inspection device equipped with N sensors, where N is a positive integer.

[0050] It should be noted that inspection tasks refer to routine or emergency inspection procedures set for fixed assets in a specific area of ​​a bank or financial institution. These procedures include the time, location, asset category, inspection frequency, and any specific inspection requirements or standards. For example, inspecting the equipment operation status and asset management information in the server room to ensure that all equipment is in its designated location and is operating normally, or conducting an initial asset inventory and spatial layout confirmation of a newly renovated office area.

[0051] An environmental map refers to a detailed geographic information model of a target area, including but not limited to: floor plans, passageway layouts, stairwell locations, obstacle information, access control areas, etc. In this embodiment, the environmental map is not limited to traditional two-dimensional or three-dimensional maps, but may also include asset location information, historical inspection routes, and other additional data that helps optimize inspections. The environmental map is updated in real time using SLAM technology to ensure accurate location and planning of inspection paths.

[0052] An inspection path refers to the route planned by asset inspection equipment within a given environmental map to complete its inspection task. The robot uses a multi-objective optimization algorithm, considering factors such as shortest time, minimum distance, and maximum asset coverage, to dynamically generate the inspection path. Combined with dynamic obstacle detection and obstacle avoidance mechanisms, the path planning ensures not only coverage of all target assets but also real-time adjustments to avoid moving pedestrians or obstacles, guaranteeing both efficiency and safety during the inspection process.

[0053] In this embodiment, the target device refers to an embodied robot (wheeled, legged, or wheel-legged hybrid mobile robot) equipped with multimodal sensing sensors, capable of autonomous movement and performing functions such as asset identification, detection, and data uploading. The target area includes all areas within a bank or financial institution where fixed assets requiring management are located, such as server rooms, office areas, warehouses, underground vaults, etc., scattered in different physical locations, such as the head office building, branch outlets, and even external logistics centers.

[0054] Before embodied robots are deployed for inspection work, a series of preparatory work needs to be carried out, including: sensor calibration to ensure that functions such as RFID reading, QR code recognition, multi-camera shooting, and LiDAR scanning are normal and accurate; navigation and positioning algorithm pre-training, that is, training the robot's navigation system based on historical environmental map data to improve its autonomous positioning and path planning capabilities; security and privacy protection measures to ensure that the robot complies with relevant regulations when performing inspection tasks, such as encrypted data transmission and setting restricted areas during non-working hours; and integration testing with existing IT systems to ensure that the data collected by the robot can be seamlessly connected to the bank's asset management system to achieve automatic data synchronization and verification.

[0055] The sensors on the embodied robot encompass various types, including RFID identification, QR code / barcode scanning, image classification and target detection, depth information acquisition, infrared temperature detection, and ultrasonic obstacle detection. Together, they form a multimodal sensing array capable of identifying and detecting various types of fixed assets in different scenarios, including tagged devices and untagged physical assets. The sensor array design must consider robustness, interoperability, and data fusion capabilities to ensure efficient and accurate data acquisition under various lighting conditions and environmental interference.

[0056] Step S202: During the movement, the target area is perceived by the sensors on the target device to obtain multimodal perception data. The multimodal perception data includes: equipment information, spatial location information and operation index data of each asset device in the target area.

[0057] It should be noted that information perception refers to the process of collecting environmental and asset information within a target area through a sensor array on the target device. In this embodiment, information perception relies on multimodal fusion sensing technology. The sensor array includes, but is not limited to, RFID readers, QR code / barcode scanners, RGB and depth cameras, LiDAR, infrared sensors and thermal imagers, and ultrasonic sensors, working together to obtain detailed perception data. For example, an RFID reader can read RFID tags on assets from a distance, while a multi-view camera can be used to identify untagged assets, classifying and extracting information through image recognition technology. LiDAR and depth cameras are combined to generate a 3D point cloud and depth map of the target area, used to construct an accurate environmental map and understand the spatial layout.

[0058] To further clarify, multimodal perception data refers to a comprehensive dataset containing different types of information acquired through the aforementioned sensor set. Specifically, this includes: equipment information, such as asset ID, model, and serial number; spatial location information, referring to the precise coordinates of the asset within the bank building space determined by SLAM technology based on LiDAR and depth cameras; operational indicator data, referring to monitoring the temperature status of equipment through infrared sensors to determine if there is any abnormal overheating, and using thermal imagers to detect the temperature distribution during equipment operation; and ultrasonic sensors for near-field obstacle detection, assisting in positioning and collision avoidance.

[0059] In addition, the assets and equipment within the target area refer to fixed assets within the bank or financial institution that require inspection and management, including but not limited to servers, computers, office furniture, safes, ATMs, etc., which can be distributed across multiple physical spaces, such as offices, server rooms, warehouses, vaults, etc.

[0060] Optionally, in the asset inspection method provided in this embodiment of the invention, the multimodal perception module includes: a radio frequency identification (RFID) reader, a visual sensor, a lidar, and an infrared thermal imager. The step of obtaining multimodal perception data by sensing information about the target area through sensors on the target device includes: reading electronic tag information on the asset device through the RFID reader; acquiring image information and depth information of the asset device through the visual sensor to obtain device information; acquiring spatial location information of the asset device through the lidar; collecting heat distribution information of the asset device through the infrared thermal imager, and judging abnormal heating indicators or power-on operation indicators of the asset device based on the heat distribution information to obtain operation indicator data.

[0061] In this embodiment, the various components of the multimodal sensing module are designed to work collaboratively to provide comprehensive, accurate, and intelligent data collection and analysis of the bank's fixed assets. Among them, the Radio Frequency Identification (RFID) reader is a wireless signal system used for contactless reading of RFID tags attached to asset equipment, containing the equipment's unique identification code and other information. The reader activates the RFID tag by emitting electromagnetic waves of a specific frequency, and the tag then transmits the stored information back to the reader in the form of a wireless signal. During the reading of electronic tag information, an anti-collision algorithm should be used to ensure that misreading or missed readings do not occur in a multi-tag environment. Furthermore, while reading RFID information, real-time updates and verification of tag information can also be performed to ensure the accuracy and timeliness of tag data.

[0062] Alternatively, visual sensors, including RGB color cameras and depth cameras, can capture images of the asset's appearance and 3D depth information for asset identification, classification, and 3D reconstruction of the environment. The RGB camera captures high-resolution visual information, while the depth camera acquires point cloud data of the asset. Image processing algorithms are then used for target detection and classification to extract information such as the device's model and status. In low-light or reflective environments, auxiliary lighting equipment or special imaging techniques (such as structured light) should be used to ensure image quality and recognition accuracy.

[0063] Furthermore, LiDAR (Light Detection and Ranging) is used to emit laser pulses and receive reflected signals to measure the distance to assets and equipment, generating accurate 3D point cloud data. LiDAR continuously scans the environment, collecting point cloud data within the target area, and uses SLAM (Simultaneous Localization and Mapping) algorithms for real-time localization and map building, thereby determining the precise spatial location of assets and equipment. During the LiDAR data processing stage, point cloud data denoising and registration should be performed to improve positioning accuracy and map quality.

[0064] In addition, infrared thermal imagers are used to detect and record the temperature distribution on the surface of objects. By analyzing thermal images, abnormal heating areas are identified to determine whether the equipment is powered on and operating. Combining the thermal images captured by the thermal imager with preset temperature thresholds and thermal image analysis algorithms, the operating status of the equipment can be identified. Multiple abnormal temperature modes can also be set to adapt to the heating characteristics of different types of equipment. At the same time, time series analysis is used to identify the periodic power-on patterns of the equipment, thereby determining whether the equipment is working properly.

[0065] Optionally, in the asset inspection method provided in this embodiment of the invention, the multimodal perception module further includes: a spectral imager and a millimeter-wave radar. The step of obtaining multimodal perception data by sensing information about the target area through sensors on the target device further includes: collecting surface spectral features of the asset device through the spectral imager; and / or collecting dielectric features of the asset device through the millimeter-wave radar.

[0066] It's important to note that a spectral imager is a high-precision imaging device capable of capturing the spectral information reflected or absorbed by an object's surface, providing richer material and condition information than RGB images. A spectral imager scans the target asset, collecting reflectance spectra at different wavelengths. By analyzing these spectral curves, it identifies the asset's material type (e.g., metal, plastic, wood) and surface treatment condition (e.g., paint, rust). The surface spectral characteristics of assets can be used for material identification, condition assessment, and early detection of potential problems. For example, changes in paint spectral characteristics may indicate corrosion or damage to the asset, facilitating timely maintenance.

[0067] In addition, millimeter-wave radar is a sensor that uses high-frequency electromagnetic waves for non-contact detection. It can penetrate non-metallic materials to measure the dielectric constant of objects, reflecting their internal structure and material properties. Millimeter-wave radar emits microwave pulses towards the target asset, receives the reflected signals, and infers the asset's internal structure and dielectric properties by analyzing the phase, amplitude, and frequency characteristics of the signals. Dielectric characteristics can help identify anomalies inside equipment, such as loose internal components, liquid leaks, or corrosion of metal layers, for assessing equipment health and safety, and are particularly suitable for non-destructive testing of enclosed electronic equipment such as servers and cabinets.

[0068] In another optional embodiment, in addition to the aforementioned multimodal perception technology, the robot needs to have a waterproof and dustproof design to adapt to various indoor and outdoor environments and ensure stable operation. It also integrates an autonomous charging module, which automatically returns to the charging station to recharge when the robot's battery level falls below a preset threshold, ensuring continuous operation.

[0069] Optionally, in the asset inspection method provided in this embodiment of the invention, after obtaining multimodal perception data by sensing information about the target area through sensors on the target device, the method further includes: constructing a three-dimensional digital twin model of the target area based on the multimodal perception data; and labeling the asset inventory results as attribute tags in the three-dimensional digital twin model.

[0070] In this embodiment, after the multimodal perception module on the target device (i.e., the embodied intelligent robot) perceives information about the target area, the resulting multimodal perception data is not only used for asset inventory but also for constructing a three-dimensional digital twin model of the target area. The asset inventory results are then integrated into the model in the form of attribute tags, thereby achieving visualized and intelligent management of the bank's fixed assets.

[0071] It should be noted that a 3D digital twin model is a precise mapping of the real physical environment of the target area in the digital world. It not only includes the spatial layout and the 3D location of assets and equipment, but also embeds the attribute information, status data and real-time operating indicators of the equipment. As the "digital brain" of the asset management system, it can provide intuitive visual display and in-depth data analysis.

[0072] The construction of a 3D digital twin model is mainly based on 3D reconstruction technology and multi-sensor data fusion. Specific steps include: S1, point cloud data acquisition and fusion, which involves fusing point cloud data generated by LiDAR and depth cameras to form a preliminary 3D framework of the target area; S2, RGB image texture mapping, which maps color information captured by RGB cameras onto the point cloud model to represent the visual appearance of the real world; S3, spatial layout and asset positioning, which involves updating and optimizing the model in real time using SLAM algorithms to ensure the accuracy of the spatial layout and asset locations within the model; and S4, attribute tag and status data integration, which involves embedding asset information and status data collected by RFID, QR codes, spectral imaging, and millimeter-wave radar as attribute tags into the corresponding asset models.

[0073] Specifically, the asset inventory results are transformed into attribute tags. These tags contain key information such as the asset's ID, type, location coordinates, and operational status, as well as surface spectral and dielectric characteristics obtained from sensors such as spectral imagers and millimeter-wave radar. The attribute tagging process is as follows: T1, Location and Matching, which uses LiDAR or a positioning system to determine the precise location of the asset equipment in the 3D digital twin model; T2, Information Extraction and Encoding, which extracts detailed asset information from multimodal sensing data, including ID, status, and material characteristics, and encodes it into attribute tags; T3, Tag Fusion and Display, which fuses the attribute tags with the corresponding asset model, displaying attribute information at each asset location on the model. This information can be displayed on the digital twin interface in the form of pop-up windows, virtual tags, or heatmaps, allowing users to quickly understand the overall picture and status of the assets.

[0074] Step S203 involves fusing and analyzing the multimodal perception data to obtain analysis results, including asset identity information and operational status assessment results. This step is jointly implemented by the edge computing unit and the cloud-based AI analysis platform, generating asset identity information and operational status assessment results through the fusing and analysis of multimodal perception data.

[0075] Optionally, in the asset inspection method provided in this embodiment of the invention, the step of fusing and analyzing multimodal sensing data to obtain analysis results includes: preprocessing the multimodal sensing data, wherein the preprocessing includes: noise filtering, data format standardization, and timestamp alignment; using a feature extraction algorithm to extract features from the preprocessed multimodal sensing data to obtain multimodal features, wherein the multimodal features include at least: geometric features, surface texture features, and temperature distribution features; calling an edge computing node and / or a cloud platform to fuse the multimodal features to obtain fused features, wherein both the edge computing node and the cloud platform are connected to a historical database; determining the asset identity information of each asset device within the target area based on the fused features, and assessing whether the operating status of each asset device is abnormal to obtain analysis results.

[0076] The purpose of data preprocessing is to ensure the quality of multimodal sensing data, eliminate noise interference, and standardize the data format for subsequent algorithm processing. Methods include: noise filtering, which removes random or unwanted signals from sensor data, such as thermal noise in thermal imaging and stray light in lidar; data format standardization, which converts raw data from different sensors into a unified format for easier algorithm processing; and timestamp alignment, ensuring that data acquired from different sensors describes the target area at the same time, avoiding inconsistencies in analysis data due to different acquisition speeds of different sensors. Data preprocessing is fundamental to multimodal data analysis, improving data usability and algorithm accuracy, and ensuring the effectiveness of fusion analysis.

[0077] Furthermore, feature extraction algorithms can include deep learning algorithms (such as CNN, convolutional neural networks) and traditional computer vision techniques (such as ORB algorithm, SURF algorithm), etc. The role of feature extraction is to extract key information representing asset identity and status from multimodal perception data, providing a basis for asset identification and status assessment.

[0078] Among the multimodal features, geometric features include the asset's size, shape, and location information, provided by LiDAR and depth cameras; surface texture features include image information captured by RGB cameras, which, combined with hyperspectral imagers, can reveal the spectral properties of the material; and temperature distribution features include temperature information captured by infrared thermal imagers, reflecting the asset's operational status and health level. This combination of multimodal features provides a comprehensive view of the asset, enabling not only asset identification but also assessment of operational status and potential problems.

[0079] Another point to note is that edge computing nodes, located on local servers or within the robot itself, are responsible for real-time processing and local decision-making to accelerate response times. After preprocessing and initial feature extraction, necessary computational tasks are assigned to edge computing nodes, utilizing computing resources for fusion analysis. The cloud platform provides remote data processing, deep learning model training, and large-capacity data storage capabilities. Edge computing nodes upload complex tasks and large amounts of data that cannot be processed locally to the cloud via network connections, calling upon the cloud's AI analysis platform for in-depth analysis and feature fusion.

[0080] Feature fusion relies on information fusion algorithms, such as Bayesian fusion, Kalman filtering, and particle filtering, to integrate feature data from different modalities and form a comprehensive understanding of assets and equipment.

[0081] In the steps of determining asset identity information and assessing operational status based on fusion characteristics, asset identity information determination relies on RFID tag reading, QR code recognition, and OCR technology combined with visual sensors to recognize text information on nameplates. Operational status assessment determines whether equipment is overheating based on temperature distribution characteristics, analyzes whether equipment has been moved or damaged through geometric and texture features, and assesses the equipment's health status and compliance by combining historical data and a rule base.

[0082] Step S204: Package the multimodal sensing data and analysis results into asset inventory results, and upload the asset inventory results to the cloud platform for storage.

[0083] In this embodiment, after the robot completes data collection and preliminary analysis, all collected multimodal perception data (including RFID information, images, point clouds, thermal distribution, etc.) and analysis results obtained through fusion processing (such as asset ID and operational status assessment) are organized into a structured data package. This package includes not only the original sensor data but also the analysis results and metadata (such as timestamps, location information, and robot ID), ensuring the integrity and traceability of the data.

[0084] By packaging multimodal sensing data and analysis results together, the aim is to provide a comprehensive and coherent view of asset management. Raw data provides direct evidence of asset status, while analysis results offer expert-level interpretation of the current condition of assets. The combination of both ensures the accuracy and reliability of asset inventory results.

[0085] Uploading to the cloud platform enables centralized storage and unified management of asset data, facilitating data sharing and collaboration among multiple robots and locations. The cloud platform provides powerful data processing capabilities and scalability, capable of accommodating massive amounts of data, supporting historical data analysis and model training, while ensuring data security and compliance.

[0086] Beyond storing asset inventory results, the cloud platform serves as the core technology and data hub of the entire system. This includes: an Asset Management System (AMS) that stores full lifecycle information of assets, processes newly added asset inventory data, and updates asset status; a Task Scheduling Platform (TSP) that intelligently plans robot inspection tasks based on asset status and historical records, optimizing resource allocation; a 3D Mapping Service Platform (3D-MSP) that maintains and updates 3D maps of various locations, providing a foundation for robot navigation and spatial analysis; and an AI Analysis Platform (AIP) that runs deep learning models to conduct in-depth analysis of uploaded multimodal data, uncovering potential problems and providing optimization suggestions.

[0087] In addition to storing data, the cloud platform also performs advanced functions such as data mining and analysis, model iteration and optimization, remote monitoring and command, ensuring that the level of intelligence and automation of asset inventory is continuously improved, and providing strong data support for asset management decisions.

[0088] Optionally, in the asset inspection method provided in the embodiments of the present invention, after packaging the multimodal perception data and analysis results into asset inventory results, the method further includes: comparing the asset inventory results obtained in this inspection with historical data, and generating alarm information and pushing the alarm information if asset loss, location change or abnormal operation status is identified during the comparison process.

[0089] The cloud platform in this embodiment stores past asset inventory results, including asset IDs, location coordinates, and operational status. The current asset inventory results are automatically compared with historical data after being uploaded, identifying discrepancies. If a previously recorded asset is not detected in the current inventory, it is marked as "missing"; if a significant displacement is found between the current and historical coordinates of an asset, it is marked as "location changed"; based on the analysis of operational status data such as temperature distribution, indicator light colors, and equipment noise, if asset operational parameters are found to be outside the normal range, it is judged as "abnormal operational status".

[0090] Alarm information includes, but is not limited to: asset ID, type, anomaly type (e.g., "missing," "location changed," "abnormal operating status"), last known location, current time, and alarm level. Alarm information is pushed to the bank's security and maintenance department and the personnel responsible for the assets in real time to ensure rapid response.

[0091] Through steps S201 to S204 above, the inspection path can be determined first based on the inspection task and environmental map, and the target equipment can be controlled to move within the target area according to the inspection path. The target equipment is an asset inspection device equipped with N sensors, where N is a positive integer. During the movement, the sensors on the target equipment perceive information about the target area and obtain multimodal perception data. The multimodal perception data includes: equipment information, spatial location information, and operational indicator data of each asset device within the target area. Then, the multimodal perception data is fused and analyzed to obtain analysis results. The analysis results include: asset identity information and operational status assessment results. Finally, the multimodal perception data and analysis results are packaged into asset inventory results and uploaded to the cloud platform for storage.

[0092] In this embodiment of the invention, an approach integrating artificial intelligence and Internet of Things technologies is adopted. By equipping intelligent inspection equipment with multimodal perception modules and autonomous decision-making algorithms, the goal of automatically and accurately collecting comprehensive information on the internal assets of financial institutions such as banks is achieved. This realizes the technical effect of intelligent asset management, greatly improves the efficiency and accuracy of fixed asset inspections, and solves the technical problem of inaccurate inspection data in the fixed asset management of financial institutions in related technologies.

[0093] Specifically, the asset inspection method proposed in this embodiment first uses intelligent control of inspection equipment to autonomously navigate to the target area based on inspection task planning and environmental map. This process relies on SLAM technology using LiDAR and visual sensors to ensure accurate equipment positioning and path planning. Subsequently, the multi-sensor array on the inspection equipment works collaboratively to collect comprehensive information on each asset device within the target area. This not only acquires the physical identification and location information of the equipment but also monitors its operating status, such as temperature, indicator light status, and other operational indicators, achieving deep perception of asset information. After data collection, the inspection equipment performs preliminary analysis of the multimodal perception data and then uses a fusion algorithm to extract and compare features from data from different sensors to obtain the asset's identity information and operational status assessment results. This fusion analysis process, combined with machine learning models and a business rule engine, can effectively identify whether there are any anomalies in the assets. Finally, the multimodal perception data and fusion analysis results are packaged into asset inventory results and uploaded to a cloud platform for storage and further analysis. The cloud platform manages the entire lifecycle of the assets and provides risk warnings, realizing intelligent and remote asset management.

[0094] The invention will now be described in conjunction with another alternative embodiment.

[0095] Example 2

[0096] This invention also provides an asset inspection system. It should be noted that the asset inspection system of this invention includes multiple implementation components, which can be used to execute the asset inspection method provided in Embodiment 1 above.

[0097] Figure 3 This is a schematic diagram of an optional asset inspection system according to an embodiment of the present invention, such as... Figure 3 As shown, the asset inspection system may include: a cloud platform 31, at least one edge computing node 32, and at least one asset inspection device 33.

[0098] The cloud platform 31 is used to generate and distribute inspection tasks, and to receive, store, and analyze asset inventory results. In this embodiment, an inspection task refers to a routine or emergency inspection procedure set for fixed assets within a specific area of ​​a bank or financial institution. This includes the inspection time, location, asset category, inspection frequency, and any specific inspection requirements or standards. For example, it could involve checking the equipment operating status and asset management information in a server room to ensure all equipment is in its designated location and operating normally, or conducting an initial asset inventory and spatial layout confirmation of a newly renovated office area. The asset inventory result is generated after data collection and preliminary analysis. All collected multimodal sensing data (including RFID information, images, point clouds, thermal distribution, etc.) and the analysis results obtained through fusion processing (such as asset IDs and operating status assessments) are organized into a structured data packet. This packet includes not only the raw sensor data but also the analysis results and metadata (such as timestamps, location information, and robot IDs).

[0099] At least one edge computing node 32, deployed locally and communicating with the cloud platform, is used to request inspections from the cloud platform and send a local environmental map, where "local" refers to the asset inspection site. The environmental map is a detailed geographic information model of the target area, including but not limited to: floor plans, passageway layouts, stairwell locations, obstacle information, access control areas, etc. In this embodiment, the environmental map is not limited to traditional two-dimensional or three-dimensional maps, but may also include asset location information, historical inspection routes, and other additional data that helps optimize inspections. The environmental map is updated in real time using SLAM technology to ensure accurate location and planning of inspection paths.

[0100] It should be further explained that the inspection path refers to the route planned by the asset inspection equipment in a given environmental map to complete the inspection task. The robot uses a multi-objective optimization algorithm to dynamically generate the inspection path, considering factors such as shortest time, minimum distance, and maximum asset coverage. Combined with dynamic obstacle detection and obstacle avoidance mechanisms, the path planning ensures that it not only covers all target assets but can also be adjusted in real time to avoid moving pedestrians or obstacles, guaranteeing the efficiency and safety of the inspection process.

[0101] At least one asset inspection device 33 is communicatively connected to a cloud platform and edge computing nodes to perform inspection tasks within a target area. In this embodiment, the asset inspection device can be an embodied robot equipped with multimodal sensing sensors, capable of autonomous movement and performing functions such as asset identification, detection, and data uploading. The target area includes all areas within the bank or financial institution where fixed assets requiring management are located, such as server rooms, office areas, warehouses, underground vaults, etc., scattered in different physical locations, such as the head office building, branch outlets, and even external logistics centers.

[0102] Asset inspection equipment 33 includes:

[0103] The autonomous navigation module 331 is used for movement control based on the inspection task and the environmental map.

[0104] The multimodal sensing module 332 is used to collect multimodal sensing data. The steps for collecting multimodal sensing data are as follows: during the movement, information is perceived in the target area through sensors, and the obtained multimodal sensing data includes: equipment information, spatial location information, and operational indicator data of each asset device in the target area.

[0105] The data processing module 333 is used to generate asset inventory results based on multimodal sensing data and upload them to the cloud platform. The steps for generating asset inventory results based on multimodal sensing data include: performing fusion analysis on the multimodal sensing data to obtain analysis results; packaging the multimodal sensing data and analysis results into asset inventory results; and uploading the asset inventory results to the cloud platform for storage. The analysis results include: asset identity information and whether there are any operational anomalies.

[0106] Optionally, in the asset inspection method provided in this embodiment of the invention, the step of fusing and analyzing multimodal sensing data to obtain analysis results includes: preprocessing the multimodal sensing data, wherein the preprocessing includes: noise filtering, data format standardization, and timestamp alignment; using a feature extraction algorithm to extract features from the preprocessed multimodal sensing data to obtain multimodal features, wherein the multimodal features include at least: geometric features, surface texture features, and temperature distribution features; calling an edge computing node and / or a cloud platform to fuse the multimodal features to obtain fused features, wherein both the edge computing node and the cloud platform are connected to a historical database; determining the asset identity information of each asset device within the target area based on the fused features, and assessing whether the operating status of each asset device is abnormal to obtain analysis results.

[0107] Optionally, in the asset inspection method provided in this embodiment of the invention, the multimodal perception module includes: a radio frequency identification card reader for reading electronic tag information on the asset equipment; a visual sensor for acquiring image information and depth information of the asset equipment; a lidar for acquiring spatial location information of the asset equipment; and an infrared thermal imager for collecting thermal distribution information of the asset equipment.

[0108] In this embodiment, the various components of the multimodal sensing module are designed to work collaboratively to provide comprehensive, accurate, and intelligent data collection and analysis of the bank's fixed assets. Among them, the Radio Frequency Identification (RFID) reader is a wireless signal system used for contactless reading of RFID tags attached to asset equipment, containing the equipment's unique identification code and other information. The reader activates the RFID tag by emitting electromagnetic waves of a specific frequency, and the tag then transmits the stored information back to the reader in the form of a wireless signal. During the reading of electronic tag information, an anti-collision algorithm should be used to ensure that misreading or missed readings do not occur in a multi-tag environment. Furthermore, while reading RFID information, real-time updates and verification of tag information can also be performed to ensure the accuracy and timeliness of tag data.

[0109] Alternatively, visual sensors, including RGB color cameras and depth cameras, can capture images of the asset's appearance and 3D depth information for asset identification, classification, and 3D reconstruction of the environment. The RGB camera captures high-resolution visual information, while the depth camera acquires point cloud data of the asset. Image processing algorithms are then used for target detection and classification to extract information such as the device's model and status. In low-light or reflective environments, auxiliary lighting equipment or special imaging techniques (such as structured light) should be used to ensure image quality and recognition accuracy.

[0110] Furthermore, LiDAR (Light Detection and Ranging) is used to emit laser pulses and receive reflected signals to measure the distance to assets and equipment, generating accurate 3D point cloud data. LiDAR continuously scans the environment, collecting point cloud data within the target area, and uses SLAM (Simultaneous Localization and Mapping) algorithms for real-time localization and map building, thereby determining the precise spatial location of assets and equipment. During the LiDAR data processing stage, point cloud data denoising and registration should be performed to improve positioning accuracy and map quality.

[0111] In addition, infrared thermal imagers are used to detect and record the temperature distribution on the surface of objects. By analyzing thermal images, abnormal heating areas are identified to determine whether the equipment is powered on and operating. Combining the thermal images captured by the thermal imager with preset temperature thresholds and thermal image analysis algorithms, the operating status of the equipment can be identified. Multiple abnormal temperature modes can also be set to adapt to the heating characteristics of different types of equipment. At the same time, time series analysis is used to identify the periodic power-on patterns of the equipment, thereby determining whether the equipment is working properly.

[0112] Optionally, in the asset inspection method provided in this embodiment of the invention, the multimodal sensing module further includes: a spectral imager for collecting surface spectral features of the asset equipment; and / or a millimeter-wave radar for collecting dielectric features of the asset equipment.

[0113] It's important to note that a spectral imager is a high-precision imaging device capable of capturing the spectral information reflected or absorbed by an object's surface, providing richer material and condition information than RGB images. A spectral imager scans the target asset, collecting reflectance spectra at different wavelengths. By analyzing these spectral curves, it identifies the asset's material type (e.g., metal, plastic, wood) and surface treatment condition (e.g., paint, rust). The surface spectral characteristics of assets can be used for material identification, condition assessment, and early detection of potential problems. For example, changes in paint spectral characteristics may indicate corrosion or damage to the asset, facilitating timely maintenance.

[0114] In addition, millimeter-wave radar is a sensor that uses high-frequency electromagnetic waves for non-contact detection. It can penetrate non-metallic materials to measure the dielectric constant of objects, reflecting their internal structure and material properties. Millimeter-wave radar emits microwave pulses towards the target asset, receives the reflected signals, and infers the asset's internal structure and dielectric properties by analyzing the phase, amplitude, and frequency characteristics of the signals. Dielectric characteristics can help identify anomalies inside equipment, such as loose internal components, liquid leaks, or corrosion of metal layers, for assessing equipment health and safety, and are particularly suitable for non-destructive testing of enclosed electronic equipment such as servers and cabinets.

[0115] Optionally, in the asset inspection method provided in this embodiment of the invention, an edge computing module is set on the edge computing node to run a lightweight algorithm model to process the received data, and a data analysis module is set on the cloud platform to run an artificial intelligence algorithm model to mine and analyze the received data.

[0116] Optionally, in the asset inspection method provided in this embodiment of the invention, the system is further configured to: compare the asset inventory results obtained in this inspection with historical data in the edge computing module and / or data analysis module; generate alarm information and push the alarm information if asset loss, location change or abnormal status is detected during the comparison process.

[0117] The cloud platform in this embodiment stores past asset inventory results, including asset IDs, location coordinates, and operational status. The current asset inventory results are automatically compared with historical data after being uploaded, identifying discrepancies. If a previously recorded asset is not detected in the current inventory, it is marked as "missing"; if a significant displacement is found between the current and historical coordinates of an asset, it is marked as "location changed"; based on the analysis of operational status data such as temperature distribution, indicator light colors, and equipment noise, if asset operational parameters are found to be outside the normal range, it is judged as "abnormal operational status".

[0118] Alarm information includes, but is not limited to: asset ID, type, anomaly type (e.g., "missing," "location changed," "abnormal operating status"), last known location, current time, and alarm level. Alarm information is pushed to the bank's security and maintenance department and the personnel responsible for the assets in real time to ensure rapid response.

[0119] Optionally, in the asset inspection method provided in this embodiment of the invention, the system further includes: a digital twin module, which is communicatively connected to the asset inspection equipment, for constructing a three-dimensional digital twin model of the target area based on the multimodal perception data collected by the asset inspection equipment; and labeling the asset inventory results as attribute tags in the three-dimensional digital twin model.

[0120] In this embodiment, after the multimodal perception module on the target device (i.e., the embodied intelligent robot) perceives information about the target area, the resulting multimodal perception data is not only used for asset inventory but also for constructing a three-dimensional digital twin model of the target area. The asset inventory results are then integrated into the model in the form of attribute tags, thereby achieving visualized and intelligent management of the bank's fixed assets.

[0121] It should be noted that a 3D digital twin model is a precise mapping of the real physical environment of the target area in the digital world. It not only includes the spatial layout and the 3D location of assets and equipment, but also embeds the attribute information, status data and real-time operating indicators of the equipment. As the "digital brain" of the asset management system, it can provide intuitive visual display and in-depth data analysis.

[0122] The construction of a 3D digital twin model is mainly based on 3D reconstruction technology and multi-sensor data fusion. Specific steps include: S1, point cloud data acquisition and fusion, which involves fusing point cloud data generated by LiDAR and depth cameras to form a preliminary 3D framework of the target area; S2, RGB image texture mapping, which maps color information captured by RGB cameras onto the point cloud model to represent the visual appearance of the real world; S3, spatial layout and asset positioning, which involves updating and optimizing the model in real time using SLAM algorithms to ensure the accuracy of the spatial layout and asset locations within the model; and S4, attribute tag and status data integration, which involves embedding asset information and status data collected by RFID, QR codes, spectral imaging, and millimeter-wave radar as attribute tags into the corresponding asset models.

[0123] Specifically, the asset inventory results are transformed into attribute tags. These tags contain key information such as the asset's ID, type, location coordinates, and operational status, as well as surface spectral and dielectric characteristics obtained from sensors such as spectral imagers and millimeter-wave radar. The attribute tagging process is as follows: T1, Location and Matching, which uses LiDAR or a positioning system to determine the precise location of the asset equipment in the 3D digital twin model; T2, Information Extraction and Encoding, which extracts detailed asset information from multimodal sensing data, including ID, status, and material characteristics, and encodes it into attribute tags; T3, Tag Fusion and Display, which fuses the attribute tags with the corresponding asset model, displaying attribute information at each asset location on the model. This information can be displayed on the digital twin interface in the form of pop-up windows, virtual tags, or heatmaps, allowing users to quickly understand the overall picture and status of the assets.

[0124] Optionally, in the asset inspection method provided in the embodiments of the present invention, the asset inspection equipment is a wheeled, legged, or wheel-legged hybrid mobile robot.

[0125] In this embodiment of the invention, by integrating artificial intelligence and Internet of Things technologies, and by using intelligent inspection equipment equipped with multimodal perception modules and autonomous decision-making algorithms, the goal of automatically and accurately collecting comprehensive information on the internal assets of financial institutions such as banks is achieved. This realizes the technical effect of intelligent asset management, greatly improves the efficiency and accuracy of fixed asset inspections, and solves the technical problem of inaccurate inspection data in the fixed asset management of financial institutions in related technologies.

[0126] Specifically, the asset inspection method proposed in this embodiment first uses intelligent control of inspection equipment to autonomously navigate to the target area based on inspection task planning and environmental map. This process relies on SLAM technology using LiDAR and visual sensors to ensure accurate equipment positioning and path planning. Subsequently, the multi-sensor array on the inspection equipment works collaboratively to collect comprehensive information on each asset device within the target area. This not only acquires the physical identification and location information of the equipment but also monitors its operating status, such as temperature, indicator light status, and other operational indicators, achieving deep perception of asset information. After data collection, the inspection equipment performs preliminary analysis of the multimodal perception data and then uses a fusion algorithm to extract and compare features from data from different sensors to obtain the asset's identity information and operational status assessment results. This fusion analysis process, combined with machine learning models and a business rule engine, can effectively identify whether there are any anomalies in the assets. Finally, the multimodal perception data and fusion analysis results are packaged into asset inventory results and uploaded to a cloud platform for storage and further analysis. The cloud platform manages the entire lifecycle of the assets and provides risk warnings, realizing intelligent and remote asset management.

[0127] The invention will now be described in conjunction with another alternative embodiment.

[0128] Example 3

[0129] This invention also provides an asset inspection device. It should be noted that the asset inspection device of this invention includes multiple implementation components, which can be used to execute the asset inspection method provided in the first embodiment above, or as the execution terminal of the asset inspection system provided in the second embodiment above.

[0130] Figure 4 This is a schematic diagram of an optional asset inspection device according to an embodiment of the present invention, such as... Figure 4 As shown, the asset inspection equipment may include:

[0131] Among them, the equipment body 41.

[0132] The mobile chassis 42 is located at the bottom of the equipment body 41 and is used to drive the equipment body 41 to move autonomously.

[0133] The multimodal sensing module 43, which is mounted on the device body 42, includes an RFID reader, a vision sensor, a lidar, and an infrared thermal imager, and is used to collect multimodal sensing data.

[0134] The communication module 44 is used to exchange data with external systems, receive inspection tasks, and upload asset inventory data.

[0135] The control module 45 is located inside the device body 41 and is electrically connected to the multimodal sensing module 43 and the mobile chassis 42. It is used to control the movement of the device, sense data, and communicate data.

[0136] The aforementioned asset inspection equipment is used to perform at least the following steps:

[0137] Step S201: Determine the inspection path based on the inspection task and the environmental map, and control the target equipment to move within the target area according to the inspection path. The target equipment is an asset inspection device equipped with N sensors, where N is a positive integer.

[0138] Step S202: During the movement, the target area is perceived by the sensors on the target device to obtain multimodal perception data. The multimodal perception data includes: equipment information, spatial location information and operation index data of each asset device in the target area.

[0139] Optionally, in the asset inspection method provided in this embodiment of the invention, the multimodal perception module includes: a radio frequency identification (RFID) reader, a visual sensor, a lidar, and an infrared thermal imager. The step of obtaining multimodal perception data by sensing information about the target area through sensors on the target device includes: reading electronic tag information on the asset device through the RFID reader; acquiring image information and depth information of the asset device through the visual sensor to obtain device information; acquiring spatial location information of the asset device through the lidar; collecting heat distribution information of the asset device through the infrared thermal imager, and judging abnormal heating indicators or power-on operation indicators of the asset device based on the heat distribution information to obtain operation indicator data.

[0140] Optionally, in the asset inspection method provided in this embodiment of the invention, the multimodal perception module further includes: a spectral imager and a millimeter-wave radar. The step of obtaining multimodal perception data by sensing information about the target area through sensors on the target device further includes: collecting surface spectral features of the asset device through the spectral imager; and / or collecting dielectric features of the asset device through the millimeter-wave radar.

[0141] Optionally, in the asset inspection method provided in this embodiment of the invention, after obtaining multimodal perception data by sensing information about the target area through sensors on the target device, the method further includes: constructing a three-dimensional digital twin model of the target area based on the multimodal perception data; and labeling the asset inventory results as attribute tags in the three-dimensional digital twin model.

[0142] Step S203: Perform fusion analysis on the multimodal perception data to obtain analysis results, including asset identity information and operational status assessment results.

[0143] Optionally, in the asset inspection method provided in this embodiment of the invention, the step of fusing and analyzing multimodal sensing data to obtain analysis results includes: preprocessing the multimodal sensing data, wherein the preprocessing includes: noise filtering, data format standardization, and timestamp alignment; using a feature extraction algorithm to extract features from the preprocessed multimodal sensing data to obtain multimodal features, wherein the multimodal features include at least: geometric features, surface texture features, and temperature distribution features; calling an edge computing node and / or a cloud platform to fuse the multimodal features to obtain fused features, wherein both the edge computing node and the cloud platform are connected to a historical database; determining the asset identity information of each asset device within the target area based on the fused features, and assessing whether the operating status of each asset device is abnormal to obtain analysis results.

[0144] Step S204: Package the multimodal sensing data and analysis results into asset inventory results, and upload the asset inventory results to the cloud platform for storage.

[0145] Optionally, in the asset inspection method provided in the embodiments of the present invention, after packaging the multimodal perception data and analysis results into asset inventory results, the method further includes: comparing the asset inventory results obtained in this inspection with historical data, and generating alarm information and pushing the alarm information if asset loss, location change or abnormal operation status is identified during the comparison process.

[0146] In this embodiment of the invention, an approach integrating artificial intelligence and Internet of Things technologies is adopted. By equipping intelligent inspection equipment with multimodal perception modules and autonomous decision-making algorithms, the goal of automatically and accurately collecting comprehensive information on the internal assets of financial institutions such as banks is achieved. This realizes the technical effect of intelligent asset management, greatly improves the efficiency and accuracy of fixed asset inspections, and solves the technical problem of inaccurate inspection data in the fixed asset management of financial institutions in related technologies.

[0147] Specifically, the asset inspection method proposed in this embodiment first uses intelligent control of inspection equipment to autonomously navigate to the target area based on inspection task planning and environmental map. This process relies on SLAM technology using LiDAR and visual sensors to ensure accurate equipment positioning and path planning. Subsequently, the multi-sensor array on the inspection equipment works collaboratively to collect comprehensive information on each asset device within the target area. This not only acquires the physical identification and location information of the equipment but also monitors its operating status, such as temperature, indicator light status, and other operational indicators, achieving deep perception of asset information. After data collection, the inspection equipment performs preliminary analysis of the multimodal perception data and then uses a fusion algorithm to extract and compare features from data from different sensors to obtain the asset's identity information and operational status assessment results. This fusion analysis process, combined with machine learning models and a business rule engine, can effectively identify whether there are any anomalies in the assets. Finally, the multimodal perception data and fusion analysis results are packaged into asset inventory results and uploaded to a cloud platform for storage and further analysis. The cloud platform manages the entire lifecycle of the assets and provides risk warnings, realizing intelligent and remote asset management.

[0148] The invention will now be described in conjunction with another alternative embodiment.

[0149] Example 4

[0150] This invention also provides an asset inspection device. It should be noted that the asset inspection device of this invention includes multiple implementation units, which can be used to execute the asset inspection method provided in the first embodiment above. Each implementation unit corresponds to each implementation step in the first embodiment above.

[0151] Figure 5 This is a schematic diagram of an optional asset inspection device according to an embodiment of the present invention, such as... Figure 5 As shown, the device may include: a determination unit 51, a sensing unit 52, an analysis unit 53, and an uploading unit 54.

[0152] The determining unit 51 is used to determine the inspection path based on the inspection task and the environmental map, and control the target equipment to move within the target area according to the inspection path. The target equipment is an asset inspection device equipped with N sensors, where N is a positive integer.

[0153] The sensing unit 52 is used to perceive information about the target area through sensors on the target device during the movement process, and obtain multimodal sensing data. The multimodal sensing data includes: equipment information, spatial location information and operation index data of each asset device in the target area.

[0154] Analysis unit 53 is used to perform fusion analysis on multimodal perception data to obtain analysis results, including asset identity information and operational status assessment results.

[0155] Upload unit 54 is used to package multimodal sensing data and analysis results into asset inventory results and upload the asset inventory results to the cloud platform for storage.

[0156] The aforementioned asset inspection device can first determine the inspection path based on the inspection task and environmental map through the determination unit 51, and control the target equipment to move within the target area according to the inspection path. The target equipment is an asset inspection device equipped with N sensors, where N is a positive integer. Then, during the movement, the perception unit 52 uses the sensors on the target equipment to perceive information about the target area and obtain multimodal perception data. The multimodal perception data includes: equipment information, spatial location information, and operational indicator data of each asset device within the target area. Then, the analysis unit 53 performs fusion analysis on the multimodal perception data to obtain analysis results. The analysis results include: asset identity information and operational status assessment results. Finally, the upload unit 54 packages the multimodal perception data and analysis results into asset inventory results and uploads the asset inventory results to the cloud platform for storage.

[0157] In this embodiment of the invention, by integrating artificial intelligence and Internet of Things technologies, and by using intelligent inspection equipment equipped with multimodal perception modules and autonomous decision-making algorithms, the goal of automatically and accurately collecting comprehensive information on the internal assets of financial institutions such as banks is achieved. This realizes the technical effect of intelligent asset management, greatly improves the efficiency and accuracy of fixed asset inspections, and solves the technical problem of inaccurate inspection data in the fixed asset management of financial institutions in related technologies.

[0158] Specifically, the asset inspection method proposed in this embodiment first uses intelligent control of inspection equipment to autonomously navigate to the target area based on inspection task planning and environmental map. This process relies on SLAM technology using LiDAR and visual sensors to ensure accurate equipment positioning and path planning. Subsequently, the multi-sensor array on the inspection equipment works collaboratively to collect comprehensive information on each asset device within the target area. This not only acquires the physical identification and location information of the equipment but also monitors its operating status, such as temperature, indicator light status, and other operational indicators, achieving deep perception of asset information. After data collection, the inspection equipment performs preliminary analysis of the multimodal perception data and then uses a fusion algorithm to extract and compare features from data from different sensors to obtain the asset's identity information and operational status assessment results. This fusion analysis process, combined with machine learning models and a business rule engine, can effectively identify whether there are any anomalies in the assets. Finally, the multimodal perception data and fusion analysis results are packaged into asset inventory results and uploaded to a cloud platform for storage and further analysis. The cloud platform manages the entire lifecycle of the assets and provides risk warnings, realizing intelligent and remote asset management.

[0159] Furthermore, the multimodal perception module includes: an RFID reader, a vision sensor, a LiDAR, and an infrared thermal imager. The perception unit includes: a reading module, used to read electronic tag information on the asset equipment through the RFID reader and acquire image and depth information of the asset equipment through the vision sensor to obtain equipment information; an acquisition module, used to acquire spatial location information of the asset equipment through the LiDAR; and a judgment module, used to collect thermal distribution information of the asset equipment through the infrared thermal imager and judge abnormal heating indicators or power-on operation indicators of the asset equipment based on the thermal distribution information to obtain operation indicator data.

[0160] Furthermore, the multimodal sensing module also includes: a spectral imager and a millimeter-wave radar. The sensing unit also includes: a first acquisition module for acquiring the surface spectral characteristics of the asset equipment through the spectral imager; and a second acquisition module for acquiring the dielectric characteristics of the asset equipment through the millimeter-wave radar.

[0161] Furthermore, the analysis unit includes: a preprocessing module for preprocessing the multimodal sensing data, wherein the preprocessing includes noise filtering, data format standardization, and timestamp alignment; an extraction module for using feature extraction algorithms to extract features from the preprocessed multimodal sensing data to obtain multimodal features, wherein the multimodal features include at least geometric features, surface texture features, and temperature distribution features; a fusion module for calling edge computing nodes and / or cloud platforms to fuse the multimodal features to obtain fused features, wherein both the edge computing nodes and the cloud platform are connected to a historical database; and an evaluation module for determining the asset identity information of each asset device within the target area based on the fused features, and evaluating whether the operating status of each asset device is abnormal to obtain the analysis results.

[0162] Furthermore, the asset inspection device also includes a comparison module, which, after packaging the multimodal sensing data and analysis results into asset inventory results, compares the asset inventory results obtained in this inspection with historical data. If asset loss, location change, or abnormal operating status is detected during the comparison process, alarm information is generated and pushed to the relevant departments.

[0163] Furthermore, the asset inspection device also includes: a construction module, used to construct a three-dimensional digital twin model of the target area based on the multimodal perception data obtained by sensing the target area through sensors on the target device; and an annotation module, used to annotate the asset inventory results as attribute labels in the three-dimensional digital twin model.

[0164] It should be noted that the aforementioned determining unit 51, sensing unit 52, analysis unit 53, and uploading unit 54 correspond to steps S201 to S204 in Embodiment 1. The instances and application scenarios implemented by the aforementioned units and corresponding steps are the same, but are not limited to the content disclosed in Embodiment 1. It should be noted that the aforementioned modules or units may be hardware or software components stored in a memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The aforementioned modules or units may also be part of a device and can run in the computer terminal 10 provided in Embodiment 1.

[0165] The invention will now be described in conjunction with another alternative embodiment.

[0166] Example 5

[0167] The present invention can also provide an electronic device. Figure 6 This is a structural block diagram of an electronic device for performing an asset inspection method according to an embodiment of the present invention, as shown below. Figure 6 As shown, the electronic device may include: one or more ( Figure 6 (Only one is shown) processor 602, memory 604, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0168] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the asset inspection method and apparatus in this application embodiment. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the aforementioned asset inspection method. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0169] The processor can access information and applications stored in memory via a transmission device to execute the following steps: Determine the inspection path based on the inspection task and environmental map, and control the target equipment to move within the target area along the inspection path. The target equipment is an asset inspection device equipped with N sensors, where N is a positive integer. During movement, the sensors on the target equipment perceive information about the target area, obtaining multimodal perception data. This multimodal perception data includes: equipment information, spatial location information, and operational indicator data for each asset within the target area. The multimodal perception data is then fused and analyzed to obtain analysis results, which include: asset identity information and operational status assessment results. The multimodal perception data and analysis results are packaged into an asset inventory result, which is then uploaded to a cloud platform for storage.

[0170] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: read the electronic tag information on the asset equipment through the radio frequency identification card reader, obtain the image information and depth information of the asset equipment through the vision sensor to obtain the equipment information; obtain the spatial location information of the asset equipment through the lidar; collect the heat distribution information of the asset equipment through the infrared thermal imager, and judge the abnormal heating index or power-on operation index of the asset equipment based on the heat distribution information to obtain the operation index data.

[0171] The processor can also invoke information and applications stored in the memory via a transmission device to perform the following steps: acquiring surface spectral characteristics of the asset equipment via a spectral imager; and / or acquiring dielectric characteristics of the asset equipment via millimeter-wave radar.

[0172] The processor can also access information and applications stored in the memory via a transmission device to perform the following steps: preprocessing the multimodal sensing data, including noise filtering, data format standardization, and timestamp alignment; extracting features from the preprocessed multimodal sensing data using a feature extraction algorithm to obtain multimodal features, which at least include geometric features, surface texture features, and temperature distribution features; fusing the multimodal features using edge computing nodes and / or a cloud platform to obtain fused features, wherein both the edge computing nodes and the cloud platform are connected to a historical database; determining the asset identity information of each asset device within the target area based on the fused features, and assessing whether the operating status of each asset device is abnormal to obtain analysis results.

[0173] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: compare the asset inventory results obtained in this inspection with historical data, and generate alarm information and push the alarm information if asset loss, location change or abnormal operation is detected during the comparison process.

[0174] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: construct a three-dimensional digital twin model of the target area based on multimodal sensing data; and label the asset inventory results as attribute tags in the three-dimensional digital twin model.

[0175] This invention provides an asset inspection solution. By integrating artificial intelligence and Internet of Things technologies, and using intelligent inspection equipment equipped with multimodal sensing modules and autonomous decision-making algorithms, it achieves the goal of automatically and accurately collecting comprehensive information on assets within financial institutions such as banks. This realizes the technical effect of intelligent asset management, greatly improves the efficiency and accuracy of fixed asset inspections, and solves the technical problem of inaccurate inspection data in the fixed asset management of financial institutions in related technologies.

[0176] Those skilled in the art will understand that Figure 6 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 6 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 6 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 6 The different configurations shown.

[0177] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0178] The invention will now be described in conjunction with another alternative embodiment.

[0179] Example 6

[0180] This invention also provides a computer-readable storage medium. Optionally, in this invention, the computer-readable storage medium can be used to store the program code executed by the asset inspection method provided in Embodiment 1.

[0181] Optionally, in this embodiment of the invention, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0182] This invention also provides a computer program product, which, when executed on a data processing device, is suitable for performing the steps of an asset inspection method: determining an inspection path based on the inspection task and an environmental map, and controlling a target device to move within a target area along the inspection path, wherein the target device is an asset inspection device equipped with N sensors, where N is a positive integer; during the movement, the target area is perceived through the sensors on the target device to obtain multimodal perception data, wherein the multimodal perception data includes: equipment information, spatial location information, and operational indicator data of each asset device within the target area; the multimodal perception data is fused and analyzed to obtain analysis results, wherein the analysis results include: asset identity information and operational status assessment results; the multimodal perception data and analysis results are packaged into an asset inventory result, and the asset inventory result is uploaded to a cloud platform for storage.

[0183] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0184] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0185] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and 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 displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0186] The units described 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.

[0187] Furthermore, the functional units in the various embodiments of this application 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.

[0188] 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 application, 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 several 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 described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0189] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. An asset inspection method, characterized in that, include: The inspection path is determined based on the inspection task and the environmental map, and the target equipment is controlled to move within the target area according to the inspection path. The target equipment is an asset inspection equipment equipped with N sensors, where N is a positive integer. During the movement, the target area is perceived by sensors on the target device to obtain multimodal perception data, which includes: equipment information, spatial location information and operation index data of each asset device in the target area; The multimodal sensing data is fused and analyzed to obtain analysis results, which include asset identity information and operational status assessment results. The multimodal sensing data and the analysis results are packaged into an asset inventory result, and the asset inventory result is uploaded to the cloud platform for storage.

2. The asset inspection method according to claim 1, characterized in that, The multimodal sensing module includes: an RFID reader, a visual sensor, a LiDAR, and an infrared thermal imager. The step of obtaining multimodal sensing data by sensing the target area through sensors on the target device includes: The electronic tag information on the asset device is read by the radio frequency identification card reader, and the image information and depth information of the asset device are obtained by the vision sensor to obtain the device information. The spatial location information of the asset equipment is obtained through the lidar. The infrared thermal imager collects the thermal distribution information of the asset equipment, and based on the thermal distribution information, it determines the abnormal heating index or power-on operation index of the asset equipment to obtain the operation index data.

3. The asset inspection method according to claim 1, characterized in that, The multimodal sensing module further includes: a spectral imager and a millimeter-wave radar. The step of obtaining multimodal sensing data by sensing the target area through sensors on the target device further includes: The surface spectral characteristics of the asset equipment are acquired using the spectral imager; and / or... The dielectric characteristics of the asset equipment are collected using the millimeter-wave radar.

4. The asset inspection method according to claim 1, characterized in that, The steps for fusing and analyzing the multimodal sensing data to obtain the analysis results include: The multimodal sensing data is preprocessed, wherein the preprocessing includes: noise filtering, data format standardization, and timestamp alignment; The preprocessed multimodal sensing data is subjected to feature extraction algorithms to obtain multimodal features, wherein the multimodal features include at least: geometric features, surface texture features and temperature distribution features; The multimodal features are fused by calling edge computing nodes and / or cloud platforms to obtain fused features, wherein both the edge computing nodes and the cloud platform are connected to a historical database; Based on the fusion features, the asset identity information of each asset device within the target area is determined, and the operational status of each asset device is assessed to determine whether it is abnormal, thereby obtaining the analysis results.

5. The asset inspection method according to claim 1, characterized in that, After packaging the multimodal sensing data and the analysis results into asset inventory results, the method further includes: The asset inventory results obtained from this inspection are compared with historical data. If any missing assets, changes in location, or abnormal operating status are detected during the comparison process, an alarm message is generated and pushed to the relevant authorities.

6. The asset inspection method according to claim 1, characterized in that, After obtaining multimodal sensing data by sensing the target area through sensors on the target device, the method further includes: Based on the multimodal sensing data, a three-dimensional digital twin model of the target area is constructed; The asset inventory results are labeled as attribute tags in the three-dimensional digital twin model.

7. An asset inspection system, characterized in that, The asset inspection system is used to perform the asset inspection method as described in any one of claims 1 to 6, and includes: The cloud platform is used to generate and distribute inspection tasks, and to receive, store and analyze asset inventory results. At least one edge computing node, deployed locally and connected to the cloud platform, is used to request inspections from the cloud platform and send the local environmental map, wherein "local" refers to the asset inspection site; At least one asset inspection device is communicatively connected to the cloud platform and the edge computing node, and is used to perform the inspection task within the target area; The asset inspection equipment includes: An autonomous navigation module is used to perform movement control based on the inspection task and the environmental map; The multimodal sensing module is used to collect multimodal sensing data; The data processing module is used to generate asset inventory results based on the multimodal perception data and upload them to the cloud platform.

8. The asset inspection system according to claim 7, characterized in that, The multimodal sensing module includes: Radio frequency identification (RFID) reader, used to read electronic tag information on the asset equipment; A visual sensor is used to acquire image and depth information of the asset equipment; LiDAR is used to acquire the spatial location information of the asset equipment; An infrared thermal imager is used to collect thermal distribution information of the assets and equipment.

9. The asset inspection system according to claim 8, characterized in that, The multimodal sensing module also includes: A spectral imager is used to acquire the surface spectral characteristics of the asset equipment; and / or, Millimeter-wave radar is used to collect the dielectric characteristics of the asset equipment.

10. The asset inspection system according to claim 7, characterized in that, The edge computing node is equipped with an edge computing module for running a lightweight algorithm model to process the received data, and the cloud platform is equipped with a data analysis module for running an artificial intelligence algorithm model to mine and analyze the received data.

11. The asset inspection system according to claim 10, characterized in that, The system is also used for: In the edge computing module and / or the data analysis module, the asset inventory results obtained from this inspection are compared with historical data. If asset loss, location change, or abnormal status is detected during the comparison process, alarm information is generated and pushed to the relevant departments.

12. The asset inspection system according to claim 7, characterized in that, The system also includes: The digital twin module is communicatively connected to the asset inspection equipment and is used to construct a three-dimensional digital twin model of the target area based on the multimodal perception data collected by the asset inspection equipment; and to annotate the asset inventory results as attribute labels in the three-dimensional digital twin model.

13. The asset inspection system according to claim 7, characterized in that, The asset inspection equipment is a wheeled, legged, or wheeled-legged hybrid mobile robot.

14. An asset inspection device, characterized in that, The asset inspection device is used to perform the asset inspection method as described in any one of claims 1 to 6, or as the execution terminal of the asset inspection system as described in any one of claims 7 to 13, the asset inspection device comprising: Equipment body; A mobile chassis is installed at the bottom of the device body to drive the device body to move autonomously; A multimodal sensing module, disposed on the device body, includes an RFID reader, a visual sensor, a lidar, and an infrared thermal imager, for collecting multimodal sensing data; The communication module is used to exchange data with external systems, receive inspection tasks, and upload asset inventory data. The control module, located within the device body, is electrically connected to the multimodal sensing module and the mobile chassis, and is used to control the device's movement, sense data, and communicate data.

15. An asset inspection device, characterized in that, include: The determining unit is used to determine the inspection path based on the inspection task and the environmental map, and control the target equipment to move within the target area according to the inspection path, wherein the target equipment is an asset inspection equipment equipped with N sensors, and N is a positive integer; The sensing unit is used to sense information about the target area through sensors on the target device during movement, and obtain multimodal sensing data, wherein the multimodal sensing data includes: equipment information, spatial location information and operational indicator data of each asset device in the target area; An analysis unit is used to perform fusion analysis on the multimodal perception data to obtain analysis results, wherein the analysis results include: asset identity information and operational status assessment results; The upload unit is used to package the multimodal sensing data and the analysis results into an asset inventory result, and upload the asset inventory result to the cloud platform for storage.

16. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the asset inspection method according to any one of claims 1 to 6.

17. An electronic device, characterized in that, It includes one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the asset inspection method according to any one of claims 1 to 6.

18. A computer program product, characterized in that, It includes computer instructions, wherein when the computer instructions are executed by a processor, they implement the steps of the asset inspection method according to any one of claims 1 to 6.

Images