🚀 Design and Implementation of Intelligent Safety Services for Personal Mobility Devices

AI-powered safety enhancement for personal mobility

---

📌 Project Overview

This project is an AI-based system designed to improve personal mobility safety through real-time hazard detection and notification, indoor parking return confirmation, and driving habit analysis. This project integrates computer vision (YOLOv5, NetVLAD), sensor-based driving habit analysis, and web/app interfaces to create a comprehensive smart mobility solution.

📖 Paper Published:
IEEE MetaCom 2023 - Design and Implementation of Intelligent Safety Services for Personal Mobility Devices

🗓 Conference Date: 26-28 June 2023

📅 Publication Date (IEEE Xplore): 06 October 2023

📍 Conference: 2023 IEEE International Conference on Metaverse Computing, Networking and Applications (MetaCom)

🔥 Features

✔ Real-time hazard detection (School zones, speed bumps, sudden stops)
✔ AI-based pothole & speed bump recognition (YOLOv5)
✔ Visual localization-based parking validation (NetVLAD)
✔ IoT integration with Raspberry Pi for GPS, gyro, and camera uploads
✔ Web-based admin dashboard & mobile app

---

🛠️ Project Structure

1️⃣ Risk Detection Modules

🚦 Hazard Detection

python School_zone_detect.py
python Speed_bump_detect.py
python Sudden_stop_detect.py
python Analysis_rating.py

# Activate environment
conda activate <your_env>

# Pothole detection
cd yolov5_pothole/get_image_Folder
python integrated.py

# Speed bump detection
cd yolov5_speedbump/get_image_Folder
python integrated.py

---

2️⃣ Web Interface 🌐

execute web/src/admin.html

---

3️⃣ IoT Device Integration (Raspberry Pi) 🖥️

Handles real-time data uploads from sensors and cameras.

# Activate Conda environment
conda activate <your_env>

# Upload camera and sensor data
python Upload_cam.py
python upload_gps.py
python upload_gyro.py

---

4️⃣ Mobile App 📱

# Download all files in the 'app' folder

---

5️⃣ Parking Validation via Visual Localization 📍

Ensures proper kickboard parking using AI-based visual localization (NetVLAD).

# Activate environment
conda activate <your_env>

# Run visual localization
python Visual_localization.py

---

🎥 Experiment Results & Hardware Specifications

For detailed experiment results, hardware specifications, and implementation details, visit the project page below:

🔗 Hackster.io Project Page

This page includes:
✅ Real-world test videos showcasing system performance
✅ Hardware specifications and sensor details
✅ Implementation insights on AI-based safety services

Check it out to explore how our intelligent safety system enhances personal mobility! 🚀