Drowsiness is a critical factor impairing drivers’ performance in driving safely. There are several approaches in dealing with this issue based on human-machine interaction to detect drivers’ dozing off state, and then alert them to keep awake by sound or visual. These techniques fundamentally measure driver’s physical changes such as head angle, fatigue level and eyes states which are the indicators of drowsy state. However, they are limited in providing accurate and reliable results. Therefore, the project aims to achieve higher accuracy rate of drowsiness detection by using a very potential technology, electroencephalography (EEG) which is used widely in medical areas. Other than providing reliable result, the final product would bring more conveniences for customers with portability, easy-to-deploy and multi-device compatibility feature. In this project, its methodology first shows the strong correlation between drowsy state with brainwave frequency. Then a proposed system and testing plan are suggested based on the project objectives and available technologies. The final product is simply comprised of a hat with attached small electronic package used to record brainwave and a handheld device placing on dashboard of the car with an installed app. Finally, project management section will present in detail the human resources, scheduling, budget plan and risk analysis to show how it will be going to complete the project in six months.

1. Anti-drowsiness system for drivers
using EEG technology
December 2012
Nguyen Van Duc
Student ID : s3411503
Teacher : Dang Trong Trinh

2. CONTENTS
1 Introduction
2 Current approaches
3 Brainwave hat
4 Project management
5 Conclusion

3. Introduction Drowsy driving
The most danger to drivers & traffic participants [1]
10-20%
Road accidents in Europe [2]
http://www.end-your-sleep-deprivation.com/
100,000 road crashes 1,550 deaths
Result from drowsy driving in US [3]

4. Current approaches
Anti-Sleep Pilot Anti-Sleep Alarm Sleep Watcher-XR
http://www.antisleeppilot.com http://www.automotto.com http://www.exeros-technologies.com
Depending too much Not work when drivers Sensitive with
on the input data tilt the head backward, alternative illumination
right or left side[4] (~70% accurate rate) [1]

5. Brainwave hat Project objectives
 Higher accurate rate detection
(>98%)
 Portable & easy to deploy
 Applicable to multiple devices
http://www.sydney-neurofeedback.com.au/

6. Brainwave hat Brainwave and cognitive states
Beta 13-30 Hz, 5-20 μV
Alert
8-13 Hz, 20-60 μV
Alpha
Relax
Theta 4-8 Hz, 20-100 μV
Drowsy
0.5 – 4 Hz, 20-200 μV
Delta
Sleep
Adapted from [5-10]

7. Brainwave hat Product concept
The product is mainly comprised of a hat and a dock for handheld device

8. Brainwave hat System description
Amplifier
Electrodes
Alarm system
Filter
Processor
ADC
Wireless Wireless
Sender Receiver
Sender module Receiver module

9. Brainwave hat Product cost
Price/Unit Cost
Items Units
(USD) (USD)
Electrodes 2 18 36
OPA333 1 1.15 1.15
MSP430 1 4.3 4.3
CC2500 1 1.7 1.7
PCB 1 4 5
Other
1 5 5
components
Dock 1 10 10
Raspberry Pi B
1 40 40
board
Total $103.15

10. Brainwave hat Testing plan
- The virtual reality of car driving model suggested by Yin Niandong [11]
- The experiment will collect three main information:
+ The fatigue level of participants
+ The response of the system on drowsiness detection
+ The estimated fatigue level of supervisor

11. Project management Human resources

12. Project management Scheduling

13. Project management Budget plan
Initial fixed cost:
Price/Unit Total cost
Items Units
(USD) (USD)
Handheld devices
4 300 1200
(smartphones, tablets)
Sender’s components 4 60 240
Software license
(Cross-platform mobile 2 995 1990
development: Sencha Touch 2)
Furniture 1 500 500
Testing instruments 1 1000 1000
Total $4,930

14. Project management Budget plan
Monthly cost:
Cost for
Cost per
Items 6 months
month (USD)
(USD)
Nguyen Van Duc 1200 7200
Tran Manh Nguyen 700 4200
Salary
Ly Tue Hai 700 4200
Van Ngoc Mai 1000 6000
Electricity 50 300
Water 20 120
Utilities Internet 25 150
Telephone 25 150
Rental fee 250 1500
Total $23,820
Total cost: $4,930 + $23,820 = $28,550

15. Conclusion
 Drowsy driving is a serious threat to drivers and traffic
participants
 Non-reliable results of current solutions
 New approach: “Brainwave Hat” is considered as a best
solution
 “Brainwave Hat” project is feasible to implement within 6
months with available resources and technologies

16. REFERENCES
[1] W. H. Fei, Cheng and G. Xueming, "Real-Time Driver Drowsiness Tracking System " Nios II Embedded
Processor Design Contest - Outstanding Designs 2005, pp. 179-188, 2005.
[2] E. BEKIARIS, "System for effective Assessment of driver vigilance and Warning According to traffic risK
Estimation ", ed: CENTRE FOR RESEARCH AND TECHNOLOGY HELLAS, 2004.
[3] (18 November 2012). Research on Drowsy Driving Available:
http://www.nhtsa.gov/Driving+Safety/Distracted+Driving/Research+on+Drowsy+Driving
[4] (2007, 18 November 2012). Minister Issues Ban on Driving Anti-Sleep Alarm. Available:
http://www.legislation.vic.gov.au/domino/Web_Notes/newmedia.nsf/798c8b072d117a01ca256c8c0019bb01/5
1439959054b623bca2572b8007b358e!OpenDocument
[5] S. K. L. Lal and A. Craig, "A critical review of the psychophysiology of driver fatigue," Biological
Psychology, vol. 55, pp. 173-194, 2001.
[6] M. J. Wagner, "Effect of music and biofeedback on alpha brainwave rhythms and attentiveness," Journal of
Research in Music Education, vol. 23, pp. 3-13, 1975.
[7] L. Dailey, N. Krause, M. Roberts, K. Schuster, and C. Vang, "Effects of Physiological Relaxation on
Response Time," 2012.
[8] U. Svensson, "Blink behaviour based drowsiness detection," 0347-6049, 2004.
[9] J. L. Andreassi, Psychophysiology: Human behavior & physiological response: Lawrence Erlbaum, 2000.
[10] E. Grandjean and D. Barker, Fitting the Task to the Man, 1988.
[11] Y. Niandong, X. Ping, and T. Jing, "Construction of virtual experiment platform for research about vehicle
driving control at energy-saving," in Computer Supported Cooperative Work in Design (CSCWD), 2010 14th
International Conference on, 2010, pp. 427-431.

17. Q&A