CYCLIOT is a smart IoT-based system designed to improve cyclist safety and riding experience using real-time sensor data, mobile applications, and AI-powered recommendations. Developed by an interdisciplinary team as part of the CME 4436 course at Dokuz Eylül University, the system integrates proximity alerts, health tracking, crash detection, and contextual suggestions through a Bluetooth-connected Android app.
- Project Title: CYCLIOT
- Institution: Dokuz Eylul University, Faculty of Engineering, Computer Engineering
- Course: CME 4436 – Basics of Internet of Things
- Date: June 2025
| Name | Role |
|---|---|
| Ahmed Cengiz Yavuz | Software Developer |
| H*****n A****n | Software Developer |
| G*****z H****u | Software Developer |
- Sensor Integration: Heart rate (MAX30100), body temperature (DHT11), proximity (HC-SR04), angular movement (MPU6050), GPS.
- Real-time Bluetooth Communication: ESP32 microcontroller to Android app.
- Mobile App:
- Live data visualization
- Ride summaries (speed, distance, heart rate)
- Crash alerts and emergency contact notification
- Firebase login and data sync
- Cloud Processing: Data is stored and analyzed via Google Firebase and Gemini LLM.
- Smart Suggestions:
- Weather-aware route planning
- Hydration reminders
- Fatigue detection and health alerts
- Crash Detection: Based on accelerometer threshold + notification system.
- AI Feedback: Performance analytics and real-time riding recommendations.
- ESP32 handles sensor reading, Bluetooth transmission, and low-power management.
- Android App (Java, Android Studio) manages UI/UX and visualizations.
- Cloud Backend using Google Firebase for real-time database and messaging.
- AI Layer using Google Gemini LLM for intelligent recommendations.
| Week | Milestone |
|---|---|
| 1 | Hardware and sensor selection |
| 2 | Firmware for ESP32 and Bluetooth setup |
| 3 | Android app prototype |
| 4 | Cloud setup and LLM integration |
| 5 | Field testing and data logging |
| 6 | System refinements (enclosures, smoothing) |
| 7 | App UI/UX improvements |
| 8 | Real-time alerts and crash notification system |
| 9 | Stress testing and optimizations |
| 10 | Final documentation and demo kit |
- Microcontroller: ESP32 DevKitC
- Sensors: MAX30100, DHT11, HC-SR04, MPU6050, GPS
- Firmware: Arduino IDE
- Mobile: Java, Android Studio, MPAndroidChart
- Backend: Firebase (Authentication, Firestore, Cloud Messaging)
- AI: Google Gemini LLM via REST API
- Design: 3D Printed Sensor Enclosures
- Sensor Noise: Implemented moving average filters and enclosures.
- Power Constraints: Optimized firmware with deep sleep modes.
- Connectivity: Reconnection logic and fallback buffer for Bluetooth.
- Crash Detection: Calibrated acceleration threshold to reduce false positives.
- Privacy Concerns: Planned support for encryption and consent-based data sharing.
- Voice control and AI-powered route prediction.
- Integration with wearable health trackers.
- City infrastructure partnerships and bike fleet compatibility.
- End-to-end data encryption for commercial deployment.
Refer to to the Final Raporu.pdf for academic references and technical documentation.
For academic or entrepreneurial inquiries, please contact:
🚴 CYCLIOT – Smart cycling, safer tomorrow.