Enhanced Driver Assist (EDA)

FYP-063-D-EDA | National University of Computer and Emerging Sciences

Making Roads Safer with AI-Powered Driver Assistance

About the Project

The Enhanced Driver Assist (EDA) system is an AI device that improves road safety by combining various advanced driver assistance features. Our system addresses key challenges such as driver fatigue, distraction, lane discipline, and risky driving behaviors by integrating AI, IoT, and edge-computing technologies.

Fatigue and distraction-related road accidents account for a significant percentage of road traffic collisions worldwide. On motorways, fatigue-related RTCs are responsible for 54% of fatal accidents, while on national highways, they cause 41% of serious injuries. In Pakistan, drowsy driving was a contributing factor in 33% of accidents on the M2 motorway.

EDA monitors driver attention, detects lane departures, and analyzes driving patterns to provide timely alerts and interventions, thereby reducing the risk of accidents. It incorporates hardware like Raspberry Pi and Coral TPU for processing, and OBD-II sensors for vehicle data collection.

Raspberry Pi
Coral TPU
OBD-II
Dashcams
Python
OpenCV
TensorFlow Lite
MediaPipe
Edge Computing

Our Team

Manahil Kamran

Manahil Kamran

21I-2668

Responsibilities: Driver Attention Assist, Lane Departure Warning, Driving Pattern Monitoring

Ali Arfa

Ali Arfa

21I-2669

Responsibilities: Hardware Implementation, Voice Assistant, OBD-II Integration

Dr. Faisal Cheema

Project Supervisor

Ms. Hira Mastoor

Project Co-Supervisor

Knowledge Discovery and Data Science (KDD) Lab

This project was developed at the KDD Lab, National University of Computer and Emerging Sciences (FAST-NUCES), Islamabad. The KDD Lab is a research facility focused on data science, machine learning, and artificial intelligence applications.

The lab provides resources and mentorship for innovative projects like EDA, supporting students in developing cutting-edge solutions to real-world problems.

System Modules

Driver Attention Monitoring

Monitors the driver's facial expressions and movements to detect signs of fatigue, drowsiness, or distraction. The system provides alerts when inattention is detected, helping to prevent accidents caused by driver fatigue.

  • Facial feature tracking using computer vision
  • Drowsiness detection through eye movements
  • Head movement tracking to monitor driver attention

Lane Detection and Departure Warning

Ensures that the vehicle remains within lane boundaries. It uses road-facing cameras to track the vehicle's position in the lane and warns the driver if the vehicle deviates from its path.

  • Lane detection using camera input
  • Vehicle position tracking within lane boundaries
  • Alert system for lane departure warnings

Driving Pattern Monitoring

Tracks driver behavior using OBD-II sensor data to monitor risky driving practices like harsh braking, speeding, and rapid acceleration. The system provides feedback to promote safer driving habits.

  • Collection of driving data from OBD-II sensors
  • Analysis of driving patterns (e.g., braking, speeding)
  • Alerts for unsafe driving practices

Voice Assistant

Enables hands-free interaction with the system. The voice assistant can control various system functions and respond to queries, helping the driver stay focused on the road.

  • Voice-controlled system interface
  • Conversational AI for normal interactions
  • Real-time processing on edge devices for quick response

Hardware Implementation and Integration

Focuses on the integration of hardware components such as dashcams, Raspberry Pi, Coral TPU, and OBD-II sensors. It ensures the synchronization of hardware and software for optimal performance.

  • Setup of road-facing and driver-facing dashcams
  • Integration of OBD-II sensor for vehicle data collection
  • Synchronization between hardware and software components

Showcasing Events

Over the past two months, we've had the opportunity to represent our Final Year Project — EDA (Enhanced Driver Assist) — at some of Pakistan's leading tech competitions. Our AI-powered driver assistance system has been recognized for its real-world applicability and innovation in improving highway safety through advanced driver assistance features.

These competitions have provided valuable validation and feedback, helping us refine our solution and demonstrate its potential impact on road safety.

COMPEC 2025 Trophy - 1st Position

1st Position – COMPEC 2025

📍 NUST CEME, Islamabad | IoT & Digital Technologies Category

EDA stood out for its real-world applicability and innovation at this national-level project competition. Competing against top-tier student projects, we were honored to secure 1st Place in the IoT & Digital Tech category.

SOFTEC 2025 Trophy - 2nd Position

2nd Position – SOFTEC AI Product Innovation Competition

📍 FAST NUCES, Lahore

This was an intense 7-day sprint, building a functional AI-based product end-to-end. EDA was recognized as one of the top AI innovations, landing us the Runner-up position in a field focused on creativity, impact, and product viability.

NumAIsh (NASCON 25) Trophy - 2nd Position

2nd Position – NumAIsh (NASCON 25)

📍 FAST NUCES, Islamabad

At NASCON's premier project exhibition judged by experts from VisionRD, Devsinc, and others, EDA earned the Runner-up position — a testament to our technical execution and pitch on highway safety innovation.

ExcITe Cup 2025 Trophy - 1st Position

1st Position – ExcITe Cup 2025 Project Exhibition

📍 CUST Islamabad

Among innovative tech solutions from across the country, EDA captured attention for its relevance to real-world road safety challenges — earning us 1st Place at CUST's national exhibition.

All EDA Competition Trophies Together

EDA Recognized Across Pakistan's Top Tech Competitions

We're grateful to each organizing body, panel of judges, and fellow innovators for the valuable feedback and support. These competitions helped us validate our idea and pushed us to deliver better, smarter, and safer solutions for real-world mobility challenges.

EDA isn't just a project. It's our vision for safer roads.

Module Demonstrations

Watch our system in action with these demonstration videos of each module:

Driver Attention Monitoring Demo

Real-time detection of driver drowsiness and distraction

Lane Detection and Departure Warning Demo

Real-time lane tracking and alert system

Driving Pattern Monitoring Demo

Analysis of driving behavior using OBD-II data

Voice Assistant Demo

Hands-free interaction with the EDA system

Complete System Demo

The entire EDA system working together in a real driving scenario

Project Timeline

September - October 2024

Iteration 1

Data Collection and Pre-processing, Hardware setup

November - December 2024

Iteration 2

Driver Attention Monitoring and Driving Pattern Analysis, Integration, and Testing

January - February 2025

Iteration 3

Lane Detection and Departure Warning, Voice Assistant, Integration, and Testing

March - April 2025

Iteration 4

Optimization, Deployment, Combined Testing

Project Impact

The Enhanced Driver Assist (EDA) system addresses critical road safety challenges with an integrated approach to driver assistance. Our project aims to make a significant impact in several areas:

Our project demonstrates how combining AI, edge computing, and IoT technologies can create practical solutions for real-world safety challenges, particularly in developing countries like Pakistan where such systems are urgently needed.