Screens can show information. But when a device starts telling a story, reacting to touch, speaking with a voice, and animating like a living object, the experience changes completely.
This project turns the ELECROW CrowPanel Advance ESP32-P4 into an interactive AI-powered storybook with animated scenes, local voice playback, and touch interaction.
Instead of static UI elements, the device itself becomes the experience:It plays animations, speaks, reacts, and guides the user through a story directly on the panel.
The result feels somewhere between:
- a smart toy
- an interactive exhibit
- a talking photo frame
- and a next-generation embedded HMI demo
Because everything runs locally on the device, the experience feels immediate and self-contained — no cloud delays, no external rendering pipeline, no remote processing.
The project is useful not only as a demo, but also as a starting point for real products and installations:
- interactive museum exhibits
- educational devices
- smart displays
- storytelling installations
- children’s products
- branded experiences
- custom embedded HMI systems
The project is fully open and designed to be modified. You can replace the assets, change the stories, redesign the UI, and build your own experience on top of the framework.
Repository: https://github.com/Grovety/Crowpanel_P4_Storybook
HardwareThis project uses:
- ELECROW CrowPanel Advance ESP32-P4 (7", 9", or 10.1")
- GRC tinyTTS
- microSD card
The ESP32-P4 platform is especially interesting for rich embedded UI projects because it combines:
- high-resolution graphics support
- smooth LVGL-based interfaces
- Wi-Fi connectivity
- local media handling
- enough performance for large animated UI scenes
For this project, we also experimented with different animation pipelines for large alpha-based transitions.Instead of relying only on traditional GIF playback, we used a compiled animation approach for heavier scenes, reducing runtime overhead and improving playback smoothness on complex transitions.
GRC AI Add-onAt the core of the project is the GRC AI Add-on — a tiny edge-AI module that brings local neural-network inference to the CrowPanel, enabling interactive storytelling and intelligent behavior without cloud processing.
The GRC AI Add-on plugs directly into the dedicated expansion slot on the back of the CrowPanel, making the setup clean, compact, and easy to assemble.
The easiest way to test the project is to use the prebuilt firmware images from GitHub.
The CrowPanel ESP32-P4 contains two controllers:
- ESP32-C6 — used for Wi-Fi communication - flash it first
- ESP32-P4 — runs the main application - flash after C6
The ESP32-C6 firmware must be updated first because some panels may come from the factory with an older Wi-Fi firmware version.
Step0. Prepare the microSD card
- Insert a microSD card into your computer.
- Format the microSD card as **FAT32**.
- Create an `assets` folder in the root of the microSD card.
- Copy the runtime asset files from the repository into this folder.
- The resulting microSD card layout should look like this:
<SD root>/
└── assets/
├── book_open.cfa
├── book_swipe.cfa
├── book_close.cfa
├── bird_idle.gif
├── bird_start_talk.gif
└── bird_talk.gifAfter copying the files, safely eject the microSD card and insert it into the CrowPanel TF card slot.
Note: The application requires both the GRC AI Add-on and the microSD card to run. If either of them is missing, the firmware can still be flashed, but the application may not start correctly after boot.
Step1. Flash ESP32-C6 Firmware
Steps:
- Download and extract the archive with ESP32-C6 firmware
- Connect the panel to your PC via the UART USB Type-C port
- Open the folder with the extracted firmware files.
- Run the Flash_tool.exe for the ESP32-C6 firmware.
- Wait until the flashing process is complete.
Step2. Flash ESP32-P4 Firmware
- Download and extract the archive for ESP32-P4
- Run the flash_from_binaries_p4.exe for the ESP32-P4 application firmware.
Step 3. Initial Setup
The setup process is intentionally simple.The only thing you can configure is the name the panel will use when talking to you.
When the device starts for the first time, the setup screen appears automatically. You can enter your own name or simply keep the default one.
Configuration is done through a lightweight web interface that can be opened from a smartphone or laptop.
Steps:
- The panel creates its own Wi-Fi network
- Scan the first QR code to connect to the panel Wi-Fi
- Scan the second QR code to open the setup page(or manually open
192.168.4.1in your browser) - Enter your name and save the settings
You can return to the name setup later at any time through the settings screen — just tap the gear icon in the corner of the panel UI.
That’s it — the panel is ready to use.
Everything in this project is customizable.
You can:
- add your own GIF/CFA assets
- change story text
- build your own storytelling experience
The easiest way to start is simply replacing the assets and text while keeping the existing framework.
The repository already contains the project structure, runtime pipeline, and asset-loading logic, making experimentation much easier.
You can use this project as:
- a template for interactive devices
- an HMI demo framework
- a smart exhibit prototype
- an educational storytelling platform
- a starting point for commercial embedded UI concepts
Full source code and instructions are available on GitHub:
If you are building a commercial product or need a custom version, feel free to reach out. We can help adapt this solution for commercial or enterprise use.
In addition, Grovety specializes in panel software development and is an official ELECROW partner. Whether you need to design a panel, develop a new application, enhance an existing app, or integrate with external sensors and services, we have the expertise to deliver tailored solutions.
Comments