SMART AUTOMOBILE

While driving or traveling, safety is our first priority. Even a single mistake by the driver may result in severe physical injuries, deaths, and significant economic losses. Many systems are available in the market today, such as navigation systems, different sensors, etc., to make driver?s work ea

Project Title

SMART AUTOMOBILE

Project Area of Specialization

Artificial Intelligence

Project Summary

While driving or traveling, safety is our first priority. Even a single mistake by the driver may result in severe physical injuries, deaths, and significant economic losses. Many systems are available in the market today, such as navigation systems, different sensors, etc., to make driver’s work easy. Our project “THE SMART AUTOMOBILE (SAM)” also serves the same purpose. “THE SMART AUTOMOBILE" is predicated on the possibility of a pre-crash framework." The aim of this project is to introduce a "SMART AUTOMOBILE" that constantly monitors the driver and the environment for real-time hypo-vigilance detection, intrusion of obstacles, and multi-parameter-based vital signs. Continuous event-related driver monitoring, effective driver-based system customization, and actual traffic situation consideration will enhance system reliability and minimize false alarm rate. Recent research suggests that driver hypo-vigilance (or under-awakening), drunk driving, sudden health issues and over-speed causing collisions with other vehicles or obstacles are among the major causes of road accidents. The proposed system detects the driver's drowsy state by using the image processing and feature extraction techniques while the driver's drunk state is detected by an alcohol sensor interfaced with the embedded board. The system will issue several warning signals upon detection of any of the above-mentioned state and will also stop the car in order to avoid any accident. In addition to this feature, the system will also monitor the driver's health parameters (by analyzing the ECG, EEG signals, blood pressure) and would inform the nearest hospital / rescue system via a message when detecting any serious abnormalities. The system also reduces the risk of accidents by using LIDAR and camera based fusion sensor to avoid any collisions. When the sensor fusion framework detects any obstacle or vehicle, it sends a signal to the embedded board. After receiving this signal, the embedded board sends a signal to the motor to automatically reduce the speed of the car, which can control the speed of the car immediately.

Project Objectives

The project aims to design an accident avoiding system consisting of following features.

1. Detection of driver’s drowsiness by image processing technique and issuing an alert and controlling the vehicle’s speed upon drowsiness detection.
2. Detection of the drunk state of driver and stopping the engine if the alcohol consumption is above a predefined threshold.
3. Non-contact monitoring of the health parameters of driver and informing the nearest rescue system in case of any emergency.
4. Constantly monitoring the distance between car and any obstacle and stops the car automatically in case the driver doesn’t stops the forward motion.

Project Implementation Method

The project has four sub-parts. This section highlights the implementation method of each sub-system.

1. Drowsiness Detection Sub-system:

The drowsiness detection sub-system contained an open source 5-megapixel digital camera supported embedded system board Raspberry-pi loaded with Raspbian-OS, and Python-IDLE with Open-CV installed. The algorithm for this sub-system employs the image acquisition, image processing, feature extraction and classification techniques to achieve its respective task.

2. Alcohol Detection Sub-system:

The hardware required for this sub-system is merely a breath-analyzer sensor attached with the ARDUINO-board. Here the algorithm necessitates that the driver rehash the test while the vehicle is in activity. Amid these moving tests, if alcohol consumption by driver is found to be greater than a predefined threshold the system first warns the driver and in case of further alcohol consumption it activates the ignition interlock system.

3. Vital Signs Monitoring Sub-system:

This sub-framework is based on a method of detection suitable for observing ECG non-contact inside a running vehicle. As a non-contact estimation mode is required, a capacitive ECG observation (cECG) system has been explored and updated. To dispense with the necessity for additional cabling, a wireless communication module obsessed on the Bluetooth standard was accustomed transmit and store information on a measurement computer for off-line analysis.

4. Anti-collision Sub-system:

 This sub-framework is intended to figure out how to actualize a a minimum spacing for vehicles in rush hour gridlock in a reasonable manner. Here we present a parallel architecture for a sensor fusion detection system that combines a camera and 1D light detection and ranging (lidar) sensor for object detection. The framework consists of two identification techniques, one dependent on an optical stream, and the other utilizing lidar. The two sensors can effectively complement the defects of the other. The exact longitudinal precision of the item's area and its horizontal development data can be accomplished all the while. We completed the development of a fusion detection system with high reliability at distances of up to 20 m using a spatio-temporal alignment and a sensor fusion policy. Whenever any obstacle or other vehicle is detected by the sensor fusion frame work it sends a signal to the embedded board. Upon receiving this flag the embedded sends a flag to the motor in order to decrease the car’s speed automatically.

Benefits of the Project

The project has following competitive advantages as compared to other similar projects in the market:

1. It is the first system that not monitors the driver’s conditions (i.e drowsiness, drunk state and vital signs) but also monitors and controls the car with respect to the road environment (obstacle avoidance sub-part).
2. It monitors vital signs of the driver inn contact less manner which ensures driver’s comfort as the wired system is not suitable for driving context.
3. The system is quite flexible as it is capable of altering the control parameters for various drivers.
4. The system’s performance is equally maintained in all the driving environments having various light, temperature and road conditions.
5. The system is adoptable in various transportation systems.
6. It is a cost-effective pre-crash system.
7. It is an SoC that ensures its compact size and that it can be easily implemented inside the car.
8. It not only ensures the safety of the driver and car but also the safety of road pedestrians.
9. Compatible in all vehicles.

Technical Details of Final Deliverable

This project is sub-divided into four sub-parts. Each of the sub-part along with its objectives, hardware and software requirements, techniques employed and features is defined below:

a. Drowsiness detection sub-part:

• Hardware Requirements: 
  • Raspberry Pi
  • PI CAM with Night Vision
  • PI Case & heat sinks
  • PI Display & Case
  • SD Card
• Software Requirements: Raspbian operating system and Python-IDLE with Open-CV installed.

• Techniques used:

• Image processing.
• Face detection.
• Feature extraction.

• Algorithm:

The algorithm of this system is further divided into four sub-sections which are as follows:

 (A) Capture of real-time frames and face detection

 (B) Eye area extraction

 (C) Detection and tracking of the eye center

 (D) Eye-blink detection

b.  Alcohol detection sub-part:

• Hardware Requirements: 
  • Arduino MEGA ATmega2560 Development Board
  • ALCOHOL SENSOR (MQ-3)
  • GSM Modem (SIM900)
• Software Requirements:

• Arduino IDE programming platform.

• Algorithm:

 MQ3, the alcohol sensor works as a breathalyzer to calculate a driver's alcohol consumption. The sensor’s testing and observation depends on the Arduino Mega programming. There are different voltage levels / samples at the output when testing the MQ-3 Alcohol Sensor. The system displays the percentage of alcohol consumption and for this purpose we program as per our condition that converts voltage samples into percentage using the concept of mapping.

c.  Anti-collision sub-part:

• Hardware Requirements:
  • A Logitech HD PRO Webcam C920R camera
  • SRL-1 lidar sensor
  • Raspberry PI Board (the Fusion system needs to be connected with PI Board for processing purpose)

• Software Requirements: Raspbian operating system and Python-IDLE with Open-CV installed.

• Techniques Used:
  • Image processing.
  • Density-based spatial clustering of applications with noise (DBSCAN)
  • optical flow
  • Sensor Fusion Policy
• Algorithm: The detection algorithm is splits into three sections:

(A) Detection algorithm of SRL-1 Lidar

(B) Detection algorithm for camera

(C) Sensor Fusion Policy

d. Vital sign monitoring sub-part:

• Hardware Requirements: Capacitive Sensor with Shield, Instrumentation Amplifier (INA116), Notch Filter (using OPA 27), Instrumentation Amplifier (LT6010)

Methodology:

The capacitive plate behaves as the actual sensor; the first amplification stage comprised of a high input impedance instrumentation amplifier; Two notch filters to filter interfering signals the second amplification stage adding further gain and filtering.

Final Deliverable of the Project

HW/SW integrated system

Type of Industry

Transportation

Technologies

Artificial Intelligence(AI)

Sustainable Development Goals

Good Health and Well-Being for People, Decent Work and Economic Growth, Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com