An automatic configuration method for the AI ​​computing module of an embedded edge smart gateway

Abstract

This invention discloses an automatic configuration method for the AI ​​computing module of an embedded edge smart gateway. By detecting the manufacturer ID and device ID of the AI ​​computing module, the method automatically downloads the corresponding driver and SDK software package and installation script from a remote server. After the download is completed, the main control processor automatically runs the installation script to complete the installation of the AI ​​computing module driver and SDK software package. During the installation process, the user does not need to manually download and install the AI ​​computing module driver and SDK software package. This solves the problem that replacing the AI ​​computing module in an edge smart gateway is complicated and requires operators to have sufficient professional knowledge. This allows the edge smart gateway to easily replace the AI ​​computing module and quickly respond to the application scenarios that require different AI computing capabilities.

Description

Technical Field

[0001] This invention relates to a configuration method for an embedded edge smart gateway, and more particularly to an automatic configuration method for the AI ​​computing module of an embedded edge smart gateway. Background Technology

[0002] With the rapid development of the Internet of Things, 5G and artificial intelligence, the demand for edge smart gateways with edge AI processing capabilities and high-speed communication capabilities is increasing.

[0003] Currently, most manufacturers' edge smart gateways use a hardware architecture with a SoC chip or module equipped with AI computing capabilities as the main control core, along with data acquisition and control modules and network communication modules. The AI ​​computing power of this type of edge smart gateway is determined by the SoC chip or module, making it suitable only for application scenarios with specific AI computing power requirements. For application scenarios requiring higher AI computing power, this type of edge smart gateway can only cope by redeveloping both hardware and software, making it unable to respond promptly to new application scenario demands. For application scenarios requiring only lower AI computing power, the cost and power consumption of this type of edge smart gateway are relatively high, increasing costs for users.

[0004] A few manufacturers' edge smart gateways use a high-performance embedded processor as the main controller, with AI computing handled by an embedded AI computing module. The main controller and the embedded AI computing module are connected via interfaces such as USB, Mini-PCIe, M.2, and PCIe, along with data acquisition and control modules and network communication modules. The AI ​​computing capability of this type of edge smart gateway is determined by the embedded AI computing module, but the corresponding driver and SDK software package for the embedded AI computing module must be installed on the main controller before AI algorithm models and AI applications can run. For application scenarios requiring different AI computing capabilities, the embedded AI computing module can be replaced with one from a manufacturer / model with the corresponding AI computing capabilities. However, users need to manually download the corresponding driver and SDK software package for the replaced embedded AI computing module and manually install it on the main controller to run different AI algorithm models and AI applications. The entire replacement process is complex, requires operators to have sufficient professional knowledge, is cumbersome for users, and carries the risk of failure. Summary of the Invention

[0005] Purpose of the invention: To provide an automatic configuration method for the AI ​​computing module of an embedded edge smart gateway, which can automatically configure and install the module after the user replaces it with a new AI computing module as needed.

[0006] Technical solution: The automatic configuration method for the AI ​​computing module of the embedded edge intelligent gateway described in this invention includes the following steps:

[0007] Step 1: The user stores the drivers, SDK packages and installation scripts for different AI computing modules on the remote server, and generates the index number corresponding to each AI computing module.

[0008] Step 2: Based on the computing power required, the user replaces the corresponding new AI computing module on the edge smart gateway and keeps the edge smart gateway running after the replacement.

[0009] Step 3: The main control processor of the edge smart gateway detects in real time whether the new AI computing module has been successfully replaced. If the replacement is successful, proceed to step 4. If the replacement is unsuccessful, send a replacement failure message to the user and return to step 2.

[0010] Step 4: The main control processor of the edge smart gateway checks whether the driver and SDK package for the new AI computing module have been installed. If not, proceed to step 5. If already installed, send a message to the user that the new AI computing module has been successfully configured and end the automatic configuration.

[0011] Step 5: The main control processor of the edge intelligent gateway generates the corresponding storage query request based on the AI ​​computing module and sends the storage query request to the remote server.

[0012] Step 6: The remote server queries whether the driver, SDK package and installation script of the new AI computing module are stored according to the storage query request, and sends the query result information to the main control processor of the edge smart gateway.

[0013] Step 7: The main control processor of the edge smart gateway obtains the query results to determine whether there are drivers, SDK packages and installation scripts for the new AI computing module available for download on the remote server. If they are available for download, proceed to step 8. If they are not available for download, send an alarm message to the user indicating an installation error of the new AI computing module and return to step 1.

[0014] Step 8: Establish a download channel between the edge smart gateway and the remote server. The main control processor of the edge smart gateway downloads the driver, SDK package and installation script of the new AI computing module from the remote server through the download channel.

[0015] Step 9: The main control processor of the edge smart gateway runs the installation script to install the downloaded new AI computing module driver and SDK package, and generates the corresponding installation success flag after successful installation.

[0016] Furthermore, in step 1, the specific steps for storing the drivers, SDK packages, and installation scripts for different AI computing modules are as follows:

[0017] Step 1.1: Package the driver, SDK package and installation script of each AI computing module into an installation package, and then store each installation package on a remote server;

[0018] Step 1.2: Check the integrity of each installation package. If each installation package contains the complete driver, SDK package and installation script, proceed to step 1.3. Otherwise, report the storage failure to the user and return to step 1.1.

[0019] Step 1.3: Use the manufacturer ID and device ID of each stored AI computing module as the index number of the installation package of each AI computing module on the remote server.

[0020] Furthermore, in step 2, when replacing the new AI computing module, if the edge smart gateway is powered on, the new AI computing module is replaced directly while it is powered on. If the edge smart gateway is powered off, the edge smart gateway is powered on and kept powered on after the new AI computing module is replaced.

[0021] Furthermore, in step 3, when detecting whether the new AI computing module has been successfully replaced, if the main control processor obtains the manufacturer ID and device ID of the new AI computing module, it is determined that the replacement was successful; otherwise, it is determined that the replacement was unsuccessful.

[0022] Furthermore, in step 4, when checking whether the driver and SDK package for the new AI computing module have been installed, it checks whether there is a successful installation flag for the new AI computing module. If there is a successful installation flag, it means that it has been installed; otherwise, it means that it has not been installed.

[0023] Furthermore, in step 5, the generated storage query request includes the edge smart gateway ID and the index number.

[0024] Furthermore, in step 6, the specific steps for the remote server to query based on the storage query request are as follows:

[0025] Step 6.1: Parse the storage query request to obtain the edge smart gateway ID and index number;

[0026] Step 6.2: Query the remote server according to the index number. If the corresponding index number exists, it means that the driver, SDK package and installation script of the new AI computing module are stored in the remote server, and the query result is "stored". Otherwise, it means that it is not stored, and the query result is "not stored".

[0027] Step 6.3: Generate query result information based on the query results. The query result information includes the index number and the query result. Then, send the query result information to the corresponding edge smart gateway based on the edge smart gateway ID.

[0028] Furthermore, in step 7, the specific steps for providing the user with an alarm message indicating an installation error of the new AI computing module are as follows:

[0029] Step 7.1: Generate alarm information for installation errors. The alarm information includes a message indicating that no resources are being downloaded and a message indicating that resources are being uploaded.

[0030] Step 7.2: Send the "No Downloaded Resources" message and the "Resource Upload" message to the user's registered email address. The resource upload message includes the ID information of the resource to be uploaded and the resource content. The ID information includes the name of the new AI computing module, the manufacturer ID, and the device ID. The resource content includes the driver, SDK package, and installation script of the new AI computing module.

[0031] Furthermore, in step 8, the specific steps for the edge smart gateway's main control processor to download data from the remote server via the download channel are as follows:

[0032] Step 8.1: The main control processor sends a download instruction to the remote server through the download channel. The download instruction includes the index number and the edge smart gateway ID.

[0033] Step 8.2: The remote server finds the installation package of the corresponding new AI computing module according to the index number, and sends the installation package to the edge smart gateway corresponding to the edge smart gateway ID through the download channel;

[0034] Step 8.3: The main control processor decompresses the sent installation package to obtain the driver, SDK package, and installation script for the new AI computing module;

[0035] Step 8.4: Perform an integrity check on the decompressed driver, SDK package, and installation script. If the driver, SDK package, and installation script are all complete and valid, proceed to step 9; otherwise, proceed to step 8.5.

[0036] Step 8.5: Determine if this is the first time decompressing. If it is, return to step 8.3; otherwise, proceed to step 8.6.

[0037] Step 8.6: Determine if this is the first download. If it is, return to step 8.1. Otherwise, provide the user with a message indicating that the download of the new AI computing module configuration software failed, and include the name, manufacturer ID, and device ID of the new AI computing module in the download failure message.

[0038] Furthermore, in step 9, the specific steps for installing the downloaded new AI computing module driver and SDK package are as follows:

[0039] Step 9.1: Construct a test dataset, which includes test data for each AI computing module and the corresponding test results;

[0040] Step 9.2: Control the installation of the new AI computing module driver and SDK software package according to the installation script;

[0041] Step 9.3: After installation, select the corresponding set of test data from the test dataset based on the name, manufacturer ID, and device ID of the new AI computing module.

[0042] Step 9.4: Run the new AI computing module using the selected test data and output the corresponding test results;

[0043] Step 9.5: Compare the output test results with the corresponding test results in the test dataset. If the test results are consistent, proceed to step 9.6; otherwise, it indicates that the installation has failed and proceed to step 9.7.

[0044] Step 9.6: Determine if this is the first test. If it is the first test, select another set of test data from the test dataset and return to step 9.4. If it is not the first test, it indicates that the installation was successful and proceed to step 9.8.

[0045] Step 9.7: Determine if this is the first installation. If it is, return to step 9.2; otherwise, proceed to step 9.8.

[0046] Step 9.8: Provide the user with feedback on whether the configuration and installation of the new AI computing module has failed or succeeded.

[0047] Compared with existing technologies, the advantages of this invention are as follows: The automatic configuration method of the AI ​​computing module of the embedded edge smart gateway of this invention detects the manufacturer ID and device ID of the AI ​​computing module, and automatically downloads the corresponding driver and SDK software package and installation script from a remote server. After the download is completed, the main control processor automatically runs the installation script to complete the installation of the AI ​​computing module driver and SDK software package. During the installation process, the user does not need to manually download and install the AI ​​computing module driver and SDK software package, which solves the problem that the operation is complicated and requires the operator to have sufficient professional knowledge when replacing the AI ​​computing module of the edge smart gateway. This makes it easy for the edge smart gateway to replace the AI ​​computing module and quickly respond to the application scenarios that require different AI computing capabilities. Attached Figure Description

[0048] Figure 1 This is a flowchart of the method of the present invention;

[0049] Figure 2 This is a diagram showing the connection relationship between the main control processor and the AI ​​computing module of the present invention. Detailed Implementation

[0050] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the embodiments described.

[0051] Example 1:

[0052] like Figure 1 As shown, the automatic configuration method for the AI ​​computing module of the embedded edge smart gateway disclosed in this invention includes the following steps:

[0053] Step 1: The user stores the drivers, SDK packages and installation scripts for different AI computing modules on the remote server, and generates the index number corresponding to each AI computing module.

[0054] Step 2: Based on the computing power required, the user replaces the corresponding new AI computing module on the edge smart gateway and keeps the edge smart gateway running after the replacement.

[0055] Step 3: The main control processor of the edge smart gateway detects in real time whether the new AI computing module has been successfully replaced. If the replacement is successful, proceed to step 4. If the replacement is unsuccessful, send a replacement failure message to the user and return to step 2.

[0056] Step 4: The main control processor of the edge smart gateway checks whether the driver and SDK package for the new AI computing module have been installed. If not, proceed to step 5. If already installed, send a message to the user that the new AI computing module has been successfully configured and end the automatic configuration.

[0057] Step 5: The main control processor of the edge intelligent gateway generates the corresponding storage query request based on the AI ​​computing module and sends the storage query request to the remote server.

[0058] Step 6: The remote server queries whether the driver, SDK package and installation script of the new AI computing module are stored according to the storage query request, and sends the query result information to the main control processor of the edge smart gateway.

[0059] Step 7: The main control processor of the edge smart gateway obtains the query results to determine whether there are drivers, SDK packages and installation scripts for the new AI computing module available for download on the remote server. If they are available for download, proceed to step 8. If they are not available for download, send an alarm message to the user indicating an installation error of the new AI computing module and return to step 1.

[0060] Step 8: Establish a download channel between the edge smart gateway and the remote server. The main control processor of the edge smart gateway downloads the driver, SDK package and installation script of the new AI computing module from the remote server through the download channel.

[0061] Step 9: The main control processor of the edge smart gateway runs the installation script to install the downloaded new AI computing module driver and SDK package, and generates the corresponding installation success flag after successful installation.

[0062] When users replace AI computing modules with those from different manufacturers / models, the main control processor in the edge smart gateway automatically downloads the corresponding driver, SDK package, and installation script from a remote server by detecting the manufacturer ID and device ID of the inserted AI computing module. Once the download is complete, the main control processor automatically runs the installation script to install the AI ​​computing module driver and SDK package. This eliminates the need for users to manually download and install the AI ​​computing module driver and SDK package, solving the problem of complex operations and the requirement for operators with sufficient professional knowledge when replacing AI computing modules in edge smart gateways. This allows edge smart gateways to easily replace AI computing modules and quickly respond to application scenarios requiring different AI computing capabilities.

[0063] Furthermore, in step 1, the specific steps for storing the drivers, SDK packages, and installation scripts for different AI computing modules are as follows:

[0064] Step 1.1: Package the driver, SDK package and installation script of each AI computing module into an installation package, and then store each installation package on a remote server;

[0065] Step 1.2: Check the integrity of each installation package. If each installation package contains the complete driver, SDK package and installation script, proceed to step 1.3. Otherwise, report the storage failure to the user and return to step 1.1.

[0066] Step 1.3: Use the manufacturer ID and device ID of each stored AI computing module as the index number of the installation package of each AI computing module on the remote server. Specifically, take the first four digits of the manufacturer ID and device ID to form an eight-digit numeric string. If the manufacturer ID or device ID is less than four digits, add zeros to the end of the corresponding string to form a four-digit numeric string.

[0067] Storing packages as installation packages facilitates storage management and enables one-time remote downloads. Integrity checks on these packages ensure their completeness, guaranteeing that edge smart gateways receive the full package and increasing the first-time configuration success rate. Using index numbers facilitates quick retrieval of installation packages for various AI computing modules, improving the configuration efficiency of new AI computing modules.

[0068] Furthermore, in step 2, when replacing the AI ​​computing module, if the edge smart gateway is powered on, the replacement is performed directly while it is powered on. If the edge smart gateway is powered off, it is powered on and kept powered on after the replacement is completed. By allowing the replacement of the new AI computing module in either powered-on or powered-off state, the replacement efficiency is effectively improved, and maintaining a powered-on state after replacement ensures the normal operation of subsequent automatic configuration.

[0069] Furthermore, in step 3, when checking whether the new AI computing module has been successfully replaced, if the main control processor obtains the manufacturer ID and device ID of the new AI computing module, it is determined that the replacement was successful; otherwise, it is determined that the replacement was unsuccessful. By checking whether the manufacturer ID and device ID can be read, it is determined whether the new AI computing module is physically connected in place, providing a reliable guarantee for subsequent automatic configuration.

[0070] Furthermore, in step 4, when checking whether the driver and SDK package for the new AI computing module have been installed, it checks for the presence of an installation success flag for the new AI computing module. If the flag exists, it indicates that the module has been installed; otherwise, it indicates that it has not. By detecting the presence of the installation success flag, it is determined that the edge smart gateway has already been configured with the corresponding software. If it has already been configured, it can be used directly, saving the subsequent software download and installation process, avoiding repeated downloads and installations, and improving configuration efficiency.

[0071] Furthermore, in step 5, the generated storage query request includes the edge smart gateway ID and the index number. Using the edge smart gateway ID facilitates the remote server in providing query results.

[0072] Furthermore, in step 6, the specific steps for the remote server to query based on the storage query request are as follows:

[0073] Step 6.1: Parse the storage query request to obtain the edge smart gateway ID and index number;

[0074] Step 6.2: Query the remote server according to the index number. If the corresponding index number exists, it means that the driver, SDK package and installation script of the new AI computing module are stored in the remote server, and the query result is "stored". Otherwise, it means that it is not stored, and the query result is "not stored".

[0075] Step 6.3: Generate query result information based on the query results. The query result information includes the index number and the query result. Then, send the query result information to the corresponding edge smart gateway based on the edge smart gateway ID.

[0076] Using the index number, a quick search can be performed on a remote server to determine whether the corresponding driver, SDK package, and installation script are stored, facilitating the establishment of a download channel for subsequent downloads.

[0077] Furthermore, in step 7, the specific steps for providing the user with an alarm message indicating an installation error of the new AI computing module are as follows:

[0078] Step 7.1: Generate alarm information for installation errors. The alarm information includes a message indicating that no resources are being downloaded and a message indicating that resources are being uploaded.

[0079] Step 7.2: Send the "No Downloaded Resources" message and the "Resource Upload" message to the user's registered email address. The resource upload message includes the ID information of the resource to be uploaded and the resource content. The ID information includes the name of the new AI computing module, the manufacturer ID, and the device ID. The resource content includes the driver, SDK package, and installation script of the new AI computing module.

[0080] Resource upload prompts can alert users to store resource content with the corresponding ID information on a remote server when no resources are available for download, thus enabling subsequent automatic configuration.

[0081] Furthermore, in step 8, the specific steps for the edge smart gateway's main control processor to download data from the remote server via the download channel are as follows:

[0082] Step 8.1: The main control processor sends a download instruction to the remote server through the download channel. The download instruction includes the index number and the edge smart gateway ID.

[0083] Step 8.2: The remote server finds the installation package of the corresponding new AI computing module according to the index number, and sends the installation package to the edge smart gateway corresponding to the edge smart gateway ID through the download channel;

[0084] Step 8.3: The main control processor decompresses the sent installation package to obtain the driver, SDK package, and installation script for the new AI computing module;

[0085] Step 8.4: Perform an integrity check on the decompressed driver, SDK package, and installation script. If the driver, SDK package, and installation script are all complete and valid, proceed to step 9; otherwise, proceed to step 8.5.

[0086] Step 8.5: Determine if this is the first time decompressing. If it is, return to step 8.3; otherwise, proceed to step 8.6.

[0087] Step 8.6: Determine if this is the first download. If it is, return to step 8.1. Otherwise, provide the user with a message indicating that the download of the new AI computing module configuration software failed, and include the name, manufacturer ID, and device ID of the new AI computing module in the download failure message.

[0088] The judgment in step 8.5 can prevent the installation package integrity check from failing due to occasional decompression problems, thus ensuring the integrity of the installation package after decompression and providing a reliable guarantee for subsequent successful installation; the judgment in step 8.6 can prevent the installation package from being downloaded incompletely due to occasional download problems, thus further ensuring the integrity of the installation package after decompression and providing a reliable guarantee for subsequent successful installation; the download failure information can be promptly fed back to the user after multiple download failures, which is convenient for the user to perform timely maintenance.

[0089] Furthermore, in step 9, the specific steps for installing the downloaded new AI computing module driver and SDK package are as follows:

[0090] Step 9.1: Construct a test dataset, which includes test data for each AI computing module and the corresponding test results;

[0091] Step 9.2: Control the installation of the new AI computing module driver and SDK software package according to the installation script;

[0092] Step 9.3: After installation, select the corresponding set of test data from the test dataset based on the name, manufacturer ID, and device ID of the new AI computing module.

[0093] Step 9.4: Run the new AI computing module using the selected test data and output the corresponding test results;

[0094] Step 9.5: Compare the output test results with the corresponding test results in the test dataset. If the test results are consistent, proceed to step 9.6; otherwise, it indicates that the installation has failed and proceed to step 9.7.

[0095] Step 9.6: Determine if this is the first test. If it is the first test, select another set of test data from the test dataset and return to step 9.4. If it is not the first test, it indicates that the installation was successful and proceed to step 9.8.

[0096] Step 9.7: Determine if this is the first installation. If it is, return to step 9.2; otherwise, proceed to step 9.8.

[0097] Step 9.8: Provide the user with feedback on whether the configuration and installation of the new AI computing module has failed or succeeded.

[0098] Using the test dataset allows for testing of the newly installed AI computing module, ensuring the reliability of data processing after normal operation. The judgment in step 9.6 can avoid test failures caused by problems with the test data itself, so verifying the success of the installation through testing with different test data is more reliable. The judgment in step 9.7 can avoid misjudgments of installation failure caused by occasional installation problems, and the problem can be solved by trying to install again.

[0099] The embedded edge intelligent gateway of this invention consists of a main control processor, a power module, an AI computing module, a data acquisition and control module, a long-range wireless data transmission module (2G / 3G / 4G / 5G / NB-IoT / LoRa, etc.), a short-range wireless communication module (Bluetooth and WIFI), and an Ethernet WAN interface. The data acquisition and control module consists of an RS232 interface, an RS485 interface, a CAN interface, an Ethernet LAN interface, and an I / O interface. The main control processor connects to different sensing node devices (including cameras used for video surveillance) through the data acquisition and control module, connects to a server platform located in a local area network through the Ethernet LAN interface and WIFI, connects to a remote server cloud platform in a wide area network through the Ethernet WAN interface, WIFI, and the long-range wireless data transmission module, and connects to the AI ​​computing module through interfaces such as USB 2.0 / USB 3.0, Mini-PCIe, M.2, and PCIe. Figure 2 As shown.

[0100] As described above, although the invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the invention itself. Various changes in form and detail may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. An automatic configuration method for the AI ​​computing module of an embedded edge intelligent gateway, characterized in that, Includes the following steps: Step 1: The user stores the drivers, SDK packages and installation scripts for different AI computing modules on the remote server, and generates the index number corresponding to each AI computing module. Step 2: Based on the computing power required, the user replaces the corresponding new AI computing module on the edge smart gateway and keeps the edge smart gateway running after the replacement. Step 3: The main control processor of the edge smart gateway detects in real time whether the new AI computing module has been successfully replaced. If the replacement is successful, proceed to step 4. If the replacement is unsuccessful, send a replacement failure message to the user and return to step 2. Step 4: The main control processor of the edge smart gateway checks whether the driver and SDK package for the new AI computing module have been installed. If not, proceed to step 5. If already installed, send a message to the user that the new AI computing module has been successfully configured and end the automatic configuration. Step 5: The main control processor of the edge intelligent gateway generates the corresponding storage query request based on the AI ​​computing module and sends the storage query request to the remote server. Step 6: The remote server queries whether the driver, SDK package and installation script of the new AI computing module are stored according to the storage query request, and sends the query result information to the main control processor of the edge smart gateway. Step 7: The main control processor of the edge smart gateway obtains the query results to determine whether there are drivers, SDK packages and installation scripts for the new AI computing module available for download on the remote server. If they are available for download, proceed to step 8. If they are not available for download, send an alarm message to the user indicating an installation error of the new AI computing module and return to step 1. Step 8: Establish a download channel between the edge smart gateway and the remote server. The main control processor of the edge smart gateway downloads the driver, SDK package and installation script of the new AI computing module from the remote server through the download channel. Step 9: The main control processor of the edge smart gateway runs the installation script to install the downloaded new AI computing module driver and SDK package, and generates the corresponding installation success flag after successful installation.

2. The automatic configuration method for the AI ​​computing module of the embedded edge intelligent gateway according to claim 1, characterized in that, In step 1, the specific steps for storing the drivers, SDK packages, and installation scripts for different AI computing modules are as follows: Step 1.1: Package the driver, SDK package and installation script of each AI computing module into an installation package, and then store each installation package on a remote server; Step 1.2: Check the integrity of each installation package. If each installation package contains the complete driver, SDK package and installation script, proceed to step 1.

3. Otherwise, report the storage failure to the user and return to step 1.

1. Step 1.3: Use the manufacturer ID and device ID of each stored AI computing module as the index number of the installation package of each AI computing module on the remote server.

3. The automatic configuration method for the AI ​​computing module of the embedded edge intelligent gateway according to claim 1, characterized in that, In step 2, when replacing the new AI computing module, if the edge smart gateway is powered on, the replacement of the new AI computing module will be performed directly while it is powered on. If the edge smart gateway is powered off, the edge smart gateway will be powered on and kept powered on after the replacement of the new AI computing module is completed.

4. The automatic configuration method for the AI ​​computing module of the embedded edge intelligent gateway according to claim 1, characterized in that, In step 3, when checking whether the new AI computing module has been successfully replaced, if the main control processor obtains the manufacturer ID and device ID of the new AI computing module, it is determined that the replacement was successful; otherwise, it is determined that the replacement was unsuccessful.

5. The automatic configuration method for the AI ​​computing module of the embedded edge intelligent gateway according to claim 1, characterized in that, In step 4, when checking whether the driver and SDK package for the new AI computing module have been installed, check whether there is a successful installation flag for the new AI computing module. If there is a successful installation flag, it means that it has been installed; otherwise, it means that it has not been installed.

6. The automatic configuration method for the AI ​​computing module of the embedded edge intelligent gateway according to claim 1, characterized in that, In step 5, the generated storage query request includes the edge smart gateway ID and the index number.

7. The automatic configuration method for the AI ​​computing module of the embedded edge intelligent gateway according to claim 6, characterized in that, In step 6, the specific steps for the remote server to query based on the storage query request are as follows: Step 6.1: Parse the storage query request to obtain the edge smart gateway ID and index number; Step 6.2: Query the remote server according to the index number. If the corresponding index number exists, it means that the driver, SDK package and installation script of the new AI computing module are stored in the remote server, and the query result is "stored". Otherwise, it means that it is not stored, and the query result is "not stored". Step 6.3: Generate query result information based on the query results. The query result information includes the index number and the query result. Then, send the query result information to the corresponding edge smart gateway based on the edge smart gateway ID.

8. The automatic configuration method for the AI ​​computing module of the embedded edge intelligent gateway according to claim 1, characterized in that, In step 7, the specific steps for providing users with alerts about installation errors in the new AI computing module are as follows: Step 7.1: Generate alarm information for installation errors. The alarm information includes a message indicating that no resources are being downloaded and a message indicating that resources are being uploaded. Step 7.2: Send the "No Downloaded Resources" message and the "Resource Upload" message to the user's registered email address. The resource upload message includes the ID information of the resource to be uploaded and the resource content. The ID information includes the name of the new AI computing module, the manufacturer ID, and the device ID. The resource content includes the driver, SDK package, and installation script of the new AI computing module.

9. The automatic configuration method for the AI ​​computing module of the embedded edge intelligent gateway according to claim 1, characterized in that, In step 8, the specific steps for the edge smart gateway's main control processor to download data from the remote server via the download channel are as follows: Step 8.1: The main control processor sends a download instruction to the remote server through the download channel. The download instruction includes the index number and the edge smart gateway ID. Step 8.2: The remote server finds the installation package of the corresponding new AI computing module according to the index number, and sends the installation package to the edge smart gateway corresponding to the edge smart gateway ID through the download channel; Step 8.3: The main control processor decompresses the sent installation package to obtain the driver, SDK package, and installation script for the new AI computing module; Step 8.4: Perform an integrity check on the decompressed driver, SDK package, and installation script. If the driver, SDK package, and installation script are all complete and valid, proceed to step 9; otherwise, proceed to step 8.

5. Step 8.5: Determine if this is the first time decompressing. If it is, return to step 8.3; otherwise, proceed to step 8.

6. Step 8.6: Determine if this is the first download. If it is, return to step 8.

1. Otherwise, provide the user with a message indicating that the download of the new AI computing module configuration software failed, and include the name, manufacturer ID, and device ID of the new AI computing module in the download failure message.

10. The automatic configuration method for the AI ​​computing module of the embedded edge intelligent gateway according to claim 1, characterized in that, In step 9, the specific steps for installing the downloaded new AI computing module driver and SDK package are as follows: Step 9.1: Construct a test dataset, which includes test data for each AI computing module and the corresponding test results; Step 9.2: Control the installation of the new AI computing module driver and SDK software package according to the installation script; Step 9.3: After installation, select the corresponding set of test data from the test dataset based on the name, manufacturer ID, and device ID of the new AI computing module. Step 9.4: Run the new AI computing module using the selected test data and output the corresponding test results; Step 9.5: Compare the output test results with the corresponding test results in the test dataset. If the test results are consistent, proceed to step 9.6; otherwise, it indicates that the installation has failed and proceed to step 9.

7. Step 9.6: Determine if this is the first test. If it is the first test, select another set of test data from the test dataset and return to step 9.

4. If it is not the first test, it indicates that the installation was successful and proceed to step 9.

8. Step 9.7: Determine if this is the first installation. If it is, return to step 9.2; otherwise, proceed to step 9.

8. Step 9.8: Provide the user with feedback on whether the configuration and installation of the new AI computing module has failed or succeeded.

Images