Enhancing Road Safety: An IoT Based Driver Sleep Detection and

Alarming System for Accident Prevention

Adnan Yusufa

1 a

, Divya Asija

1 b

and R. K. Viral

1 c

Amity School of Engineering and Technology, Amity University Uttar Pradesh Noida, India

Keywords: IoT, Eye Blink, Driver Safety, Microcontroller PIC.

Abstract: This paper describes a driver drowsiness detection and alarming system based on the Internet of Things (IoT)

that uses an eye blink sensor to improve driver safety by monitoring drowsiness levels and providing timely

notifications to prevent accidents caused by driver exhaustion. To collect relevant driving-related data, the

system combines a number of sensors, including an eye motion sensor. The driver's eye movements and blink

patterns are recorded by the eye blink sensor and analysed by advanced algorithms to figure out sleepiness

levels. To collect data on the driver's behaviour and vehicle dynamics, additional sensors such as steering

angle sensors and accelerometers may be included. The acquired data is processed and analysed in real-time

using machine learning techniques to uncover patterns linked with lethargy. For proper interpretation,

algorithms evaluate characteristics such as blink duration, blink frequency, and eye closure duration when

determining fatigue degree. By continuously monitoring these characteristics, the system can detect early

signs of lethargy and take appropriate action. Finally, the IoT-based driver nap detection and alarm system

with an eye blink sensor is an excellent way for detecting driver weariness and notifying drivers in real time.

By utilizing innovative sensor technology, data processing algorithms, and networking, this system greatly

improves driver safety and mitigates the risks associated with fatigued driving.

1 INTRODUCTION

The IoT/AI-based driver sleep detection and alarm

system is a cutting-edge solution designed to address

the critical issue of drowsy driving, which poses a

significant risk to road safety. Fatigue and drowsiness

can impair a driver's cognitive abilities, reaction

times, and decision-making skills, leading to

accidents and potentially fatal outcomes. This

advanced system leverages the power of Internet of

Things (IoT) and Artificial Intelligence (AI)

technologies to monitor driver behavior, detect signs

of drowsiness, and provide timely alerts to prevent

accidents (Hussein et al.2022), (Knapik et al.,2019),

(Liu et al.,2019). By integrating a network of sensors

within the vehicle, the system continuously collects

data on various parameters, including eye

movements, steering patterns, vehicle acceleration,

and even the driver's heart rate. This wealth of data is

then processed and analysed in real-time using

https://orcid.org/0009-0007-6817-9632

https://orcid.org/0000-0003-4978-4216

https://orcid.org/0009-0001-7220-2742

sophisticated AI algorithms. By detecting patterns

indicative of drowsiness, such as prolonged eye

closures, erratic steering, or decreased heart rate, the

system can accurately assess the driver's level of

alertness. Once drowsiness is detected, the system

triggers an alarming mechanism to alert the driver and

prompt them to take immediate action (You et

al.,2020), (Gwak at al.,2020). This can involve

audible alarms, visual notifications, or even physical

feedback through vibrations or seat adjustments. By

providing timely alerts, the system aims to prevent

accidents caused by drowsiness-related factors and

promote driver safety. The IoT/AI-based driver sleep

detection and alarm system not only helps safeguard

the lives of drivers but also contributes to the overall

improvement of road safety (Tamanani et al. ,2021)

(Abbas et al.,2020), (Dong et al.,2019). By

proactively addressing the issue of drowsy driving,

this technology has the potential to significantly

reduce the number of accidents caused by driver

fatigue and save countless lives on the road.

Yusuf, A., Asija, D. and Viral, R.

Enhancing Road Safety: An IoT Based Driver Sleep Detection and Alarming System for Accident Prevention.

DOI: 10.5220/0012526100003808

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Intelligent and Sustainable Power and Energy Systems (ISPES 2023), pages 131-135

ISBN: 978-989-758-689-7

131

The major purpose of this study is to build and

implement a dependable, cost-effective, and real-time

IoT system to detect driver tiredness based on

physiological and behavioural data. The secondary

objective of this research is to identify the factors that

contribute to driver sleepiness. The effectiveness of

the system in reducing the number of collisions that

are the result of sleepy driving is another objective of

the research project.

2 METHODOLOGY

2.1 Architecture of the System

The Internet of Things-based driver sleep detection

system will include a variety of components, such as

physiological sensors, driving behavior sensors, data

processing units, and an alerting mechanism. The

architecture is developed in such a manner that it can

assure the capture and analysis of data in real-time

(Hussein et al.2022),(Knapik et al.,2019).Fig. 1

shows block diagram representation of the proposed

system where microcontroller PIC unit plays a major

role in detection drowsiness via eye blink sensor then

sending control and monitoring signals to end device

creating alarm signals and stopping of car as per

predefined program instructions.

2.2 The Accumulation and Processing

of Data

Wearable non-invasive sensors will be used to gather

physiological signals such as heart rate, eye

movement, and electroencephalogram (EEG) data

(Hussein et al.2022). These signals and data will be

collected. In parallel, data on the driver's behaviour,

such as the movement of the steering wheel and

departure from the lane, will be gathered by the

onboard sensors of the car. Processing and

interpretation of these data streams will be handled by

algorithms that are designed for machine learning.

2.3 Signal Processing and Analysis

The data that was acquired on the subject's

physiological state and driving behavior will first be

pre-processed to eliminate artefacts and noise. There

will be an application of techniques for feature

extraction to derive pertinent properties from the

signals. This will make it possible to identify patterns

that are related with driver sleepiness.

Fig. 2.1: Block Diagram Representation of the Proposed

System.

Fig 2.2: Flow chart of Algorithm.

2.4 Drowsiness Detection Algorithm

Based on the characteristics that were retrieved, a

dependable drowsiness detection algorithm will be

constructed. In order to recognize the early warning

indications of sleepiness, the algorithm will be trained

using a labelled dataset. This will ensure that the

detection will be both prompt and accurate.

2.5 Real-Time Alerting Mechanism

If the IoT system determines that the driver is

becoming sleepy, it will immediately send a warning

message to the driver's device. It is planned to

conduct research into a number of different kinds of

warning, including audio-visual signals and haptic

input, in order to identify the one that is best capable

of rousing a sleepy driver.

2.6 Evaluation and Testing

One of the important stages in moving forward with

the prototype implementation is the collection and

preparation of datasets. As an outcome, in this

section, we will go over the technical aspects of the

data-gathering technique, including participant

selection and the various driving scenarios. In

ISPES 2023 - International Conference on Intelligent and Sustainable Power and Energy Systems

132

addition, the techniques used to ensure the dataset's

accuracy and consistency will be described.

3 WORKING MECHANISM

The process of an IoT/AI-based driver sleep detection

and alarming system involves several key steps.

Here's an overview of the process: Sensor data

collection: The system incorporates various sensors

placed strategically within the vehicle to capture

relevant data. These sensors continuously monitor

and record information such as eye movements,

steering patterns, vehicle acceleration, heart rate, and

other parameters related to driver behaviour and

vehicle dynamics.

Data Preprocessing. The collected sensor data is

pre-processed to remove noise, normalize values, and

ensure data consistency. This step may involve

filtering, scaling, and feature extraction techniques to

prepare the data for further analysis. Feature

extraction: AI algorithms are applied to the pre-

processed data to extract meaningful features that can

be indicative of driver drowsiness. For example, eye-

tracking data may be analysed to identify patterns

such as prolonged eye closures or a decrease in blink

rate. Steering angle data may be examined for

irregular or inconsistent patterns.

Drowsiness Detection. Machine learning algorithms,

such as deep neural networks or support vector

machines, are trained on labelled datasets to classify

the driver's alertness level based on the extracted

features. These algorithms learn patterns and

correlations between the sensor data and drowsiness

indicators, enabling accurate drowsiness detection.

Real-Time Monitoring. The trained model is

deployed in real-time to continuously monitor the

driver's behavior. As the system receives new sensor

data, it processes it through the trained model to

predict the driver's alertness level. This monitoring

occurs in real-time, allowing for immediate detection

of drowsiness indicators.

Alarming Mechanism. When the system detects

signs of drowsiness, it activates the alarming

mechanism to alert the driver and prevent potential

accidents. The alarms can take various forms, such as

audible alerts, visual notifications on the dashboard,

or physical feedback through seat vibrations or

adjustments. The alarming mechanism aims to

capture the driver's attention and prompt them to take

corrective action.

Continuous Monitoring and Adaptation. The

system continuously monitors the driver's state

throughout the journey and adapts its detection

algorithms as necessary. It can learn from new data

and update its models to improve accuracy over time.

This adaptive capability ensures that the system

remains effective even as the driver's behaviour or

environmental conditions change. By following this

process, the IoT/AI-based driver sleep detection and

alarming system can reliably identify drowsiness in

drivers and provide timely alerts, thereby enhancing

road safety and preventing accidents caused by driver

fatigue.

4 RESULT AND ANALYSIS

The findings and analyses that were reported in the

study article revealed that the Internet of Things-

based driver sleep detection and warning system was

effective in increasing overall road safety.

Technology can avoid accidents that are caused by

drivers who are drowsy by properly identifying their

state of tiredness and taking measures that are suitable

for the situation. This would make the roads safer for

everyone who uses them. The findings highlight how

important it is to continue research and development

in this field to enhance technology and encourage the

automobile sector to use it.

The development stages of the prototype are shown

in the subsequent section. It will showcase how the

hardware of prototype has undergone through

different stages before reaching the final version.

4.1 Developing Stages of the Driver

Sleep Detection and Alarming

System

Fig. 4.1: Development stage 1.

Enhancing Road Safety: An IoT Based Driver Sleep Detection and Alarming System for Accident Prevention

133

Fig. 4.2: Development stage 2.

Fig. 4.1 and 4.2 show the prototype's development

at various stages leading up to the final version.

Fig. 4.3: Eyeglasses for sensing driver drowsiness.

Table 5.1: Experimental Observations at different Time

Intervals.

Time (sec) | Eye State | Action

Taken

0 | Open | None

1 | Open | None

2 | Closed | None

3 | Closed | None

4 | Closed | None

5 | Closed | Car Stopped

6 | Open | Brakes

Applied

7 | Open | Brakes

Applied

8 | Open | Brakes

Applied

9 | Open | Brakes

Applied

10 | Open | Brakes

Applied

11 | Open | Brakes

Applied

12 | Open | None

13 | Open | None

14 | Closed | None

15 | Closed | None

16 | Closed | None

17 | Closed | Car Stopped

18 | Open | Brakes

Applied

19 | Open | Brakes

Applied

20 | Open | Brakes

Applied

21 | Open | Brakes

Applied

22 | Open | Brakes

Applied

23 | Open | None

24 | Open | None

The eye state is shown in table 5.1 as "Open" or

"Closed," depending on whether the driver's eyes are

open or closed at the given moment. The "Action

Taken" column displays the relevant action that the

system took in accordance with the specified rules:

a. If the eyes are closed for longer than three

consecutive seconds, the system will halt the

vehicle after five seconds (row 5 in the

table).

b. As long as the eyes are closed once the car

has stopped, the system applies the brakes

once every second (rows 6 to 11).

c. The car resumes normal operation when the

driver's eyes open (row 12), and nothing

more happens.

ISPES 2023 - International Conference on Intelligent and Sustainable Power and Energy Systems

134

5 CONCLUSION AND FUTURE

WORK

In conclusion, the IoT/AI-based driver sleep detection

and alarm system is a powerful technology that

addresses the critical issue of drowsy driving. By

leveraging sensors, data processing techniques, and

machine learning algorithms, the system can monitor

the driver's behaviour in real-time, detect signs of

drowsiness, and provide timely alerts to prevent

accidents. The system collects data from various

sensors placed within the vehicle, including eye-

tracking cameras, steering angle sensors,

accelerometers, and heart rate monitors. This data is

processed and analyzed using AI algorithms to extract

relevant features indicative of drowsiness. Machine

learning models are trained to classify the driver's

alertness level based on these features.

When drowsiness is detected, the system activates

an alarming mechanism, alerting the driver through

audible, visual, or physical means. This prompt

intervention aims to awaken the driver, increase their

alertness, and prevent accidents caused by

drowsiness-related factors. The IoT/AI-based driver

sleep detection and alarming system not only

enhances driver safety but also contributes to overall

road safety. By proactively addressing the issue of

drowsy driving, it has the potential to save lives,

reduce accidents, and minimize the devastating

consequences of driver fatigue. With further

advancements in IoT and AI technologies, this system

holds promise for continued improvement and

refinement, making roads safer for everyone. By

combining technology and human vigilance, we can

create a future where drowsy driving becomes a thing

of the past, ensuring safer and more secure journeys

for all.

REFERENCES

Hussein, R. M., Miften, F. S., & George, L. E. (2022).

Driver drowsiness detection methods using EEG

signals: a systematic review. Computer Methods in

Biomechanics and Biomedical Engineering, 1–13.

https://doi.org/10.1080/10255842.2022.2112574

Knapik, M., & Cyganek, B. (2019). Driver’s fatigue

recognition based on yawn detection in thermal

images. Neurocomputing, 338, 274–292.

https://doi.org/10.1016/j.neucom.2019.02.014

Liu, W., Qian, J., Yao, Z., Jiao, X., & Pan, J. (2019).

Convolutional Two-Stream Network Using Multi-

Facial Feature Fusion for Driver Fatigue

Detection. Future Internet, 11(5), 115.

https://doi.org/10.3390/fi11050115

You, F., Gong, Y., Tu, H., Liang, J., & Wang, H. (2020). A

Fatigue Driving Detection Algorithm Based on Facial

Motion Information Entropy. Journal of Advanced

Transportation, 2020, 1–17. https://doi.org/10.1155/

2020/8851485

Gwak, J., Hirao, A., & Shino, M. (2020). An Investigation

of Early Detection of Driver Drowsiness Using

Ensemble Machine Learning Based on Hybrid

Sensing. Applied Sciences, 10(8), 2890.

https://doi.org/10.3390/app10082890

Tamanani, R., Muresan, R., & Al-Dweik, A. (2021).

Estimation of Driver Vigilance Status Using Real-Time

Facial Expression and Deep Learning. IEEE Sensors

Letters, 5(5), 1–4. https://doi.org/10.1109/

lsens.2021.3070419

Abbas, Q., & Alsheddy, A. (2020). Driver Fatigue

Detection Systems Using Multi-Sensors, Smartphone,

and Cloud-Based Computing Platforms: A

Comparative Analysis. Sensors (Basel,

Switzerland), 21(1). https://doi.org/10.3390/s21010056

Dong, N., Li, Y., Gao, Z., Ip, W. H., & Yung, K. L. (2019).

A WPCA-Based Method for Detecting Fatigue Driving

From EEG-Based Internet of Vehicles System. IEEE

Access, 7, 124702–124711. https://doi.org/10.1109/

access.2019.2937914

Enhancing Road Safety: An IoT Based Driver Sleep Detection and Alarming System for Accident Prevention

135