Sentinel: Smart ADAS-Integrated Fleet Monitoring System

Sentinel is an advanced fleet monitoring solution designed to improve the safety and management of buses and trucks. It integrates cutting-edge Advanced Driver Assistance Systems (ADAS) such as collision detection, driver monitoring, and adaptive cruise control with real-time fleet management capabilities, ensuring both safety and operational efficiency.

Features

1. Driver Monitoring:

   • Uses dlib and OpenCV to monitor driver attentiveness and behavior.
   • Detects signs of drowsiness or distraction, triggering warnings.

2. Collision Detection:

   • Real-time detection of potential collisions using LIDAR sensors.
   • Provides early warnings to avoid accidents.

3. Passenger Overcrowding Detection:

   • Utilizes YOLO object detection for real-time monitoring of passenger overcrowding in vehicles.
   • Ensures adherence to safety guidelines by preventing overcrowding.

4. Fleet Tracking Simulation:

   • GPS-based vehicle tracking system implemented using a Flask backend.
   • Provides real-time location data for fleet management.

5. Adaptive Cruise Control with LIDAR:

   • Enables adaptive speed regulation based on distance from other vehicles.
   • Ensures smooth and safe driving conditions.

6. Web Dashboard for Fleet Management:

   • Provides a comprehensive overview of vehicle locations, driver behavior, and passenger conditions.
   • Enables real-time decision-making and fleet oversight.

System Architecture

Sentinel-Smart-ADAS-Integrated-Fleet-Monitoring-System
├── src
│   ├── coco.names
│   ├── crowded_bus.webm
│   ├── shape_predictor_68_face_landmarks.dat 
│   ├── yolov3.cfg
│   └── yolov3.weights                #Needs to be downloaded and placed here         
├── driver_monitoring.py          # Driver attentiveness system
├── collision_detection_lidar.py  # Collision detection using LIDAR
├── adaptive_cruise_control_lidar.py # Adaptive cruise control system
├── passenger_counting.py         # Passenger overcrowding detection
├── fleet_tracker_simulation.py   # Fleet tracking and simulation
├── app.py                        # Flask API for fleet tracking
├── index.html                    # React or plain HTML dashboard (based on use case)
├── README.md                         # Project documentation
├── main.py                           # Script to run all the Python scripts simultaneously
├── requirements.txt                  # Dependencies
└── .vscode                           # VS Code settings

System Flow

1. Driver Monitoring: Uses the driver_monitoring.py script to ensure driver attentiveness through face tracking and blink detection.
2. Collision Detection: The collision_detection.py script continuously monitors the vehicle's surroundings using LIDAR sensors to prevent collisions.
3. Adaptive Cruise Control: The adaptive_cruise_control_lidar.py script dynamically adjusts the vehicle's speed based on LIDAR input.
4. Passenger Counting: The passenger_counting.py script monitors the vehicle’s interior to track overcrowding using YOLO detection models.
5. Fleet Tracker Simulation: The fleet_tracker_simulation.py script simulates real-time vehicle location and reports it to the Flask backend.
6. Web Dashboard: Fleet managers can access real-time data from all subsystems through the web dashboard (backend hosted in app.py).

How to Run

Setup

Download YOLOv3 Weights

To use the YOLO model for passenger overcrowding detection, you need to download the pre-trained weights file yolov3.weights. This file is not included in the repository due to its size.

Follow these steps to download it:

1. Download the yolov3.weights file from the official YOLO website or using the following link:

   • Download YOLOv3 Weights

2. After downloading, place the yolov3.weights file in the /src directory

Prerequisites

• Python 3.10+
• Flask for the backend
• OpenCV, Dlib, YOLO for computer vision tasks
• LIDAR simulation libraries

Installation

1. Clone the repository:

   git clone https://github.com/icebelly29/Sentinel-Smart-ADAS-Integrated-Fleet-Monitoring-System.git
   cd Sentinel-Smart-ADAS-Integrated-Fleet-Monitoring-System

2. Install the required dependencies:

   pip install -r requirements.txt

Running the System

1. Run the Flask backend:

   python backend/app.py

2. Run the main system components simultaneously using the main script:

   python main.py

3. Access the web dashboard:
   Open a browser and navigate to http://127.0.0.1:5000 to view fleet and driver data.

Tech Stack

• Python: Core language used for ADAS, driver monitoring, and passenger counting.
• Flask: Backend framework to handle vehicle tracking and dashboard API.
• OpenCV + Dlib: For real-time computer vision in driver monitoring.
• YOLO: Used for passenger overcrowding detection.
• LIDAR: Simulated LIDAR data for collision detection and adaptive cruise control.
• HTML + JavaScript: Used for the web dashboard frontend.

License

This project is licensed under the MIT License - see the LICENSE file for details.