Embedded AI edge computing system structure and external component configuration method

Abstract

The invention discloses an embedded AI edge computing system structure and an external component configuration method, and belongs to the field of AI edge computing equipment design. The system is based on a Jetson AGX Orin platform, comprises a plurality of standardized interfaces, and can flexibly access different types of external equipment. The system is compact in structural design, space can be effectively saved, and meanwhile stable operation of equipment in different environments is supported. Through the design of an extensible interface, the system can flexibly add external equipment, such as a sensor and a camera, according to actual demands, and supports future technology upgrading. Each interface is subjected to optimization design, high efficiency and stability of data transmission are ensured, and the problem of signal loss or interference is avoided. The external component configuration method is flexible and adjustable, application requirements of different users can be met, and wide expansion capability is provided. The system structure not only improves the computing power of the equipment, but also optimizes the aspects of power consumption management, heat dissipation, system stability and the like, and is suitable for multiple fields of automatic driving, intelligent robots, industrial Internet of Things and the like.

Description

Technical Field

[0001] This application relates to the field of AI edge computing device design, and particularly to an embedded AI edge computing system structure and an external component configuration method. Background Art

[0002] With the rapid development of technologies such as smart devices, autonomous driving, and intelligent robots, embedded AI systems play an important role in various application scenarios. Especially for intelligent edge computing systems, they not only need to have powerful computing capabilities but also be able to support the access and flexible expansion of multiple external devices.

[0003] However, there are some problems in the layout of the hardware structure, interface configuration, and scalability of the external component configuration method in existing embedded systems, which limit the compatibility and stability of the system.

[0004] To meet diverse application requirements, a new system architecture and interface configuration scheme are urgently needed to ensure that the system can operate efficiently and support future expansion. Summary of the Invention

[0005] An embodiment of this application provides an embedded AI edge computing system structure and an external component configuration method. The technical solution is as follows:

[0006] According to one aspect of this application, an embedded AI edge computing system structure is provided, and the system structure includes:

[0007] Chassis;

[0008] Main board, including a processing platform;

[0009] USB interface, used to connect external devices;

[0010] GMSL expansion card area, used to connect camera devices;

[0011] Ethernet expansion card area, used to connect network devices;

[0012] Power module, providing the voltage required by the system.

[0013] Optionally, the main board is connected to external devices through a standardized interface.

[0014] Optionally, the USB interface is a USB3.0 interface.

[0015] Optionally, the GMSL expansion card area supports 4 native GMSL interfaces and supports expansion to 8 channels by connecting an additional expansion card.

[0016] Optionally, the power module has a 24V DC input and has overvoltage and undervoltage protection functions.

[0017] Optionally, the Ethernet expansion card area is a gigabit Ethernet interface, supporting the PTP timing function.

[0018] Optionally, the system structure further includes:

[0019] An HDMI interface for connecting to a monitor;

[0020] A 12-pin general-purpose input / output interface GPIO;

[0021] A Debug interface;

[0022] A combined inertial navigation system connection port.

[0023] Optionally, the USB interface includes a USB camera interface and a reserved interface;

[0024] The Ethernet expansion card area is provided with an ETH network port and a reserved network port, and the master station corresponding to the Ethernet expansion card is connected to one of the ETH network ports.

[0025] On the other hand, a method for configuring external components is also provided. The method is used for the above-mentioned embedded AI edge computing system structure, and the method includes:

[0026] Determine the required sensor types, including cameras and lidar;

[0027] Select a suitable expansion card according to the system interface to connect external devices;

[0028] After completing the connection of external devices, perform system initialization and configuration.

[0029] Optionally, the selection of a suitable expansion card includes selecting a GMSL expansion card or a USB expansion card.

[0030] The system is based on the Jetson AGX Orin platform, includes multiple standardized interfaces, and can flexibly access different types of external devices. The system structure is designed compactly, can effectively save space, and at the same time supports the stable operation of the device in different environments. Through the scalable interface design, the system can flexibly add external devices such as sensors and cameras according to actual needs, supporting future technology upgrades. Each interface is optimized to ensure the efficiency and stability of data transmission, avoiding signal loss or interference problems. The configuration method of the external component configuration method is flexible and adjustable, can meet the application needs of different users, and provides extensive expansion capabilities. The system structure of the invention not only improves the computing power of the device, but also optimizes aspects such as power consumption management, heat dissipation, and system stability, and is applicable to multiple fields such as autonomous driving, intelligent robots, and industrial Internet of Things. Description of the Drawings

[0031] Figure 1 Shows the system architecture and hardware module diagram provided by an exemplary embodiment of the present application;

[0032] Figure 2 Shows the physical diagram of the camera expansion card provided by an exemplary embodiment of the present application;

[0033] Figure 3 Shows the physical diagram of the Ethernet expansion card provided by an exemplary embodiment of the present application;

[0034] Figure 4 Shows the physical diagram of the USB expansion card provided by another exemplary embodiment of the present application. Detailed implementation manners

[0035] To make the objectives, technical solutions and advantages of the present application clearer, the following will further describe the embodiments of the present application in detail with reference to the accompanying drawings.

[0036] As used herein, "a plurality of" means two or more. "And / or" describes the association relationship of associated objects and indicates that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects before and after.

[0037] First, the respective pictures will be introduced and explained.

[0038] Figure 1 It is the system architecture and hardware module diagram, which shows the overall architecture of the system. The system is equipped with the Jetson AGX Orin platform. At the overall structure, in order to achieve the connection of multiple external devices, there are USB interfaces, GMSL camera interfaces, Ethernet interfaces, power modules, etc.

[0039] Figure 2 It is the physical diagram of the camera expansion card, which shows the actual appearance of the camera expansion card. It is suitable for the expansion part of the GMSL camera interface mentioned in the present invention and shows how to connect an external camera device through the expansion card.

[0040] Figure 3 It is the physical diagram of the Ethernet expansion card, which shows the physical diagram of the Ethernet expansion card. It is suitable for the Ethernet interface part described in the present invention and shows how to increase Ethernet ports through the expansion card to support the connection of more network devices.

[0041] Figure 4 It is the physical diagram of the USB expansion card, which shows the physical diagram of the USB expansion card. It is suitable for showing how to expand the number of USB interfaces of the system to support the connection of more external devices.

[0042] It should be noted that for the general type of expansion cards used for practical purposes, of course, those skilled in the art can also develop self-developed expansion card connections according to their own expansion card requirements, and this is not limited herein.

[0043] Embodiment 1

[0044] An embedded AI edge computing system structure, the system structure includes:

[0045] A chassis; a main board, including a processing platform; a USB interface for connecting external devices; a GMSL expansion card area for connecting camera devices; an Ethernet expansion card area for connecting network devices; and a power module for providing the voltage required by the system.

[0046] Optionally, the main board is connected to external devices through a standardized interface.

[0047] The system structure of the present invention includes a main board (based on the Jetson AGX Orin platform), multiple interfaces (USB interface, GMSL camera interface, Ethernet interface), and a power module. When designing the system, the selection of the main board and the power module must meet the requirements of efficient data processing and stable power supply. The interface part ensures compatibility with external devices (such as cameras, sensors, computing modules, etc.).

[0048] In actual implementation, the Jetson AGX Orin platform is installed on the main board, and the corresponding interfaces (such as USB interface, GMSL interface, and Ethernet interface) are connected. The power module provides 24V DC voltage required by the system and is connected to other modules through a standardized interface.

[0049] According to application requirements, external devices such as cameras and sensors can be connected, and data transmission is carried out through a standardized interface.

[0050] The Jetson AGX Orin platform, as the core computing module, has high-performance computing capabilities and supports tasks such as AI inference and image processing. The USB interface, GMSL camera interface, and Ethernet interface ensure the flexibility and efficiency of data input and output.

[0051] The power module provides a stable 24V DC input for the system, ensuring the stability of power supply and avoiding damage caused by system overload or voltage fluctuations.

[0052] Through the high-performance computing of the Jetson AGX Orin platform, the system can perform real-time AI inference and computing tasks, meeting the high-load application requirements of industries, robotics, autonomous driving, etc.

[0053] The standardized interface design enables the system to flexibly connect different external devices and meet the requirements of different application scenarios.

[0054] The power supply module ensures the stability of the system during long-term operation, avoiding hardware failures caused by voltage fluctuations or insufficient power.

[0055] The interface part of the main board in the system (including USB interfaces, GMSL camera interfaces, and Ethernet interfaces) adopts a standardized interface design, enabling all external devices to be easily connected to the main board.

[0056] When implementing the system, users can select appropriate external devices according to their needs and connect them through standardized interfaces. For example, select a USB interface to connect sensors, a GMSL interface to connect cameras, or use an Ethernet interface for data transmission.

[0057] The standardized interface design makes the connection between devices no longer dependent on specific hardware modules, providing stronger compatibility and flexibility, and facilitating later device replacement and upgrade.

[0058] Using standardized interfaces, users do not need to worry about compatibility issues when connecting external devices, simplifying the hardware installation and configuration process.

[0059] The system can quickly access various external devices according to actual needs, enhancing the scalability and adaptability of the system.

[0060] The system design includes a main board, multiple interfaces (USB, GMSL, Ethernet), and a power supply module, ensuring that the system has efficient data processing capabilities and stable power supply. The Jetson AGX Orin platform, as the core computing module, can meet the needs of high-load applications such as industry, robotics, and autonomous driving. The standardized interface design improves the compatibility and flexibility of the system, facilitates the upgrade and replacement of later devices, and can quickly adapt to different external devices, enhancing the scalability and adaptability of the system.

[0061] Embodiment 2

[0062] Optionally, the USB interface is a USB3.0 interface.

[0063] The system uses the USB3.0 interface to provide a data transmission channel, which is suitable for high-bandwidth application scenarios. The USB3.0 interface is used in conjunction with sensors, external storage devices, etc., supporting high-speed data transmission.

[0064] In the system, the USB3.0 interface connects to external sensors, storage devices, etc., ensuring the stable transmission of high-speed data streams, especially suitable for applications that require high-frequency data transmission.

[0065] USB3.0 provides a higher transmission speed than USB2.0 (up to 5Gbps), having significant advantages in data-intensive applications such as image processing and video stream transmission.

[0066] The high transmission speed of the USB3.0 interface ensures the efficient operation of the system in applications with large amounts of data, reducing data transmission latency.

[0067] The USB3.0 interface has wide compatibility and can connect a variety of external devices, enhancing the versatility of the system.

[0068] By adopting the USB3.0 interface to provide a higher data transmission rate (up to 5Gbps), it is especially suitable for high-bandwidth application scenarios. The USB3.0 interface shows significant advantages in data-intensive applications, supporting high-speed data transmission between sensors and storage devices, ensuring the efficient operation of the system when processing large amounts of data. Its wide compatibility enables the system to connect a variety of external devices, improving the versatility of the system.

[0069] Embodiment 3

[0070] Optionally, the GMSL camera interface supports 4 native GMSL interfaces and can be extended to 8 interfaces through an expansion card.

[0071] In the system design, the GMSL camera interface supports connecting up to 4 native GMSL cameras. According to application requirements, users can choose an expansion card to expand the system to 8 camera inputs to adapt to different scales of monitoring needs.

[0072] During implementation, the GMSL interface is connected to the camera through an expansion card to transmit high-quality image data for computing and processing.

[0073] The GMSL (Gigabit Multimedia Serial Link) interface can achieve high-speed image data transmission and is suitable for scenarios with high bandwidth requirements such as high-definition video monitoring and autonomous driving systems.

[0074] Supporting multiple GMSL camera inputs, it can process high-resolution video streams and meet the high-bandwidth requirements of intelligent monitoring, autonomous driving, etc.

[0075] Supporting an expansion card that can expand the system to 8 cameras, it enhances the application scope and flexibility of the system.

[0076] The design of the GMSL camera interface supports a maximum of 4 native camera inputs and can be expanded to 8 through an expansion card. The GMSL interface provides high-speed image data transmission and is suitable for applications with high bandwidth requirements such as high-definition video monitoring and autonomous driving. The design supporting the expansion card enhances the application flexibility of the system and can meet different scales of monitoring needs.

[0077] Embodiment 4

[0078] Optionally, the power module has a 24V DC input and has overvoltage and undervoltage protection functions.

[0079] The system uses a 24V DC power supply module to provide a stable voltage. This module includes overvoltage and undervoltage protection functions to ensure a stable power supply.

[0080] The system monitors the voltage of the power supply module in real time to avoid equipment damage caused by voltage fluctuations.

[0081] The overvoltage and undervoltage protection functions of the power supply module can automatically adjust the voltage, avoid damage to the system caused by abnormal voltage, and ensure the long-term stable operation of the equipment.

[0082] The overvoltage and undervoltage protection functions ensure the stable operation of the system under various voltage fluctuations, reducing faults caused by power problems.

[0083] Through voltage protection, hardware damage caused by abnormal voltage is avoided, and the service life of the equipment is extended.

[0084] By configuring the power supply module (24V DC input with overvoltage and undervoltage protection functions), the stability of the system's power supply is ensured. Overvoltage and undervoltage protection ensure that the system is not damaged when the power supply voltage fluctuates, thereby extending the service life of the equipment and reducing faults caused by power problems.

[0085] Example 5

[0086] Optionally, the Ethernet interface is a Gigabit Ethernet interface and supports the PTP time synchronization function.

[0087] The system is configured with a Gigabit Ethernet interface to connect to other devices via Ethernet and transfer data. This interface supports the PTP (Precision Time Protocol) time synchronization function to ensure time synchronization between multiple devices.

[0088] In scenarios that require high-precision time synchronization (such as industrial automation systems), the PTP function is used to ensure efficient collaboration and synchronization between devices.

[0089] The Gigabit Ethernet interface provides a high-speed network connection, supports large-bandwidth applications, and is particularly suitable for data-intensive tasks.

[0090] The PTP function ensures that multiple devices can work together under strict time requirements through precise time synchronization, avoiding data inconsistencies caused by time deviations.

[0091] The Gigabit Ethernet interface ensures an efficient data transfer rate and can meet the demand for large amounts of data.

[0092] The PTP time synchronization function ensures the precise synchronization of each device in the system, improving the coordination and reliability of the overall system.

[0093] In addition, the system structure further includes an HDMI interface for connecting to a display, a 12-pin General-Purpose Input / Output interface GPIO, a Debug interface, a combined inertial navigation system connection port. The USB interface includes a USB camera interface and a reserved interface. In the Ethernet expansion card area, there are an ETH network port and a reserved network port, and the master station corresponding to the Ethernet expansion card is connected to one of the ETH network ports.

[0094] It adopts a Gigabit Ethernet interface and supports the PTP (Precision Time Protocol) time synchronization function, which is suitable for applications that require high-precision time synchronization (such as industrial automation systems). This interface ensures that multiple devices can work efficiently together under strict time requirements, and provides high-speed data transmission capabilities, supports large-bandwidth applications, and improves the overall coordination and reliability of the system.

[0095] Embodiment 6

[0096] On the other hand, a method for configuring external components is provided. The method is used for the embedded AI edge computing system structure of the above embodiment. The method includes:

[0097] Determine the required sensor types, including cameras and lidars;

[0098] Select a suitable expansion card according to the system interface to connect external devices;

[0099] After completing the connection of external devices, perform system initialization and configuration.

[0100] In a possible implementation, according to the system application requirements, external sensors such as cameras and lidars are selected. A suitable expansion card is selected for device connection according to the interface type (such as GMSL, USB, etc.). After completing the device connection, system initialization is performed to ensure that the devices can work properly, and various parameters are adjusted through software configuration. A suitable expansion card is selected according to the device interface type to ensure smooth communication between the device and the host. When the system is initialized, the control module checks the device status and performs necessary configurations.

[0101] The selection of the expansion card simplifies the device access process and reduces the configuration difficulty. System initialization and configuration ensure the normal operation of the devices and improve the overall efficiency of the system.

[0102] Optionally, the expansion card includes a GMSL camera expansion card or a USB expansion card.

[0103] Select a suitable expansion card type according to the required devices (such as cameras or sensors).

[0104] After selecting the expansion card, install it on the system motherboard and configure it as needed to ensure that the system can correctly identify and use the new device.

[0105] The expansion card provides more interfaces, enabling the system to connect to different external devices and expanding the application scenarios of the system. The expansion card provides more device connection options for the system, increasing the compatibility and flexibility of the system.

[0106] By selecting a suitable expansion card, the system can add new functions according to different application requirements.

[0107] An external component configuration method according to an embodiment of the present application simplifies the access process of external devices (such as cameras, lidar, etc.). By selecting a suitable expansion card and performing system initialization and configuration, the normal operation of external devices is ensured, improving the overall efficiency and flexibility of the system. This method can quickly add new devices according to different application requirements, enhancing the compatibility and adaptability of the system.

[0108] The following is an example of the application scenario of the present application.

[0109] Application Scenario 1, External Component Configuration in an Autonomous Driving System

[0110] In an autonomous driving system, the required external sensors include lidar (LiDAR), cameras, GPS modules, etc. The system needs to select these device types and perform precise configuration according to the requirements of autonomous driving.

[0111] According to the interface requirements (for example, lidar may require a GMSL interface, and cameras and GPS devices may require a USB interface), select a suitable expansion card to connect these external devices. GMSL expansion cards and USB expansion cards can ensure the compatibility of sensors with the edge computing system.

[0112] After the devices are connected, the system will automatically initialize. The control module will check the status of the hardware module and verify whether the connection is successful to ensure that all sensors are working properly.

[0113] GMSL expansion cards and USB expansion cards ensure that different types of sensors can be smoothly connected to the system, especially for high-speed image data transmission and real-time computing.

[0114] The system ensures that each external device can be efficiently and stably connected to the embedded AI edge computing system through a standardized interface.

[0115] Through the selection of standardized interfaces and expansion cards, the system can flexibly connect different sensors and support complex autonomous driving environments.

[0116] High-speed GMSL and USB interfaces ensure the real-time transmission of a large amount of data, improving the response speed of autonomous driving decisions.

[0117] Application Scenario 2, External Component Configuration in an Intelligent Manufacturing System

[0118] In an intelligent manufacturing system, the system may require devices such as temperature sensors, humidity sensors, pressure sensors, cameras, etc. The system will select appropriate sensors according to the specific requirements of the production line.

[0119] According to the type of sensor interface (such as GMSL for cameras and USB for sensors), select the appropriate expansion card. Through these expansion cards, the system can expand the device interface to ensure the efficient access and data transmission of sensors.

[0120] After all devices are connected, the system will be initialized through the control module to check the status of all sensors and external devices to ensure that the devices can operate correctly.

[0121] According to the requirements of intelligent manufacturing, the system needs to flexibly select and configure different external sensor interfaces. The selection of GMSL and USB interfaces ensures the speed and reliability of data transmission.

[0122] Through the initialization of the system and the verification of the hardware status, ensure that the sensors can work as expected and be synchronized with the main control system.

[0123] Through the precise configuration of sensors and the support of expansion cards, the system can efficiently collect production line data for real-time monitoring and adjustment.

[0124] Efficient data transmission and real-time calculation enable the intelligent manufacturing system to quickly respond to changes and improve the efficiency and reliability of the production line.

[0125] Application Scenario 3, External Component Configuration in a Smart City Monitoring System

[0126] In a smart city monitoring system, the required external devices include high-definition cameras, environmental monitoring sensors, traffic flow monitoring devices, etc. The system will determine the type of sensors according to the specific monitoring requirements.

[0127] The system selects suitable expansion cards according to different sensor interface requirements. For example, GMSL expansion cards are used to connect high-resolution cameras, and USB expansion cards are used to connect environmental monitoring sensors.

[0128] After the device connection is completed, the system performs initialization configuration to ensure that all devices work properly and can transmit data in real time.

[0129] The system ensures that different types of external devices can be successfully connected through standardized interfaces such as GMSL and USB to achieve efficient data collection and transmission.

[0130] The system verifies whether the external devices are successfully connected to the host through the initialization process and ensures that the functions of each device can operate normally.

[0131] Flexibly select and configure external sensors to ensure the efficient execution of various monitoring tasks in the smart city.

[0132] The use of high-bandwidth interfaces and high-speed expansion cards ensures the real-time transmission of a large amount of data, enhancing the system's response speed.

[0133] Application Scenario 4, Robot Vision and Perception System

[0134] In the robot vision and perception system, the required sensors include RGB cameras, depth cameras, lidars, etc. Select appropriate sensor devices according to the complexity of the tasks performed by the robot.

[0135] According to the device interface requirements, select a GMSL expansion card to connect the camera, or select a USB expansion card to connect other sensors to ensure that the system can quickly access and process data from external devices.

[0136] After the device connection is completed, the system will automatically complete the initialization configuration to ensure that all external devices can work smoothly in the robot system.

[0137] The system connects different sensors through standardized interfaces to ensure the efficient integration of data from multiple sensors and improve the robot's perception ability of the environment.

[0138] During the initialization phase, check the hardware status to ensure the smooth data transmission of each external device and avoid data loss or transmission delay caused by connection problems.

[0139] Through the configuration of multiple sensors and expansion cards, the robot can more accurately perceive and understand the surrounding environment. The real-time data processing ability ensures that the robot can quickly respond to environmental changes and make decisions.

[0140] The embodiment of the present application also provides a computer-readable medium, which stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the embedded AI edge computing method described in each of the above embodiments.

[0141] The above are only optional embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. An embedded AI edge computing system structure, characterized in that: The system structure includes: Chassis; The motherboard, which contains the processing platform; USB interface, used to connect external devices; GMSL expansion card area, used to connect camera equipment; Ethernet expansion card area, used to connect network devices; The power module provides the voltage required by the system.

2. The system structure according to claim 1, characterized in that: The mainboard is connected to external devices via a standardized interface.

3. The system structure according to claim 1, characterized in that: The USB interface is a USB 3.0 interface.

4. The system structure according to claim 1, characterized in that: The GMSL expansion card area supports 4 native GMSL interfaces and supports expansion to 8 interfaces by connecting an expansion card.

5. The system structure according to claim 1, characterized in that: The power module has a 24V DC input and has overvoltage and undervoltage protection functions.

6. The system structure according to claim 1, characterized in that: The Ethernet expansion card area is a Gigabit Ethernet interface and supports the PTP timing function.

7. The system structure according to claim 1, characterized in that: The system structure also includes: HDMI interface, used to connect to a monitor; 12-pin general purpose input and output interface GPIO; Debug interface; Combined inertial navigation system connection port.

8. The system structure according to claim 1, characterized in that: The USB interface includes a USB camera interface and a reserved interface; The Ethernet expansion card area is provided with an ETH network port and a reserved network port, and the master station corresponding to the Ethernet expansion card is connected to one of the ETH network ports.

9. A method for configuring an external component, characterized in that: The method is used for the embedded AI edge computing system structure according to any one of claims 1 to 8, and the method comprises: Determine the types of sensors needed, including cameras and lidar; Select appropriate expansion cards to connect external devices according to the system interface; After completing the external device connection, initialize and configure the system.

10. The method according to claim 9, characterized in that The selecting a suitable expansion card includes selecting a GMSL expansion card or a USB expansion card.

Images