Zecko ES

Embedded System Built on the Unihiker K10

Zecko ES is a multifunctional embedded system built from scratch using the Unihiker K10.

The goal of this project was to maximize the capabilities of the board and create a fully interactive daily-use device combining hardware control, AI features, sensor monitoring, and multimedia functions — all written in C++.

---

🚀 Features

Zecko ES includes:

• 📷 Camera – Capture and save photos to SD card
• 🎵 Music Player – Play .wav files from SD card
• 🏋️ Workout Station – Chronometer, pedometer, sets & reps counter
• 🤖 Zecko AI – Voice-interactive chatbot
• 🌤 Weather Station – Temperature, humidity, light intensity & noise monitoring
• 👁 AI Recognition – Face detection and analysis
• 🎮 Reaction Game – Reflex-based color game
• 🎙 Journal – Voice recorder with SD card storage

---

🧠 Hardware Overview

The project is built entirely around the:

Unihiker K10

Manufactured by DFRobot, the board integrates:

• ESP32-S3 MCU (Xtensa LX7)
• 2.8” 240x320 color display
• Wi-Fi (2.4 GHz)
• Bluetooth 5.0 (BLE)
• 2MP Camera
• Microphone & Speaker
• RGB LED
• Temperature & Humidity sensor
• Ambient Light Sensor
• Accelerometer
• SD Card support

No external PCB or breadboard was required — everything is integrated on the board.

---

🏗 System Architecture

The system is built around:

• Persistent state counters
• Button-based navigation logic
• Sensor polling in loop()
• Mode-based function switching
• AI mode switching (Face / NoMode)
• SD card file interaction

The navigation system relies on internal counters (pressA, pressB) to switch between modules.

While intentionally simple in structure, the logic demonstrates:

• State management
• Event callbacks
• Hardware abstraction usage
• Sensor integration
• AI module interaction

---

🛠 Development Environment

• Language: C++ (Arduino framework)
• Platform: PlatformIO
• Board: unihiker_k10
• Framework: Arduino

platformio.ini

[env:unihiker]
platform = https://github.com/DFRobot/platform-unihiker.git
board = unihiker_k10
framework = arduino
build_flags = 
    -DARDUINO_USB_CDC_ON_BOOT=1
    -DARDUINO_USB_MODE=1
    -DModel=None

---

📦 Required Libraries

#include "unihiker_k10.h"
#include "math.h"
#include <ctime>
#include <Wire.h>
#include "AIRecognition.h"
#include "asr.h"

---

🎨 3D Enclosure

A custom enclosure was designed to:

• Keep GPIO pins accessible
• Avoid blocking Wi-Fi antenna
• Maintain sensor exposure
• Preserve speaker output
• Allow press-fit assembly

You can download the case files on Printables (link in repository description).

---

🧪 Design Philosophy

Zecko ES was built under a personal challenge:

• No heavy AI assistance for coding
• Fast development (under 6 hours)
• Focus on core C++ logic understanding
• Manual state handling
• Minimal abstraction layers

The logic is intentionally raw but functional — the objective was learning and system integration.

---

⚠ Known Limitations

• No structured state machine implementation
• No error-handling safeguards
• AI + Wi-Fi cannot run reliably simultaneously (memory constraints)
• Button-counter navigation can fail under unintended usage
• Not optimized for scalability

Future refactoring could include:

• Proper state machine
• Debouncing logic
• Modular class-based architecture
• Memory optimization
• Power management improvements

---

📈 Future Improvements

• Refactor main.cpp
• Implement structured state handling
• Add OTA updates
• Expand AI interactions
• Add file manager system
• Improve UI transitions

---

📜 License

This project is open-source.
See LICENSE file for details.

---

🙌 Acknowledgment

Special thanks to DFRobot for providing the Unihiker K10 through their community program.