EDGE INTELLIGENCE NODE DEVELOPMENT¶
ARCHITECTURE¶
+------------------------------+
| Application Layer | <-- Application Code
+------------------------------+
| System Services Layer | <-- OS + Middleware
+------------------------------+
| Platform Abstraction Layer | <-- Startup Code + HAL (register-level) + BSP (board-level)
+------------------------------+
| Hardware Layer | <-- Main Control + Peripherals
+------------------------------+
Info
Except for the application layer content, which is strongly bound to specific uses, the contents of other layers are universal, so the development of hardware, platform abstraction layer, and system services layer can be universal and can be ported to different application scenarios. They are placed in different repositories for easy management and maintenance.
NODE HARDWARE¶
Hardware Development
To achieve edge intelligence computing, we have developed two types of MCU nodes, based on STM32 and ESP32. These nodes have high-performance edge computing capabilities and can be used in IoT, smart home, smart city, and other application scenarios. The current development focus is on ESP32.
DRIVER DEVELOPMENT¶
Driver Development
Driver development is an important part of edge intelligence node development. We have developed various drivers, including sensor drivers, communication drivers, display drivers, etc. These drivers provide rich peripheral support for edge intelligence nodes, enabling nodes to better adapt to different application scenarios.
Note
The drivers included in this repository are actually included in the node repositories of ESP32 and STM32 above. Here, they are only organized and included for easy porting to other devices and platforms.
-
MCU_BSP๐ฏ๐
Currently Supported Development Boards:
-
STM32
-
ESP32
-
TinySHM DEVELOPMENT¶
TinySHM Development
TinySHM is an edge computing enablement framework developed specifically for structural health monitoring, containing various modules and services, integrated in the form of middleware in edge intelligence node software. TinySHM provides rich edge computing support for structural health monitoring, enabling nodes to better adapt to different structural health monitoring application scenarios. It has the following characteristics:
- Modular design, easy to expand
- Cross-platform support, strong portability, reserved areas for custom implementation on different platforms
- Rich middleware services, including data acquisition, data processing, data storage, data transmission, etc.
-
TinySHM๐ฏ๐
Currently Supported Platforms:
-
STM32
-
ESP32
-