International Journal of Engineering Advanced Research eISSN: 2710-7167 [Vol. 4 No. 1 March 2022]
Journal website: http://myjms.mohe.gov.my/index.php/ijear
DEVELOPMENT OF DRIVING FATIGUE STRAIN INDEX USING FUZZY LOGIC TO QUANTIFY IMPAIRMENT RISK
LEVELS OF COGNITIVE SKILLS IN VEHICLE DRIVING
Muhammad Shafiq Ibrahim1, Seri Rahayu Kamat2*, Syamimi Shamsuddin3 and Mohd Hafzi Md Isa4
1 2 3 Fakulti Kejuruteraan Pembuatan, Universiti Teknikal Malaysia Melaka, Hang Tuah Jaya, Durian Tunggal,
Melaka, MALAYSIA
4 Work-Related Road Safety Management Cluster, Malaysian Institute of Road Safety Research (MIROS), Kajang, Selangor, MALAYSIA
*Corresponding author: [email protected]
Article Information:
Article history:
Received date : 24 February 2022 Revised date : 4 March 2022 Accepted date : 6 March 2022 Published date : 19 March 2022 To cite this document:
Ibrahim, M. S., Kamat, S. R., Shamsuddin, S., & Md Isa, H. (2022).
DEVELOPMENT OF DRIVING FATIGUE STRAIN INDEX USING FUZZY LOGIC TO QUANTIFY IMPAIRMENT RISK LEVELS OF COGNITIVE SKILLS IN VEHICLE DRIVING. International Journal of Engineering Advanced Research, 4(1), 121-139.
Abstract: Fatigue-related road accident is now being an alarming issue in many countries, including Malaysia.
This paper developed a driving fatigue strain index (DFSI) by using fuzzy logic approach to analyze the impairment risk levels of three significant cognitive skills that influence driving fatigue, namely (i) alertness, (ii) attention and (ii) decision making. A human-related factor which is oxygen saturation was also analysed. The DFSI projected an advance warning method to alert users through a systematic analysis, proposing a solution to avoid road crashes. The fuzzy logic is developed by building fuzzy inference system (FIS) through Fuzzy Logic Toolbox function by MATLAB software. The input variables (alertness, attention, decision making and oxygen saturation) and an output variable (DFSI) were first divided into fuzzy sets by using fuzzy logic designer interface. Next, the shapes of all the membership functions associated with the input and output variables were developed by using membership function (MF) editor. IF- THEN rules statement was then employed to formulate the conditional statements of membership functions. Finally, the rule viewer interpreted all the rules and illustrated how the shape of particular membership functions influenced the overall results, eventually, reaching the DFSI degree.
1. Introduction
Traffic accident is a worldwide tragedy with an ever-increasing trend. It is predicted that more than 1.3 million fatalities and 20 to 50 million injuries occur every year due to global traffic accidents (WHO, 2018). Most experts agree that driving fatigue is a leading factor of road crashes (Li et al., 2020) and is believed to account for 20-30% of all vehicle crashes (Rau, 2005). This is in agreement with the report released by the Malaysian Institute of Road Safety Research (MIROS) as 80.6% of road accidents in Malaysia in year 2016 are attributed to driving fatigue (MIROS, 2021). The number of road accidents in Malaysia has increased alarmingly with a record of 414,421 in 2010 to 567,516 cases in 2019 as shown in Figure 1. Driving fatigue is defined as a feeling of extreme physical and mental tiredness which commonly occur due to lack of adequate sleep (Davidović et al., 2018) and long distance of continuous driving (Fell, 2019). In such situation, vital driver’s cognitive skills like alertness, attention and decision making (Mullette- Gillman et al., 2015) are adversely impaired, hence, enhancing the probability of being involved in accidents (Alonso et al., 2016).
Figure 1: Statistics of Malaysia Road Accidents From 2010-2019 (MIROS, 2021)
Cognitive skills or general intelligence, is essential for human adaptation and survival especially when involving complex activities that require multiple interaction of cognitive abilities like driving. This ability includes the capacity to reason, plan, solve problems, think abstractly, understand complex suggestion, learn fast and learn from previous experience (Plomin, 1999).
414,421 449,040
462,426 477,204
476,196 489,606
521,466 533,875
548,598 567,516
400,000 420,000 440,000 460,000 480,000 500,000 520,000 540,000 560,000 580,000
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Road accident, cases
Year
and unsafe. The DFSI degree is expected to be useful to quantify the current driver’s condition in terms of cognitive abilities. Hence, this provides an advance warning to the driver to help avoid accidents and to ensure everyone reaches their destination without harm.
Keywords: driving fatigue, cognitive skills, fuzzy logic, risk levels.
identified as one of the factors that increases the risk of safety errors (Anderson et al., 2005) and deteriorates driving performance (Lafont et al., 2008), resulting in driving fatigue. Morris &
Miller, (1996) who performed an experiment using a flight simulator, concluded that pilots that suffer cognitive skills impairment, result in drowsiness after about four and a half hours, hence, increasing the possibility in making errors during operations.
Therefore, this study has developed a driving fatigue strain index (DFSI) to quantify and analyse the impairment risk level of three significant cognitive skills that contribute to the driving fatigue, namely (i) alertness, (ii) attention and (ii) decision making. Oxygen saturation factor which is a human factor that influenced driving fatigue was also analysed. The index provides an advance warning to alert road drivers of the risk level and condition, whether the driving was safe, slightly unsafe or unsafe, thus, minimizing the traffic crash probability.
2. Literature Review
Three imperative cognitive skills for driving have been found to be investigated in literature, namely (i) alertness, (ii) attention and (iii) decision making.
2.1 Alertness
Driving involves both physical and mental activities, where drivers need to maintain their alertness at an optimum level (May & Baldwin, 2009). Alertness is a general condition of cognitive readiness, reflected in cortical arousal. The rise in alertness will benefit specific cognitive abilities, including vigilance performance (Shahid et al., 2016) which is good for driving activity. Poor alertness levels are related with higher risks for accidents, mistake and injury (Folkard & Tucker, 2003). A study showed that, the fatigue-related road crashes mostly occur during the two periods of physiological decrease of alertness: (i) early morning hours from 2:00 am to 6:00 am and (ii) afternoon period from 1:00 pm to 4:00 pm (Di Milia et al., 2011). (Mohamed et al., 2012) found that most of the fatal road crashes in Malaysia occur in the early morning hours. The alertness decline may be due to accumulated fatigue caused by two factors. The first factor is time-on-task effect like long driving period. (Jovanis et al., 2012) found that drivers tend to fatigue with increased accident risk after four hours behind the wheel without rest. This is in agreement with (Jing et al., 2013) as the analysis on four driver’s physiological information for six hours simulated driving test, concluded that 235 minutes should be an ideal consecutive driving time threshold.
Meanwhile, (Yan-li & Yu-long, 2009) studied the driving performance of forty drivers and found significant drop in driving performance after 3.5 hours consecutive driving time. The second factor is sleep deprivation. (Demirdöğen Çetinoğlu et al., 2015) found that the neurocognitive measures of reaction time and driving performance are impaired after sleep deprivation. Another study found that the shorter the drivers sleep within the 24-hours before accident, the more odds that the drivers should be responsible for accidents (Tefft, 2018).
2.2 Attention
Attention is a vital cognitive process to engage with surroundings, interacts with arousal and the sleep-wake spectrum. In simpler terms, attention is the ability to stay focus in a particular task while avoiding distractions from surrounding (Lindsay, 2020). Attention is not the same as alertness, as alertness is the state of being aware and prepared to react, while attention is more to mental focus (VandenBos, 2007). However, it is often difficult to maintain high attention level while driving for an extended period of time. Under this circumstance, the drivers tend to get distract, diverting the attention away from the driving task, eventually causing road accidents (Walker & Trick, 2018). Driver distraction due to attention deterioration is a leading cause of road collision (Klauer et al., 2006). A study examined brain activity while engaging in simulated driving tasks, found that even minor lapses in task attention can diminish the ability to recognize hazards (Schweizer et al., 2013). For example, (Strayer et al., 2003) investigated the impact of speaking on a phone while driving diminished attention and response to a vehicle braking in front of the driver. Mind-wandering while behind the wheels can also divert attention from the visual field (Barron et al., 2011). Mind-wandering is a universal human experience in which the focus of attention is temporarily shifts away from the perceptual demands of the main task towards the processing of more personal goals (Smallwood et al., 2006). (Walker & Trick, 2018) found the tendency to mind-wandering while driving is related to the previous night’s sleep rather than the tiredness. (Burdett et al., 2016) concluded that young adults are prone to mind-wandering than young drivers due to their greater propensity to sleep deprivation (Pediatrics, 2014).
2.3 Decision Making
Decision making is a high-level cognitive process resulting in the selection of a belief or a course of action among a set of alternatives based on particular criteria (Prezenski et al., 2017). During driving, drivers are destined to cope infrequent, yet, safety-critical situations where fatalities are inevitable and predictable (Goodall, 2014). Therefore, making the right decisions in a limited amount of time is crucial (Samuel et al., 2020). For example, within a short period of time to brake, drivers may have to decide to save a group of pedestrians crossing the road by veering to the side or striking only one pedestrian. Under this intense situation, it is infeasible for the drivers to make the best decision as the remaining time to perform an action (swerve right or left) decreases. The decision making while driving is mainly influenced by three factors, namely (i) driver, (ii) vehicle and (iii) road environment. The driver’s physiological and psychological characteristics are the main factors influencing decision making (Wang et al., 2019), such as gender (Barr Jr et al., 2015), age (Körber et al., 2016), emotion (Bogdan et al., 2016), attitude (Ramisetty-Mikler &
Almakadma, 2016), cognitive skills (Gish et al., 2017), habit (Chung, 2015) and driving behaviour (Sentoff et al., 2015). In terms of vehicle factors, vehicle motion condition and performance are the main causes (Suh et al., 2016). Road-environment factors are categorized into three, namely (i) external weather environment such as ice, fog and rain (Li et al., 2015) (ii) road driving environment like road alignment (Bella, 2015), road facilities (Shackel & Parkin, 2014) and roadside environment (Bella, 2013) and (iii) cockpit environment including physical environment like temperature (Simion et al., 2016), noise (Ulrich et al., 2014) and driving sub-tasks like using a mobile phone (Li et al., 2016) and listening to a radio (Shen & Neyens, 2017).
3. Method
The ultimate objective of this study is to develop a driving fatigue strain index (DFSI) to quantify and analyse impairment risk level of three significant cognitive skills which consists of, (i) alertness, (ii) attention and (iii) decision making and a human factor, which was oxygen saturation level that influenced driving fatigue. This section provides a brief explanation of the DFSI development, comprising (i) knowledge acquisition, (ii) Fuzzy Inference System (FIS) development and (iii) experimental set up.
3.1 Knowledge Acquisition
The process started with knowledge acquisition to determine the significant driving cognitive skills that influenced driving fatigue. Acquiring knowledge is a vital process for intellectual growth and innovation, comprising elicitation, modeling and knowledge validation (Kaba & Ramaiah, 2020).
In this study, the significant cognitive skills that influenced driving fatigue were determined by gathering information from reliable sources like previous studies, pre-survey and interview.
Journal articles from online databases like Google Scholar and Scopus, survey reports and editorials were reviewed and analyzed to identify the cognitive skills that influenced driving fatigue. Three cognitive skills were determined, namely (i) alertness, (ii) attention and (iii) decision making. Another human factor that influences driving fatigue like oxygen saturation level was also analysed. The impairment risk levels for each cognitive skill and oxygen saturation level were quantified as found in previous studies as reflected in Table 1. A pre-survey and interview were then conducted among twenty Malaysian road users with the age range of 25 to 40 years old. The objective is to examine and gain information regarding the cognitive factors that influenced driving fatigue. A set of questionnaires involving how the alertness, attention and decision making affect driving fatigue were provided and interview sessions were conducted to acquire each respondent’s personal experience while driving, history of accidents caused by fatigue and suggestions to prevent the fatigue-related road accidents
Table 1: Previous Studies Used as References Cognitive Skills
Type Frequency, Hz
Physiological Condition
Explanation/Reference
Delta, δ 0-4 Deep rest, sleep Visible during sober state usually when sleeping and under deep anesthesia (Liao et al., 2018).
Theta, θ 4-8 Deeply relaxed,
inward focused
Visible during early stage of sleep, which influence the alertness (Kulkarni et al., 2019).
Alpha, α 8-13 Very relaxed, calm, passive attention
The frequency promotes a feeling of relaxation, which is good for attention (Peylo et al., 2021)
Beta, β 13-30 Alert, relaxed,
anxiety dominant, active
Visible when involved in complex thought processes like problem solving or decision making (Koudelková &
Strmiska, 2018).
Oxygen Saturation
Condition Range, % Reference
Normal 95-100 (Peterson, 2007)
Abnormal <95
Table 2 summarizes the impairment risk level for identified cognitive skills and oxygen saturation to indicate driving fatigue based on previous studies in the Table 1 as the guidelines. The summary of the risk level was used later for the DFSI development.
Table 2: Cognitive Skills Risk Levels
3.2 Fuzzy Inference System (FIS) development
The DFSI was developed by using an artificial intelligence technique, known as fuzzy logic. Fuzzy logic is similar to human reasoning system (Thakkar et al., 2021), therefore it can manage the uncertainties found in the data. The fuzzy logic was developed by building the fuzzy inference system (FIS) through the Fuzzy Logic Toolbox function of MATLAB software. FIS is an approach that formulates the values in the input vector and assigns values to the output vector by using some sets of rules (Kalogirou, 2014). Four steps were used to build, edit and view FIS, namely (i) fuzzy logic designer, (ii) membership function (MF) editor, (iii) rule editor and (iv) rule viewer.
3.2.1 Step 1: Fuzzy Logic Designer
Fuzzy logic designer opened and displayed a diagram of the FIS with the names of each input variable on the left and each output variable on the right. This study used Mamdani-type inference as it is intuitive, well suited to human input and has widespread acceptance (Garza-Ulloa, 2018).
In this study, the input variables were (i) alertness, (ii) attention and (iii) decision making and (iv) oxygen saturation level. The output variable was the DFSI linguistic variable. In short, four inputs and one output system were constructed as shown in Figure 2. There are numerous techniques offered for data defuzzification in this interface such as bisector, centroid, medium of maximum (MOM), large of maximum (LOM) and small of maximum (SOM). However, this study employed centroid method for data defuzzification because the approach determines the center of area of fuzzy set and returns the corresponding crisp number (Wang, 2001).
Impairment Risk Level(s)
Alertness, Hz Attention, Hz Decision Making, Hz Oxygen saturation, 100 Fairly fit: 4-5 Fairly fit: 8-9 Fairy fit: 13-18 Normal 95-100
Fit: 6-8 Fit: 10-11 Fit: 19-24 Abnormal <95
- Very fit: 12-13 Very fit: 25-30 -
Figure 2: Fuzzy Inference System (FIS) Editor
3.2.2 Step 2: Membership Function (MF) Editor
The purpose of membership function (MF) editor is to define the shapes of all the membership functions associated with all the input and output variables. In this study, there were four inputs, namely (i) alertness, (ii) attention, (iii) decision making and (iv) oxygen saturation, whereas DFSI was the only output of these variables. The input and output variables were divided into fuzzy sets, defined by the membership function. Fuzzy sets are sets that indicate the degree of membership functions in that sets. In this study, the degree of membership functions was based on the risk criteria and risk levels of the cognitive skills as stated in previous studies as illustrated in Table 2.
For example, the linguistic variable of ‘alertness’ was divided into two membership functions, namely (i) fairly fit and (ii) fit with the degree of memberships of 4-5 Hz and 6-8 Hz respectively.
Meanwhile, the output variable of ‘DFSI’ was divided into three membership functions, namely (i) safety, (ii) slightly unsafe and (iii) unsafe with the degree of memberships of (i) 0-10 Hz, (ii) 10-20 and (iii) 20-30 Hz respectively. The membership functions for all variables are show in Figure 3-7.
There are approximately ten membership functions to represent a fuzzy set, however, in this study, two typical types of functions were used to represent the input and output variables, namely (i) triangular MF and (ii) Gaussian MF. Triangular MF is the most favorable MF in practice as these functions are formed by straight lines, offering an advantage of simplicity to graphically represent the fuzzy system and fast for computation. Meanwhile, Gaussian MF is a popular technique for specifying fuzzy sets due to the smoothness and concise notation characteristics, resulting in smooth and nonzero curve at all points. In comparison, triangular and Gaussian MF are more feasible than other MFs for many kinds of problems (Sadollah, 2018). However, (Zhao & Bose, 2002) found that the triangular MF is better than Gaussian MF by comparing the response of the system.
DFSI DFSI_Driving
(mamdani) Alertness
Attention
Oxygen Saturation Decision Making
Figure 3: Membership Function (Alertness) _ Triangular Function
Figure 4: Membership Function (Attention) _ Triangular Function
Figure 5: Membership Function (Decision Making) _ Gaussian Function 0.5
1
4 4.5 5 5.5 6 6.5 7 7.5 8
Fairly Fit Fit
0
0.5 1
8 8.5 9 9.5 10 10.5 11 11.5 12
Fairly Fit Fit
0
Very Fit
12.5 13
0.5 1
14 16 18 20 22 24 26 28
Fairly_Fit Fit
0
30
Fairly Fit Fit Very Fit
Figure 6: Membership Function (Oxygen Saturation) _ Gaussian Function
Figure 7: Membership Function (DFSI) _ Gaussian Function
3.2.3 Step 3: Rule editor
The aim of rule editor is to edit and define the behavior of the system by automatically constructing the rule statements. The rules development was based on the descriptions of the input and output variables defined by using Fuzzy Logic Designer. The IF-THEN rules statement was employed to formulate the conditional statements that comprised fuzzy logic. The Mamdani-type model was used as the inference tool. This system comprised IF-THEN rules of the form “IF A is X THEN B is Y”. The IF part “A is X” is defined as the antecedent of the rule, where in this study, “A” was the input variable (alertness, attention, decision making and oxygen saturation) and “X” was the membership functions. Meanwhile, THEN part “B is Y” is known as the consequent or alternatives recommendation of the rule, where in this study, “B” was the output variable (DFSI) and “Y” was the outcome (safe, slightly unsafe or unsafe). The rule editor GUI was utilized to edit the list of inference rules as shown in Figure 8. Thirty rules’ statements were constructed in this study. Table 3 shows the three examples from thirty rules generated by using Mamdani- type inference model.
0.5 1
93 94 95 96 97 98 99 100
0
Normal Abnormal
0.5
1
0 0.5
5 10 15 20 25 30
0
Safe Slightly unsafe Unsafe
Figure 8: Rule Editor
Table 3: Example of Rules Set by Mamdani-type Inference Model
3.2.4 Step 4: Rule viewer
The purpose of the rule viewer is to view the road map of the whole fuzzy inference process. As mentioned earlier, the output was represented by the linguistic variable DFSI, described by three membership functions, namely (i) safe, (ii) slightly unsafe and (iii) unsafe. The rule viewer interpreted the overall fuzzy inference process and showed how the shape of particular membership functions influenced the overall results as shown in Figure 9. Each rule defined a row of plots, and each column represented the variable. In this study, 30 rules were constructed by using Mamdani-type inference model, generated by the interaction of the four input and one output variables. The first four columns of plots (yellow color) indicated the membership functions (IF- part). Meanwhile, the fifth column of plots (blue color) represented the membership functions (THEN-part). The de-fuzzified output value represented the measure of the DFSI degree, derived from the combination of the input values as shown on top of the diagram and displayed as a bold vertical line at the bottom of the plot.
No Rules set
Rule 3 IF Alertness is Fairly Fit and Attention is Very Fit and Decision Making is Fairly Fit AND Oxygen Saturation is Normal THEN DFSI is Safe
Rule 5 IF Alertness is Fairly Fit and Attention is Fit and Decision Making is Fit AND Oxygen Saturation is Abnormal THEN DFSI is Slightly Unsafe
Rule 26 IF Alertness is Fit and Attention is Very Fit and Decision Making is Very Fit AND Oxygen Saturation is Abnormal THEN DFSI is Unsafe
Figure 9: Rule Editor: Road Map for the Whole FIS on Rule Views
3.3 Experimental Set Up
Series of experiments utilizing driving simulator study were conducted to obtain samples of data sets to test and demonstrate the functioning and reliability of the developed fuzzy inference system.
The cognitive ability refers to a person’s ability to perform various mental activities. In this study, electroencephalogram (EEG) was used to examine the driver’s brain activity. Recently, a great number of researchers conduct investigations on the advantages of EEG due to its characteristics that can reflect ongoing reaction of the brain status (Liu et al., 2020). Two subjects (one male and one female) who at least, had two years driving experiences were selected. All the subjects had normal BMI (18.5 to 24.9), healthy bodies, no daily medication and been refrained from drinking beverages containing caffeine and alcohol. Emotiv Epoc X, a wireless EEG device with 14 electrodes, was used to examine participant’s brain signals. The EEG categorizes the signal into four frequencies, namely (i) delta (0- 4Hz), (ii) theta (4-8Hz), (iii) alpha (8-13Hz) and (iv) beta (13-30Hz). Each frequency contained major characteristic waveforms, indicating the cognitive abilities (alertness, focusing and decision making) as shown in the Table 1. In this study, delta (0- 4Hz) was ignored as this frequency indicated deep sleep, which was not considered a cognitive skill. Meanwhile, to measure the subject’s oxygen saturation, pulse oximeter was used. Thirty-two experimental runs were carried out. Table 4 shows the samples of experimental run and the best
IF-part (Input variable) = Value
THEN-part (DFSI degree) Alertness=4.57 Attention= 12.3 Decision Making= 18 Oxygen Saturation= 93 DFSI=9.42
DFSI degree
plot
Figure 10 shows the driving simulator experimental set up. First, the Emotiv Epoc X headset was charged until the indicator light on the headset turned green. The sensor felts were then hydrated by placing them in a glass contained-saline solution before squeezed out the excess fluid to obtain a good contact with the scalp. Next, each hydrated sensor felt was fitted into each sensor arm. The headset was then set on the subject’s head. The headset was connected into a laptop by using a USB receiver dongle. The driving simulator consists of a steering wheel, gas and brake pedals. A desktop was placed at a distance about 80 cm from the participant’s eye, displayed the road environment and the current speed. The participants were required to drive on the simulator for twenty minutes and the brain frequencies were recorded and analyzed. The oxygen saturation level was recorded before and after driving by using a pulse oximeter.
Figure 10: Experimental Set Up
Table 4: Sample of Experimental Run and the Best Data Sets to Demonstrate the FIS Run Factor 1 Factor 2 Factor
3
Response 1 (Alertness)
Response 2 (Attention)
Response 3 (Decision
Making)
Response 4
Data Set
A: Time Exposure,
min
B: Type of road
C:
Gender
Theta, θ (Hz)
Alpha, α Hz
Beta, β (Hz)
Oxygen saturation
(%)
3 15 Winding Male 4.57 10.6 13.9 97.7 1
13 30 Uphill Female 4.57 12.3 18 93.4 2
21 30 Straight Female 4.48 10.5 24.9 98.13 3
Emotiv Epoc X
Pulse Oximeter
4. Results and Discussion
The fuzzy inference system was tested and demonstrated through three data sets obtained by series of experiments through a driving simulator study as presented in Table 4. The values of each input variable were entered into the rule viewer to get the defuzzification value, representing the degree of DFSI. Figures 11 (a), (b) and (c) show the FIS rule viewer, which reflected the results of the analysis. The rule viewer automatically showed the DFSI values once the input variables were entered into the system. The yellow color in the first four column indicated the IF-part and the blue color in the fifth column represented the THEN-part of the generated rules for each membership function. The DFSI value was shown on the top fifth column and displayed as a bold vertical line at the bottom of the plot. The DFSI values for data set 1, set 2 and set 3 were 5.68, 11.3 and 24.3 respectively. Figure 12 depicts the membership function plot for the output variables of set 1, set 2 and set 3. Table 5 summarizes the results from the demonstration.
(a) (b)
(c)
Figure 11: FIS Rule Viewer of (a) Set 1, (b) Set 2 and (c) Set 3
Figure 12: Memberships Function Plots for Output Variable of Set 1, Set 2 and Set 3 Table 5: Summary of Results
5. Conclusion and Recommendation
The present study introduces a novel approach to minimize road accidents through the development of DFSI using fuzzy logic method. The application of fuzzy logic system has more benefits than a traditional mathematical method due to its characteristic that deals with uncertainties derived from fuzziness and is capable to provide the most effective solution for complex issues, including road safety. Three significant cognitive skills, namely alertness, attention and decision making together with one human factor that influence driving fatigue which is oxygen saturation were investigated. The developed system was tested to examine its reliability by using three data sets obtained by a driving simulator study, yielding 5.68, 11.3 and 24.3 DFSI values, representing safe, slightly unsafe and unsafe respectively. The results show a great capability of the developed system to provide an advance-warning signal regarding the user’s current condition through a systematic analysis.
Recently, there is growth in the use of computer-based decision support system (DSS) to control the impacts of transportation. DSS is an interactive computer-based information system with an organized collection of analytical tools, graphical user interfaces and databases to support decisions making and timely problem-solving. The system enables decision makers to integrate directly with computers to collect useful information from raw data, personal knowledge and documents to determine and solve problems for structured and unstructured issues and perform decision making. The DSS for transportation offers many functionalities and solve numerous categories of transportation decision problems including road safety. Thus, in realizing the urgent
Set DFSI Driving condition
1 5.68 It is safe to drive. The driver is in excellent condition and need to maintain throughout the driving.
2 11.3 It is slightly unsafe to drive. The driver is in condition where there is high possibility to involve in road accidents.
3 24.3 It is unsafe to drive. The driver is in bad condition and strictly not allowed to drive as this condition can cause road accident.
1
0 0.5
5 10 15 20 25 30
0
Safe Slightly unsafe Unsafe
Set 1 (5.68)
Set 2 11.3
Set 3 24.3
need for solving the fatigue-related road accidents, the developed FIS system will be used for the DSS development in the future study.
6. Acknowledgement
The authors would like to thank the Ministry of Higher Education (MOHE), for sponsoring this work under the Fundamental Research Grant Scheme (FRGS) COSSID/F00446.
References
Alonso, F., Esteban, C., Useche, S. A., & López de Cózar, E. (2016). Prevalence of physical and mental fatigue symptoms on Spanish drivers and its incidence on driving safety. Advances in psychology and neuroscience, 1(2), 10-18.
Anderson, S. W., Rizzo, M., Shi, Q., Uc, E. Y., & Dawson, J. D. (2005). Cognitive abilities related to driving performance in a simulator and crashing on the road. Proceedings of the Third International Driving Symposium on Human Factors in Driver Assessment, Training and Vehicle Design,
Barr Jr, G. C., Kane, K. E., Barraco, R. D., Rayburg, T., Demers, L., Kraus, C. K., Greenberg, M.
R., Rupp, V. A., Hamilton, K. M., & Kane, B. G. (2015). Gender differences in perceptions and self-reported driving behaviors among teenagers. The Journal of emergency medicine, 48(3), 366-370. e363.
Barron, E., Riby, L. M., Greer, J., & Smallwood, J. (2011). Absorbed in thought: The effect of mind wandering on the processing of relevant and irrelevant events. Psychological science, 22(5), 596-601.
Bella, F. (2013). Driver perception of roadside configurations on two-lane rural roads: Effects on speed and lateral placement. Accident Analysis & Prevention, 50, 251-262.
Bella, F. (2015). Coordination of horizontal and sag vertical curves on two-lane rural roads:
Driving simulator study. IATSS research, 39(1), 51-57.
Bogdan, S. R., Măirean, C., & Havârneanu, C.-E. (2016). A meta-analysis of the association between anger and aggressive driving. Transportation research part F: traffic psychology and behaviour, 42, 350-364.
Burdett, B. R., Charlton, S. G., & Starkey, N. J. (2016). Not all minds wander equally: The influence of traits, states and road environment factors on self-reported mind wandering during everyday driving. Accident Analysis & Prevention, 95, 1-7.
Chung, Y.-S. (2015). Seemingly irrational driving behavior model: The effect of habit strength and anticipated affective reactions. Accident Analysis & Prevention, 82, 79-89.
Davidović, J., Pešić, D., & Antić, B. (2018). Professional drivers’ fatigue as a problem of the modern era. Transportation research part F: traffic psychology and behaviour, 55, 199-209.
Demirdöğen Çetinoğlu, E., Görek Dilektaşlı, A., Demir, N. A., Özkaya, G., Acet, N. A., Durmuş, E., Ursavaş, A., Karadağ, M., & Ege, E. (2015). The relationship between driving simulation performance and obstructive sleep apnoea risk, daytime sleepiness, obesity and road traffic accident history of commercial drivers in Turkey. Sleep and Breathing, 19(3), 865-872.
Di Milia, L., Smolensky, M. H., Costa, G., Howarth, H. D., Ohayon, M. M., & Philip, P. (2011).
Demographic factors, fatigue, and driving accidents: An examination of the published literature. Accident Analysis & Prevention, 43(2), 516-532.
Fell, D. (2019). The road to fatigue: circumstances leading to fatigue accidents. In Fatigue and driving (pp. 97-105). Routledge.
Folkard, S., & Tucker, P. (2003). Shift work, safety and productivity. Occupational medicine, 53(2), 95-101.
Garza-Ulloa, J. (2018). Chapter 6 - Application of mathematical models in biomechatronics:
artificial intelligence and time-frequency analysis. In J. Garza-Ulloa (Ed.), Applied Biomechatronics using Mathematical Models (pp. 373-524). Academic Press.
https://doi.org/https://doi.org/10.1016/B978-0-12-812594-6.00006-8
Gish, J., Vrkljan, B., Grenier, A., & Van Miltenburg, B. (2017). Driving with advanced vehicle technology: A qualitative investigation of older drivers’ perceptions and motivations for use.
Accident Analysis & Prevention, 106, 498-504.
Goodall, N. J. (2014). Machine ethics and automated vehicles. In Road vehicle automation (pp.
93-102). Springer.
Jing, T., Xinghao, S., Lijuan, J., Qizong, S., Chang, P., & Chong, G. (2013). Experimental study on fatigue characteristics of bus drivers under continuous driving conditions [J]. Traffic information and safety, 31(03), 87-92.
Jovanis, P. P., Wu, K.-F., & Chen, C. (2012). Effects of hours of service and driving patterns on motor carrier crashes. Transportation research record, 2281(1), 119-127.
Kaba, A., & Ramaiah, C. K. (2020). Measuring knowledge acquisition and knowledge creation: a review of the literature.
Kalogirou, S. A. (2014). Chapter 11 - Designing and Modeling Solar Energy Systems. In S. A.
Kalogirou (Ed.), Solar Energy Engineering (Second Edition) (pp. 583-699). Academic Press.
https://doi.org/https://doi.org/10.1016/B978-0-12-397270-5.00011-X
Klauer, C., Dingus, T. A., Neale, V. L., Sudweeks, J. D., & Ramsey, D. J. (2006). The impact of driver inattention on near-crash/crash risk: An analysis using the 100-car naturalistic driving study data.
Körber, M., Gold, C., Lechner, D., & Bengler, K. (2016). The influence of age on the take-over of vehicle control in highly automated driving. Transportation research part F: traffic psychology and behaviour, 39, 19-32.
Koudelková, Z., & Strmiska, M. (2018). Introduction to the identification of brain waves based on their frequency. MATEC Web of Conferences,
Kulkarni, A. M., Nandi, A. V., & Nissimagoudar, P. (2019). Driver state analysis for ADAS using EEG signals. 2019 2nd International Conference on Signal Processing and Communication (ICSPC),
Lafont, S., Laumon, B., Helmer, C., Dartigues, J.-F., & Fabrigoule, C. (2008). Driving cessation and self-reported car crashes in older drivers: the impact of cognitive impairment and dementia in a population-based study. Journal of Geriatric Psychiatry and Neurology, 21(3), 171-182.
Li, K., Gong, Y., & Ren, Z. (2020). A fatigue driving detection algorithm based on facial multi- feature fusion. IEEE Access, 8, 101244-101259.
Li, X., Yan, X., & Wong, S. C. (2015). Effects of fog, driver experience and gender on driving behavior on S-curved road segments. Accident Analysis & Prevention, 77, 91-104.
Li, X., Yan, X., Wu, J., Radwan, E., & Zhang, Y. (2016). A rear-end collision risk assessment model based on drivers’ collision avoidance process under influences of cell phone use and
Liao, C.-Y., Chen, R.-C., & Tai, S.-K. (2018). Emotion stress detection using EEG signal and deep learning technologies. 2018 IEEE International Conference on Applied System Invention (ICASI),
Lindsay, G. W. (2020). Attention in Psychology, Neuroscience, and Machine Learning [Review].
Frontiers in Computational Neuroscience, 14. https://doi.org/10.3389/fncom.2020.00029 Liu, Y., Lan, Z., Cui, J., Sourina, O., & Müller-Wittig, W. (2020). Inter-subject transfer learning
for EEG-based mental fatigue recognition. Advanced Engineering Informatics, 46, 101157.
May, J. F., & Baldwin, C. L. (2009). Driver fatigue: The importance of identifying causal factors of fatigue when considering detection and countermeasure technologies. Transportation research part F: traffic psychology and behaviour, 12(3), 218-224.
MIROS. (2021). Road Accidents and Fatalities in Malaysia
https://www.mot.gov.my/en/land/safety/road-accident-and-facilities
Mohamed, N., Mohd-Yusoff, M.-F., Othman, I., Zulkipli, Z.-H., Osman, M. R., & Voon, W. S.
(2012). Fatigue-related crashes involving express buses in Malaysia: Will the proposed policy of banning the early-hour operation reduce fatigue-related crashes and benefit overall road safety? Accident Analysis & Prevention, 45, 45-49.
Mullette-Gillman, O. D. A., Leong, R. L., & Kurnianingsih, Y. A. (2015). Cognitive fatigue destabilizes economic decision making preferences and strategies. PloS one, 10(7), e0132022.
Pediatrics, A. A. o. (2014). School start times for adolescents. Pediatrics, 134(3), 642-649.
Peterson, B. K. (2007). Chapter 22 - Vital Signs. In M. H. Cameron & L. G. Monroe (Eds.),
Physical Rehabilitation (pp. 598-624). W.B. Saunders.
https://doi.org/https://doi.org/10.1016/B978-072160361-2.50025-9
Peylo, C., Hilla, Y., & Sauseng, P. (2021). Cause or consequence? Alpha oscillations in visuospatial attention. Trends in Neurosciences, 44(9), 705-713.
https://doi.org/https://doi.org/10.1016/j.tins.2021.05.004
Plomin, R. (1999). Genetics and general cognitive ability. Nature, 402(6761), C25-C29.
Prezenski, S., Brechmann, A., Wolff, S., & Russwinkel, N. (2017). A cognitive modeling approach to strategy formation in dynamic decision making. Frontiers in psychology, 8, 1335.
Ramisetty-Mikler, S., & Almakadma, A. (2016). Attitudes and behaviors towards risky driving among adolescents in Saudi Arabia. International Journal of Pediatrics and Adolescent Medicine, 3(2), 55-63.
Rau, P. S. (2005). Drowsy Driver Detection and Warning System for Commercial Vehicle Drivers:
Field Operational Test Design, Data Analyses, and Progress.
Roy, E. (2013). Cognitive Impairment. In M. D. Gellman & J. R. Turner (Eds.), Encyclopedia of Behavioral Medicine (pp. 449-451). Springer New York. https://doi.org/10.1007/978-1-4419- 1005-9_1118
Sadollah, A. (2018). Introductory chapter: which membership function is appropriate in fuzzy system? IntechOpen.
Samuel, S., Yahoodik, S., Yamani, Y., Valluru, K., & Fisher, D. L. (2020). Ethical decision making behind the wheel – A driving simulator study. Transportation Research Interdisciplinary Perspectives, 5, 100147. https://doi.org/https://doi.org/10.1016/j.trip.2020.100147
Schweizer, T. A., Kan, K., Hung, Y., Tam, F., Naglie, G., & Graham, S. J. (2013). Brain activity during driving with distraction: an immersive fMRI study. Front Hum Neurosci, 7, 53.
https://doi.org/10.3389/fnhum.2013.00053
Sentoff, K. M., Aultman-Hall, L., & Holmén, B. A. (2015). Implications of driving style and road grade for accurate vehicle activity data and emissions estimates. Transportation research part D: transport and environment, 35, 175-188.
Shackel, S. C., & Parkin, J. (2014). Influence of road markings, lane widths and driver behaviour on proximity and speed of vehicles overtaking cyclists. Accident Analysis & Prevention, 73, 100-108.
Shahid, A., Chung, S. A., Maresky, L., Danish, A., Bingeliene, A., Shen, J., & Shapiro, C. M.
(2016). The Toronto Hospital Alertness Test scale: relationship to daytime sleepiness, fatigue, and symptoms of depression and anxiety. Nature and science of sleep, 8, 41-45.
https://doi.org/10.2147/NSS.S91928
Shen, S., & Neyens, D. M. (2017). Assessing drivers' response during automated driver support system failures with non-driving tasks. Journal of safety research, 61, 149-155.
Simion, M., Socaciu, L., & Unguresan, P. (2016). Factors which influence the thermal comfort inside of vehicles. Energy Procedia, 85, 472-480.
Smallwood, J., Riby, L., Heim, D., & Davies, J. B. (2006). Encoding during the attentional lapse:
Accuracy of encoding during the semantic sustained attention to response task. Consciousness and cognition, 15(1), 218-231. https://doi.org/https://doi.org/10.1016/j.concog.2005.03.003 Strayer, D. L., Drews, F. A., & Johnston, W. A. (2003). Cell phone-induced failures of visual
attention during simulated driving. J Exp Psychol Appl, 9(1), 23-32.
https://doi.org/10.1037/1076-898x.9.1.23
Suh, J., Yi, K., Jung, J., Lee, K., Chong, H., & Ko, B. (2016). Design and evaluation of a model predictive vehicle control algorithm for automated driving using a vehicle traffic simulator.
Control Engineering Practice, 51, 92-107.
Tefft, B. C. (2018). Acute sleep deprivation and culpable motor vehicle crash involvement. Sleep, 41(10), zsy144.
Thakkar, H., Shah, V., Yagnik, H., & Shah, M. (2021). Comparative anatomization of data mining and fuzzy logic techniques used in diabetes prognosis. Clinical eHealth, 4, 12-23.
https://doi.org/https://doi.org/10.1016/j.ceh.2020.11.001
Ulrich, T. A., Barton, B. K., & Lew, R. (2014). Detection and localization of approaching vehicles in the presence of competing vehicle noise. Transportation research part F: traffic psychology and behaviour, 26, 151-159.
VandenBos, G. R. (2007). APA dictionary of psychology. American Psychological Association.
Walker, H. E., & Trick, L. M. (2018). Mind-wandering while driving: The impact of fatigue, task length, and sustained attention abilities. Transportation research part F: traffic psychology and behaviour, 59, 81-97.
Wang, K. (2001). Computational Intelligence in Agile Manufacturing Engineering. In A.
Gunasekaran (Ed.), Agile Manufacturing: The 21st Century Competitive Strategy (pp. 297- 315). Elsevier Science Ltd. https://doi.org/https://doi.org/10.1016/B978-008043567-1/50016- 4
Wang, X., Liu, Y., Wang, J., & Zhang, J. (2019). Study on influencing factors selection of driver’s propensity. Transportation research part D: transport and environment, 66, 35-48.
WHO. (2018). Global status report on road safety 2018: summary.
Yan-li, M., & Yu-long, P. (2009). Influences of continuous driving time on test indicators of driving characteristics. China journal of highway and transport, 22(1), 84.
Zhao, J., & Bose, B. K. (2002). Evaluation of membership functions for fuzzy logic controlled induction motor drive. IEEE 2002 28th Annual Conference of the Industrial Electronics Society. IECON 02,