The Sky Star · XiaoZhi AI provides a simple example of using the Ai-WV01-32S XiaoZhi AI MCP to control a microcontroller. Through MCP programming, the AI can control peripherals such as:
- AI-controlled LED
- AI-controlled motor
- AI-controlled volume settings
Key Features
- Full voice interaction: offline voice control + AI conversation
- Quick network provisioning via mini-program
- Emoji display
- Subtitle display
- AI control of WS2812 color adjustments
Feature Demonstration
Color Control
Play Music
System Block Diagram
Hardware Design
The most complex circuit in this project is the OLED driver circuit, which only requires following the specifications provided.
Speaker & Accessories
The speaker used is a 2W 8Ω mobile phone speaker. We created and verified the footprint for it.
Purchase link (Taobao):
Touch spring contacts: Recommended height is 2.5–3 mm to ensure proper contact.
Mounting posts: Use M1 PCB nuts. These require heating with a hot air gun for proper soldering.
Note: Do not apply too much solder, or the hole may become blocked.
Software Information
This project uses the Ai-WV01-32S, but not the default firmware. Instead, it uses the UART-MCP firmware developed by Ai-Thinker’s open-source team (firmware link provided below).
The first step after soldering is flashing the UART-MCP firmware; otherwise, control functions will not work.
Ai-Thinker provides a dedicated adaptation library for UART-MCP: emMCP.
It requires only a simple porting process, allowing Sky Star to focus mainly on OLED display and peripheral control logic.
Additional Resources
· Source code:https://gitee.com/Ai-Thinker-Open/emMCP
· UART-MCP firmware download: download
· MCP reference documentation:
· https://docs.ai-thinker.com/uart_mcp/emMCP.html
· emMCP programming guide:
https://docs.ai-thinker.com/uart_mcp/emMCP_getStart.html
· Sky Star firmware: see attachments
· Community, replication & technical discussion:
Ai-Thinker
djsb
Pablo Garrido
sfowen