Psychophysiological responses to short-term cooling during a simulated monotonous driving task

Applied Ergonomics 62 (2017) 9e18

Contents lists available at ScienceDirect

Applied Ergonomics journal homepage: www.elsevier.com/locate/apergo

Psychophysiological responses to short-term cooling during a simulated monotonous driving task Elisabeth Schmidt a, b, *, Ralf Decke a, Ralph Rasshofer a, Angelika C. Bullinger b a b

BMW AG, Knorrstrasse 147, 80788 Munich, Germany Technical University Chemnitz, 09107 Chemnitz, Germany

a r t i c l e i n f o

a b s t r a c t

Article history: Received 13 May 2016 Received in revised form 17 January 2017 Accepted 20 January 2017

For drivers on monotonous routes, cognitive fatigue causes discomfort and poses an important risk for traffic safety. Countermeasures against this type of fatigue are required and thermal stimulation is one intervention method. Surprisingly, there are hardly studies available to measure the effect of cooling while driving. Hence, to better understand the effect of short-term cooling on the perceived sleepiness of car drivers, a driving simulator study (n ¼ 34) was conducted in which physiological and vehicular data during cooling and control conditions were compared. The evaluation of the study showed that cooling applied during a monotonous drive increased the alertness of the car driver. The sleepiness rankings were significantly lower for the cooling condition. Furthermore, the significant pupillary and electrodermal responses were physiological indicators for increased sympathetic activation. In addition, during cooling a better driving performance was observed. In conclusion, the study shows generally that cooling has a positive short-term effect on drivers’ wakefulness; in detail, a cooling period of 3 min delivers best results. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Fatigue countermeasures Cooling Driver sleepiness Thermal stimulation

1. Introduction During long monotonous routes, drivers may find themselves in a state of low arousal due to lack of sensory stimulation. Low arousal describes an affective state in which individuals feel little activation, for example when being sleepy, bored or droopy (Russell, 1980). In the circumplex model of affect of Russell (1980) as well as in the mood theory of Watson and Tellegen (1985) emotions are viewed as combinations of the two basic orthogonal constructs valence and arousal. Later studies (Bradley and Lang, 1994; Thayer, 1989) report on semantic differential methods to subjectively measure valence and arousal states. These techniques become important when researching external factors e such as monotonous driving e on perceived arousal. Besides causing a deactivated state of the driver, the repetitive and predictable nature of monotonous drives turns the driving task also in an automatic process - a mode, also referred to as driving without awareness (DWA, Brown, 1994; Charlton and Starkey, 2011). The studies of Briest et al. (2006) and Karrer et al. (2005) proved that DWA and sleepiness are closely related. They found for example that DWA

* Corresponding author. BMW AG, Knorrstrasse 147, 80788 Munich, Germany. E-mail address: [email protected] (E. Schmidt). http://dx.doi.org/10.1016/j.apergo.2017.01.017 0003-6870/© 2017 Elsevier Ltd. All rights reserved.

often precedes microsleep and that both DWA and sleepiness are indicated by changes in blink behavior. Another study supporting that monotonous routes foster the development of fatigue was done by Thiffault and Bergeron (2003), who found that monotonously simulated highway routes cause a vigilance decrement. That task underload on monotonous routes causes cognitive sleepiness is also reflected in several fatigue models (Lal and Craig, 2001; May and Baldwin, 2009; Van Veen et al., 2014) which distinguish different types of driver fatigue and their causes. As a result of cognitive sleepiness, the driver grows uncomfortable during the trip. This not only inhibits pleasurable user experience but can also cause traffic risks due to lower levels of alertness of the driver. Tejero Gimeno et al. (2006) and Van Veen et al. (2014) list thermal stimulation as a possible countermeasure for cognitive sleepiness due to task underload in their reviews. The underlying principle for this is that outside of comfortable temperatures, the arousal of a human increases steadily (Parsons, 2003; Wyon, 1973). This means, that arousal is a function of ambient temperature, whereas the minimum arousal is at a comfortable temperature. It has also been shown that performance is a function of arousal (Hygge, 1992; Parsons, 2003), whereas there is an optimum level of arousal in which performance peaks. Higher or lower levels of arousal worsen the performance. The model of

10

E. Schmidt et al. / Applied Ergonomics 62 (2017) 9e18

Hygge (1992) specifies the relationship between arousal and performance for easy and difficult tasks, in a way that the performance in easier tasks is generally higher than in difficult tasks, and that the optimum arousal level for easy tasks is higher than the required arousal for difficult tasks. From these theories, we draw that arousal and performance are influenced by the thermal environment, and that easy tasks require higher levels of arousal for best performance. In practice, some studies can be identified investigating thermal stimulation of drivers. Surveys about countermeasures against fatigue report that “opening a window” or “turning on the AC” (air conditioning) are common practice among drivers (Anund et al., 2008; Gershon et al., 2011; Oron-Gilad and Shinar, 2000; Royal, 2003). Besides interview studies, a few controlled laboratory studies investigated the effect of cooling on the perceived sleepiness of car drivers with different outcomes. Wyon et al. (1996) compared the task performance in a driving study at a temperature of 21  C and 27  C and found a better performance at the lower temperature. In another example Van Veen (2016) investigated the effect of intermittent local hand cooling by a temperature difference of 5  C to the ambient temperature on drowsiness and heart rate (HR) during simulated driving. While a significant increase of HR after 3 min of cooling was reported, indicating an activation of the sympathetic nervous system (SNS), the effect of short-term local cooling of the hands on perceived cognitive sleepiness was not significant. Another laboratory study was conducted by €m et al. (1999) in which the effect of short-term room Landstro temperature drops from 28  C to 18  C on perceived sleep-related sleepiness and electroencephalogram (EEG) data was measured. Here they found a significant decrease in perceived sleepiness with cooling and EEG data proved an enhanced wakefulness. The study did not, however, include a driving task and focused on sleep€m et al. (2002) related fatigue. In a later field study, Landstro found that a temperature control system which cooled down the truck cabin temperature repeatedly by 8  C to 10  C with starting temperatures between 25  C and 30  C, increased the alertness in professional drivers. Since the truck drivers used the temperature control system during the night and mainly after long hours of driving, the study addressed sleep-related fatigue. Reyner and Horne (1998) found in a car simulator study, that 10  C cold air neither had a significant effect on perceived sleepiness nor on EEG data after two and a half hours of driving. Their study also focused on drowsiness caused by sleep deprivation. Schwarz et al. (2012) worked with subjects which were not sleep-deprived and investigated the effect of intermittent 10 min-periods of opening the window for 2 cm at speeds of 120 km/h in a real-driving study. Subjective sleepiness and blink duration were not affected by this countermeasure. Given these studies, it is surprising to see that the effect of thermal stimulation to relieve the strain of monotonous driving has hardly been explored. In order to gain knowledge on the effect of short-term cooling on cognitive fatigue in a controlled vehicle setting, a simulator study was performed. The aim of the study was to investigate a decrease in sleepiness caused by a task underload during a monotonous drive by means of cooling as well as exploring the effect of cooling on the emotions of the drivers using the semantic differential. 2. Experimental design 2.1. Apparatus It was possible to use a street-legal car, changed only for experimental reasons. As the studies of Philip et al. (2005) and Hallvig et al. (2013) have shown, perceived sleepiness and

physiological sleepiness is higher in simulated driving than in real driving for both sleep-deprived and non-sleep-deprived subjects due to lower levels of visuomotor stimulation. Since this study required to induce sleepiness due to monotony in non-sleepdeprived subjects, we decided to run a simulator study instead of a real-driving study which allowed also for reproducible traffic scenarios. The cold air was provided by an external AC unit which was attached to the fresh air intake of the vehicle. Via remote control of the circulation flap and the interior fans of the vehicle, the cold air could be let in the cabin as required. Air at a temperature of 17  C was used for the cooling condition because studies of Van Veen (2016) have shown a physiological effect at similar temperatures. In the control condition the cabin stayed at a thermo-neutral temperature (23  C). After flipping the circulation flap, it took about 60 s before the measured temperature at the air inlets dropped from 23  C to 17  C, because the air channels of the car were warmed up to room temperature during the control period. 2.2. Participants Fifty BMW Group employees recruited via a mailing list voluntarily participated in the study in February 2016. Of those, n ¼ 5 participants were excluded from analysis by listwise deletion due to insufficient sleepiness level and n ¼ 9 were excluded because of technical errors. Another n ¼ 2 participants were not able to continue the study after the familiarization drive due to simulator sickness, resulting in a sample size of 34 participants for analysis. Study participants were 24 male and 10 female healthy subjects aged between 21 and 59 (M ¼ 31.8, SD ¼ 11.2). The participants kept their regular sleeping schedule, but were instructed to avoid tobacco or caffeinated beverages on the day of the study. The participants were examined at different times of the day. Seven participants took part from 8am to 10am, eight from 10am to 12pm, seven from 12pm to 2pm, eight from 2pm to 4pm and four from 4pm to 6pm. A Kruskal-Wallis test of the initial sleepiness ratings between the subjects of the five test times supports that the test time did not affect the sleepiness level (H (4) ¼ 1.93, p ¼ 0.75). The subjects were dressed in underwear, socks, shoes, pants and Tshirts in all thermal conditions of the experiment. The outside temperature during the period of the experiments ranged between 2  C and 12  C and the relative humidity ranged between 63% and 81%. Through the initial questionnaires, attachment and calibration of all sensors, the baseline recording of the electrocardiogram (ECG) and the familiarization drive, subjects could acclimate for 30 min to the thermal conditions of the simulator (23  C and 35% relative humidity) before the start of the first test drive. The subjects did not know that cooling would be applied during the experiment. 2.3. Study design The study employed a one-factor within-subject-design, in which all participants drove two monotonous highway routeseone with short-term cooling (COOL) at the end and one without (CONT) (Fig. 1). The order of the conditions was counterbalanced. The study began with one 5-min familiarization drive on a highway in which subjects could get used to the simulation environment. After the familiarization, the two test drives of 26 min followed. The duration of these drives was limited by the fact that total length of the experiment was limited to 90 min because the study took place during the participants’ working hours. Both test drives were highway-routes with very little traffic in order to cause a monotonous task. Furthermore, subjects were instructed to drive no faster than 120 km/h.

E. Schmidt et al. / Applied Ergonomics 62 (2017) 9e18

11

Fig. 1. Test drives with marked times of verbal assessment of sleepiness, cooling period and questionnaires.

In the COOL condition, cold air was blown with 100% fan intensity in the cabin starting after 20 min for 6 min (Fig. 1). A duration of 6 min was chosen because of the recommendation of Van Veen (2016) that cooling should last at least 3 min and the €m et al. (1999) that longer durations go at the warning of Landstro expense of thermal comfort. Severe cooling at 5  C for 30 min has been proven to deteriorate driving performance by Daanen et al. (2003). Only the middle air inlets (not the feet or defrost air inlets) of the car were active and the air flow was directed towards the face of the driver in order to target as much exposed skin as possible for maximum cooling sensation. Since the air flow gets distributed in the cabin, also neck and upper chest were partly hit by the air flow. The alternative option of cooling the legroom would have also been worthy of examination, however we viewed the practical impact of the study results of facial cooling higher. This is because the face is usually exposed, whereas the legs are often covered by clothes and therefore the transfer of study results would have been limited to a certain kind of leg clothing. We chose to vary both temperature and fan intensity as the COOL treatment to further increase the cooling effect due to the wind chill effect on the exposed skin and to prevent the cold air from mixing with surrounding thermo-neutral air before reaching the driver. The disadvantage in doing so, is, that all measured effects cannot be solely attributed to the air temperature difference, but to a change in thermal environment consisting of air temperature change and air movement change. Furthermore, the fanning caused a constant noise for the 6 min of cooling on top of the driving noise. At all other times the fans blew at an intensity of 20% with thermo-neutral air. After the first monotonous test drive the participants performed a wakening dexterity mastication taskenamely shelling and eating sunflower seeds as in the study of Gershon et al. (2009)ein an attempt to activate the participants and to create a similar initial condition before the second monotonous drive.

2.4. Questionnaires After each drive, the subjects completed several questionnaires including the Karolinska Sleepiness Scale (KSS) (Åkerstedt and Gillberg, 1990; translated into German by Niederl (2007)) for measuring perceived sleepiness. Furthermore, the activationdeactivation adjective checklist (ADACL) following Thayer (1989) (German version by Imhof (1998)) was asked to measure the drivers’ emotions in terms of their perceived positive or negative activation, and the thermal sensation and preference scale described in EN ISO 14505e3, 2006. In the comment fields of the questionnaire, subjects could also

describe their impressions of the cooling period and its effectiveness against cognitive sleepiness during the monotonous drive. The KSS was also asked before the start of the experiment as well as verbally via a microphone-speaker system after the 6th and 16th minute of the drives without stopping the drive. It has been proven by Schmidt et al. (2011) that sleepiness can be verbally assessed every 5 min without continued activation of the driver through this type of interaction. This is because the verbal assessment only has a short-term effect on the subjects’ vigilance level and the physiological vigilance measures are back to precommunication levels after a maximum of 2 min. In case the KSS-rating stayed lower than 5 (¼ neither alert nor sleepy) for a subject throughout the experiment the datasets were excluded from the statistical analysis because the study requirement of cognitive driver fatigue was not met. 2.5. Psychophysiological measurements A 3-channel-ECG, breathing frequency (BF), skin conductance level (SCL), skin temperature on the upper arm, oxygen saturation, plethysmogram and pulse were measured with medical sensors (g.tech, Austria) with a sampling frequency of 512 Hz. The ECG electrodes were placed under the right and left clavicle and on the lower left abdomen within the rib cage frame. A baseline signal of all physiological measures was recorded for 10 min in a resting position of the subjects in the driver's seat without the simulation projection. Head position and emotions were extracted at 100 Hz from a frontal camera (IDS Imaging Development Systems, Germany) facing the subjects from the dashboard with a facial recognition software (iMotions, Denmark) which is based on the FACS (facial action coding system) of Ekman and Rosenberg (1997). Gaze coordinates and pupil diameters of each eye were recorded at 60 Hz using a Tobii Pro X2-60 eye tracker (Tobii, Sweden). 2.6. Data analysis The data was processed and analyzed using Matlab 2013b. From the ECG recording, the HR and heart rate variability (HRV) measures were obtained. Both time and frequency domain HRV were extracted because studies (Egelund, 1982; Kaida et al., 2007; Patel et al., 2011) have proven those as indicators of sleepiness. The time HRV measures SDNN (standard deviation of normal to normal intervals) and RMSSD (root mean square of successive differences) were calculated for the period of 3 min. Furthermore, a spectral analyses of 3-min sequences of interbeat intervals were performed and the frequency domain HRV measures LF (low frequency component, 0.04e0.15 Hz) and HF (high frequency component, 0.15e0.4 Hz) were obtained. The raw SCL was processed using the convex optimization

12

E. Schmidt et al. / Applied Ergonomics 62 (2017) 9e18

approach of Greco et al. (2014). Pupil diameters of the left and right eye were averaged for each subject. The lane keeping quality and jerkiness of steering were extracted from the vehicle data. The lane keeping quality is a percentage value indicating how well subjects drive in the middle of the lane. The value ranges linearly from 100%, when driving perfectly in the middle of one highway lane, to 0%, when driving in the middle between two lanes. For the statistical analysis, all continuously recorded data and evaluated signals were averaged for each minute and for each subject. A significance level of 0.05 was used for all statistical tests, unless stated otherwise. Questionnaire results, as well as physiological and driving data were evaluated using two-sided Wilcoxon tests because the data was not normally distributed.

Table 1 Means (M) and standard deviations (SD) of the subjective positive and negative activation perceived after each drive and Wilcoxon test results for the conditions CONT and COOL. (enot significantly different, * significantly different p < 0.05, ** highly significantly different p < 0.001). Psychological construct Positive activation Positive deactivation Negative deactivation Negative activation

M (SD) (based on 1e4 scales) Energy CONT Energy COOL Calmness CONT Calmness COOL Sleepiness CONT Sleepiness COOL Tension CONT Tension COOL

1.51 1.95 3.14 3.06 3.25 2.77 1.42 1.49

(0.57) (0.78) (0.41) (0.61) (0.59) (0.76) (0.41) (0.56)

** e ** e

3. Results 3.1. Subjective sleepiness and activation The results of the KSS are shown in Fig. 2. There are no statistically significant differences between the sleepiness levels after minute 6 and 16 of the drives COOL and CONT. The sleepiness increase from minute 6 to 16 is significant in both conditions (Z ¼ 4.46, p < 0.001 for COOL and Z ¼ 4.92, p < 0.001 for CONT), and so is the increase from minute 16 to 26 in the CONT drive (Z ¼ 3.90, p < 0.001). As hypothesized, the average sleepiness level after 26 min is significantly different between the two conditions (Z ¼ 3.64, p < 0.001). In the COOL drive, the sleepiness level after 26 min is not different (Z ¼ 0.88, p ¼ 0.38) from the level after 16 min of driving. The results of the ADACL showed that the construct ‘energy’ in the COOL drive increased compared to the CONT drive (Z ¼ 3.35, p < 0.001) (Table 1). The construct ‘sleepiness’ decreased significantly through the short-term cooling (Z ¼ 3.40, p < 0.001). There were no significant differences in ‘tension’ (Z ¼ 1.05, p ¼ 0.30) and ‘calmness’ (Z ¼ 0.66, p ¼ 0.51). Table 2 shows the answers of the questionnaires regarding the impressions on the cooling. A percentage of 59% of the participants liked the cooling and the comments of these people often included that the cooling was perceived as refreshing and that it had a wakening effect. The comments of the people disliking the cooling revealed that the cooling felt too cold in their faces. This aligns with an average cool thermal sensation of 1.5 (1 meaning ‘slightly cool’, 2 meaning ‘cool’ according to EN ISO 14505e3, 2006)

compared to the neutral thermal sensation of 0.2 in the control condition. The majority of people (88%) thought that cooling reduced their sleepiness. The comments here included hints that the wakening effect of the cooling is only short-term and subjects indicated that they would have felt tired again if the drive had continued for a longer duration because they got used to the stimulus. The last question, which was asked after subjects completed both drives, revealed that 78% of the drivers would prefer a monotonous car drive with short-term cooling.

3.2. Physiological and vehicular measures Fig. 3 shows the graphs of the HR (averaged over all subjects) for both drives. HR decreased with cooling onset and stayed significantly lower than in the comparison drive in the first 3 min of cooling. The vertical lines at minute 6 and 16 in the graph indicate where sleepiness was verbally assessed. The means and standard deviations of all signals and the statistical results for the last 7 min of the drives are listed in Table 3. The two time domain HRV measures, SDNN and RMSSD, are increasing over the course of both drives which is an indicator for increasing sleepiness (Fig. 4). SDNN and RMSSD correlated fairly well with the KSS score (rs (32) ¼ 0.24, p < 0.001, rs (32) ¼ 0.31, p < 0.001). There are no significant differences between the two conditions at any point of the drive. The cooling did not affect BF (Table 3). The SCL increased significantly with onset of cooling and stayed higher for the first

Fig. 2. KSS ratings (mean and standard error) over the course of the drives CONT and COOL. Marks indicate Wilcoxon test results for points of time in which CONT and COOL are significantly different.

E. Schmidt et al. / Applied Ergonomics 62 (2017) 9e18

13

Table 2 Questionnaire answers to the impressions on the cooling.

Did you like the cooling? Do you believe that the cooling reduced sleepiness? Would you prefer a monotonous drive with or without short-term cooling?

Yes [%]

No [%]

59 88 With cooling [%] 78

41 12 Without cooling [%] 22

Fig. 3. Mean HR over the course of the drives CONT and COOL. Marks indicate Wilcoxon test results for points of time in which CONT and COOL are significantly different.

Table 3 Means and standard deviations of physiological and driving data during the last 7 min of the two drives and Wilcoxon test results of the minute-wise comparison of the means for the drives CONT and COOL. (enot significantly different, * significantly different p < 0.05, ** highly significantly different p < 0.001). Parameter (CONT and COOL)

1 min before cooling

1st min of cooling

HR [bpm] CONT HR [bpm] COOL SDNN [ms] CONT SDNN [ms] COOL RMSSD [ms] CONT RMSSD [ms] COOL LF/HF [-] CONT LF/HF [-] COOL LF [ms2] CONT LF [ms2] COOL HF [ms2] CONT HF [ms2] COOL BF [1/min] CONT BF [1/min] COOL SCL [mS] CONT SCL [mS] COOL SC phasic [-] CONT SC phasic [-] COOL SC tonic [-] CONT SC tonic [-] COOL Pupil diameter [mm] CONT Pupil diameter [mm] COOL Eyes closed [-] CONT Eyes closed [-] COOL Gaze variability [mm2 $103] CONT Gaze variability [mm2$103] COOL Lane keeping quality [%] CONT Lane keeping quality [%] COOL Steer jerkiness [ /s4] CONT Steer jerkiness [ /s4] COOL

68.6 (10.5) 69.5 (10.3) 58.1 (27.0) 54.0 (24.4) 45.8 (33.8) 44.1 (32.1) 1.348 (0.004) 1.347 (0.002) 762 (183) 745 (184) 565 (136) 553 (136) 17.3 (3.6) 17.4 (2.3) 1.28 (1.07) 1.39 (1.12) 3.11 (3.55) 5.06 (6.66) 3.46 (3.38) 5.30 (6.21) 4.47 (0.86) 4.54 (0.88) 0.63 (0.91) 0.64 (0.91) 2.08 (1.57) 1.91 (1.34) 75.5 (9.3) 77.1 (7.6) 0.095 (0.061) 0.088 (0.048)

e 68.5 (10.5) 66.6 (9.6) e 56.3 (25.5) 57.1 (25.3) e 46.0 (35.4) 44.2 (33.4) e 1.347 (0.001) 1.347 (0.001) e 763 (204) 751 (198) e 566 (152) 557 (147) e 17.6 (2.9) 17.2 (2.6) e 1.24 (1.00) 1.53 (1.32) e 2.29 (2.94) 7.76 (6.94) e 2.76 (2.75) 7.52 (6.43) e 4.44 (0.85) 4.86 (0.82) e 0.70 (0.86) 0.50 (0.96) e 1.87 (0.92) 1.98 (1.06) e 76.5 (7.5) 79.5 (7.5) e 0.084 (0.089) 0.054 (0.033)

2nd min of cooling * e e e e e e ** ** ** ** * e * *

3rd min of cooling

69.0 (10.1) 67.4 (10.1) 59.1 (31.7) 56.9 (25.4) 47.2 (38.2) 46.4 (35.4) 1.348 (0.001) 1.348 (0.001) 767 (228) 759 (207) 569 (170) 563 (153) 17.3 (2.8) 17.2 (3.3) 1.25 (1.00) 1.38 (1.16) 3.81 (4.63) 4.35 (5.09) 4.13 (4.57) 4.68 (4.74) 4.40 (0.80) 4.71 (0.81) 0.69 (0.85) 0.51 (0.89) 2.31 (1.95) 1.99 (1.09) 76.9 (8.5) 79.6 (9.1) 0.052 (0.038) 0.051 (0.067)

2 min of cooling compared to the CONT drive (Fig. 5). The phasic and tonic components of skin conductance changed immediately with cooling and the effect lasted for the first minute of cooling

*

69.1 (11.1) 67.8 (9.5) e 59.0 (31.0) 56.7 (25.6) e 47.2 (38.2) 48.0 (38.1) e 1.348 (0.001) 1.348 (0.001) e 761 (222) 770 (218) e 564 (165) 571 (162) e 17.4 (2.6) 17.1 (2.8) * 1.24 (0.98) 1.33 (1.13) e 3.85 (5.08) 4.45 (5.51) e 4.17 (4.80) 4.96 (5.21) ** 4.49 (0.76) 4.66 (0.85) * 0.66 (0.89) 0.54 (0.87) e 1.82 (0.92) 2.03 (0.84) e 75.5 (8.2) 78.4 (7.0) e 0.098 (0.074) 0.086 (0.061)

4th min of cooling * e e e e e e e e e * * e * e

69.5 (10.7) 68.7 (10.1) 62.6 (34.9) 58.9 (26.4) 48.2 (39.6) 48.6 (38.6) 1.348 (0.001) 1.348 (0.001) 773 (227) 770 (230) 573 (169) 571 (171) 17.2 (2.8) 16.9 (2.5) 1.27 (1.04) 1.30 (1.06) 4.57 (5.29) 3.52 (4.23) 4.82 (4.96) 4.12 (4.02) 4.49 (0.83) 4.70 (0.80) 0.61 (0.87) 0.51 (0.86) 2.23 (1.09) 1.91 (0.83) 76.3 (7.9) 76.1 (8.7) 0.114 (0.101) 0.095 (0.055)

5th min of cooling

6th min of cooling

69.2 (10.6) 68.5 (9.6) e 61.5 (33.9) 61.0 (27.4) e 49.3 (41.1) 48.4 (36.1) e 1.348 (0.001) 1.348 (0.002) e 758 (240) 767 (215) e 562 (178) 569 (160) e 17.4 (3.0) 16.5 (2.8) e 1.23 (0.96) 1.33 (1.14) e 5.31 (7.25) 3.52 (3.69) e 5.62 (7.03) 4.08 (3.39) ** 4.36 (0.85) 4.64 (0.78) e 0.66 (0.86) 0.53 (0.89) e 2.32 (1.47) 2.20 (1.11) e 74.2 (8.7) 76.7 (8.8) e 0.096 (0.081) 0.092 (0.124)

68.9 (10.4) 68.3 (9.7) e 62.5 (36.2) 60.5 (26.7) e 51.0 (44.6) 47.8 (35.3) e 1.348 (0.001) 1.348 (0.001) e 763 (259) 764 (212) e 566 (192) 567 (157) e 17.5 (3.1) 17.5 (2.8) e 1.25 (0.99) 1.26 (1.05) e 3.87 (3.97) 3.28 (4.12) e 4.10 (3.96) 4.05 (3.95) ** 4.35 (0.82) 4.52 (0.79) e 0.65 (0.86) 0.48 (0.97) e 2.26 (1.16) 2.15 (0.96) * 73.3 (11.4) 75.7 (9.6) * 0.130 (0.150) 0.104 (0.063)

e

e

e e e e e e e e e e * e e e e

only (Table 3). The pupil diameter is significantly increased during the entire length of the cooling period. Fig. 6 shows that the increase in

14

E. Schmidt et al. / Applied Ergonomics 62 (2017) 9e18

Fig. 4. Mean SDNN and RMSSD over the course of the drives CONT and COOL. Marks indicate Wilcoxon test results for points of time in which CONT and COOL are significantly different.

Fig. 5. Mean skin conductance over the course of the drives CONT and COOL. Marks indicate Wilcoxon test results for points of time in which CONT and COOL are significantly different.

Fig. 6. Mean pupil diameter over the course of the drives CONT and COOL. Marks indicate Wilcoxon test results for points of time in which CONT and COOL are significantly different.

E. Schmidt et al. / Applied Ergonomics 62 (2017) 9e18

diameter is the largest at the beginning of the cooling period. With the facial expression software, based on the FACS, evidence values for diverse facial actions could be extracted. In the FACS, the action unit no. 43 indicates closed eyes, which was of interest for objective sleepiness evaluation. Evidence values provided by the software are on a logarithmic scale. Cooling had an immediate effect on the evidence of eye closures (Fig. 7). The evidence of eye closures in the COOL condition was significantly lower for the first 3 min of cooling than in the CONT condition. Gaze variability was not affected by the cooling (Table 3). The average lane keeping quality is shown in Fig. 8. The drops in the 3rd, 13th and 18th minute are due to an overtaking maneuver and two slight curves of the highway. The quality was significantly increased with onset of cooling and stayed higher throughout the cooling period, except in the 4th cooling minute. Similarly, the steering movements (listed in Table 3) were less jerky during the onset of cooling which indicates higher alertness. Other vehicle parameters such as speed, acceleration, longitudinal jerkiness and gas pedal position were not correlated to the KSS ratings in our study and were also not affected by the cooling. 4. Discussion The goal of this study is to investigate the psychophysiological effect of short-term cooling when the driver is sleepy due to monotony. There are studies using similar in-car treatments € m et al., 1999, 2002; Reyner and Horne, 1998), but only a (Landstro few studies investigated the effect of cooling on non-sleepdeprived subjects (Schwarz et al., 2012; Van Veen, 2016). In those studies, subjectively rated sleepiness was not significantly decreased. Therefore, we conducted a study with a larger temperature difference, fan intensity and a different fanning direction. One requirement on the study was the induction of sleepiness in the participants by means of monotonous driving. As the analysis of the KSS ratings (Fig. 2) showed, subjects perceived increasing sleepiness through the drive. This is supported by the increase of the HRV measures SDNN and RMSSD (Fig. 4). The verbal assessment of sleepiness activated the drivers physiologically because HR (Fig. 3), SCL (Fig. 5) and pupil diameter (Fig. 6) were increased after minutes 6 and 16 when subjects were required to respond to the KSS. The activation lasted only shortly, though, because the parameters return to the pre-assessment levels after 1 min, which aligns with the observations of Schmidt et al. (2011).

15

The evaluation of the questionnaires measuring sleepiness and the activation of the driver shows that the cooling decreased perceived sleepiness (Fig. 2) and subjects felt significantly more energetic (Table 1). The majority of participants believed that cooling reduced sleepiness. However, a not negligible percentage of the drivers did not like the cooling because it felt too cool (Table 2). These enhanced subjective perceptions on vigilance could also be confirmed in several physiological responses of the drivers. A SNS activation is indicated by the increase of SCL and pupil diameter (Figs. 5 and 6) during the cooling period. The two peaks in minute 6 and 16 in both graphs are due to the verbal KSS-ratings done at these times which caused these task-evolved pupillary responses (Beatty, 1982). Both the increase of pupil diameter and SCL during the cooling period are symptoms of autonomic activation due to sensory stimulation (Bradley et al., 2008; Goldwater, 1972). Bradley et al. (2008) conclude that the pupillary dilatation, the change in skin conductance, as well as HR deceleration reflect the emotional arousal linked to increased sympathetic activity. These three patterns of physiological reactions occurred in this study with onset of cooling. The increased arousal as indicated by the physiological response can furthermore constitute higher alertness of the subjects during the monotonous drive. Goldwater (1972) found that an increase in pupil diameter is associated with increased attention. Another reaction supporting the reduction in sleepiness through cooling, is the reduced evidence of eye closures during the first 3 min of cooling (Fig. 7). This result has special weight because the facial air flow could have also caused a sleepiness-unrelated increase in the number of blinks as for example reported by Schwarz et al. (2012) and as we could observe with two subjects wearing contact lenses. In our study, the SCL and its phasic and tonic components started to decrease to the baseline after 2 min at the latest (Table 3). This is most likely because subjects got used to the thermal stimulation after some time and therefore the stimulus has become monotonous. This assumption can partly be substantiated by the comments of the participants stating that one gets used to the stimulus after some time. The reduction of HR (Fig. 3) during cooling does not seem to align with the increased wakefulness of the driver at the first glance. In past studies (Schmidt et al., 2016), however, HR did not prove to be a good indicator for cognitive sleepiness during simulated driving. The reduction of HR is a physiological response of the short-term cooling, which has been observed with facial fanning in

Fig. 7. Mean evidence of eye closures over the course of the drives CONT and COOL. Marks indicate Wilcoxon test results for points of time in which CONT and COOL are significantly different.

16

E. Schmidt et al. / Applied Ergonomics 62 (2017) 9e18

Fig. 8. Lane keeping quality over the course of the drives course of the drives CONT and COOL. Marks indicate Wilcoxon test results for points of time in which CONT and COOL are significantly different.

non-vehicle settings (Collins et al., 1996; Hayward et al., 1976; LeBlanc et al., 1975, 1976, 1978). Researchers have suggested that the decrease of HR is caused by vagal reflex in which facial receptors initiate a stimulation of the trigeminal nerve (Collins et al., 1996; LeBlanc et al., 1976). The studies of LeBlanc et al. (1975, 1978) concluded that both branches of the autonomic nervous system e sympathetic and parasympathetic e are activated by facial cooling, and that the parasympathetic activation preponderates, leading to the decrease in HR. Heath and Downey (1990) and Lossius et al. (1994) also found that the sympathetic effect on HR is inferior to the vagal-parasympathetic deceleration of HR. In general, Heath and Downey (1990) found that facial cooling activates both the trigeminal-vagal-cardiac and trigeminal-sympathetic peripheral vasculature pathway function. Therefore, we conclude that the mild cooling during driving activated both the sympathetic and parasympathetic system. The increase in HR through hand cooling observed by Van Veen (2016) in a simulator study is not contradictory to our study since it is proven that hand and facial cooling cause different HR responses. Several studies comparing facial and hand cooling with colder temperatures were performed by LeBlanc et al. (1975, 1978) yielding similar results in non-vehicular settings. The limitation of the study is that both air temperature and air movement were varied in the COOL condition and the stronger fanning came along with an increased background noise. The change in noise level impacted very likely the first seconds of the physiological measurements due to the orienting response. The orienting response is a phenomenon which describes a set of physiological reactions to an unexpected stimulus of auditory, visual, tactual or thermal nature (Bradley, 2009; Frith and Allen, 1983; Graham and Clifton, 1966). As the studies of Bradley (2009) have shown, the orienting response lasts less than a few seconds and causes HR deceleration and an increase in SCL. This means that the reported reactions of these measures in our study were affected by this phenomenon because of the undesired effect of the increased noise level in the COOL condition. A pre-study in which we tested 6 min of fanning without a change in temperature on a small sample, can substantiate these hypotheses on the secondary effect of the orienting response caused by the intense fanning. In this pilot study we changed fanning intensity from 20% to 100% with a constant temperature of 23  C. In this case the physiological responses are as follows: SCL and pupil diameter increased noticeably in the first minute of fanning, but returned to pre-fanning levels after 1 min. HR

decelerated with onset of fanning and returned to pre-fanning level after 2 min. The HR reaction cannot solely be attributed to the orienting reflex which only lasts for seconds. Instead this reaction is the result of the vagal reflex explained earlier. As the study of LeBlanc et al. (1976) has shown, the reflex is caused by the stimulation of the trigeminal nerve of the face and it occurs with fanning at temperatures as high as 25  C. Both increasing wind speeds and decreasing temperatures were found to strengthen the bradycardia. Taking our pilot testing with constant temperature and results from the literature into account we can conclude that the secondary effect of changes in fan intensity and noise level affected the first minute of the electrodermal and pupillary responses. The HR measured in our study was affected for the entire COOL period by this secondary effect because it is known that both facial fanning and cooling separately decrease HR over the course of several minutes. A combined fanning and cooling lowers the HR even more (LeBlanc et al., 1976). Since in this study, a continued effect on SCL and especially on pupil diameter during the COOL treatment was measured, we assume that this continued activation is due the lower air temperature. As a consequence of the SNS activation in this study, drivers were more alert. A higher alertness is reflected in the increase of lane keeping quality (Fig. 8) and less jerky steering movements (Table 3). Against our expectations the variability in gaze direction, time and frequency domain HRV measures and BF did not change through the cooling. This differs from the results of Lal and Craig (2000) who found that eye movements are fast in a state of wakefulness compared to little eye movements during fatigue. The lack of effect on gaze variability, HRV and BF is also in contrast to previous simulator studies (Schmidt et al., 2016) in which those were correlated with subjective sleepiness. In that past study, however, changing traffic scenarios affected these parameters instead of in-vehicle treatments like AC. We therefore assume that the effect of cooling is not strong enough to be reflected in gaze variability, HRV and BF. 5. Conclusions In our study both objective and subjective responses indicated that cooling caused psychophysiological arousal and decreased sleepiness. According to subjects’ KSS ratings and self-evaluation via the ADACL the subjects felt more awake and activated after the cooling. The pupillary and electrodermal response indicated a SNS activation which leads to a state of alertness in the driver. This

E. Schmidt et al. / Applied Ergonomics 62 (2017) 9e18

increased short-term alertness could also be shown in the driving behavior parameters, e.g. in increased lane keeping quality and smoother steering movements. Furthermore, the amount of eye closures was decreased through the cooling, which is also a sign of decreased sleepiness. The decrease of HR in the cooling period indicates an activation of the parasympathetic branch of the autonomic nervous system next to the indicated sympathetic activation. In the case of a simultaneous activation of both sympathetic and parasympathetic system activation, the parasympathetic effect preponderates the sympathetic effect on HR which therefore decelerates. The COOL treatment consisted of a 6 min temperature reduction to 17  C and an increased air flow entailing a constant audible noise. The driver was therefore exposed to a multi-sensory stimulation, comprising thermal, tactile and auditory stimuli. Therefore, all measured subjective and objective differences between the conditions cannot be attributed solely to the cooling per se. Instead, the effects are due to a combination of thermal and auditory stimuli as well as tactile stimuli because of the air pressure on the skin. Based on the length of the effects on HR, SCL, pupil diameter, eye closure evidence and driving parameters, we recommend a cooling duration of 3 min. After this period, continued cooling will not activate the driver any further, as the objective data evaluation suggests. This would also align with the subjects’ impression that they acclimated to the cooling after a few minutes and that they only felt more awake in the short-term. These impressions can also be reasoned with the results of Reyner and Horne (1998) who found that a cooling period of 2 h at an even lower temperature did not affect sleepiness, whereas in the first minutes significantly lower KSS-ratings were recorded. In our opinion, a duration of cooling longer than a few minutes, will worsen the driving experience since it goes at the expense of the thermal comfort and deteriorates driving performance (Daanen et al., 2003). Further research should also focus on identifying the minimum temperature drop for which an activation of the driver can be measured. We conclude from the study results that short-term cooling at 17  C is a countermeasure against cognitive fatigue while driving in monotonous road conditions because it causes physiologically measureable activation. We also draw from the collected data that cooling only has a short-lasting effect on sleepiness. To highlight the effect of the cooling on the perceived sleepiness, we would like to point out that the sleepiness level after 26 min in the cooling condition is comparable to the subjects’ sleepiness level after 15 min of driving. Although a wakening effect of cooling during a monotonous driving task could be shown in our sample size, these results should be considered carefully and cannot be generalized for different samples. As the studies of LeBlanc et al. (1975, 1978) have shown, facial cooling effects differ between people of different ages, physical fitness and temperature acclimatization. Furthermore, the study did not investigate whether repeated cooling would still be effective against cognitive sleepiness. There is reason to assume that the change of thermal conditions is especially stimulating for the driver and an effect on sleepiness may be visible with every change of thermal conditions. Based on the studies of the orienting response with repeated stimuli (Goldwater, 1972; Graham and Clifton, 1966), we may also assume that immediate physiological responses like pupillary dilatation and HR decrease will be less apparent with repeated cooling. Therefore, it seems worthwhile to study the effect of repeated cooling over a longer period of time on the driver's subjective and physiological sleepiness. If a continued effect of repeatedly applied thermal stimuli on cognitive sleepiness could be proven, this may create new opportunities to increase alertness for longer than just a few minutes. Providing a variety of thermal stimuli that will

17

repeatedly cause arousal during monotonous driving could not only benefit car drivers, but also professional truck drivers experiencing sleepiness due to task underload. However, after hours of monotonous drivingeas it is often the case with professional drivers e sleepiness cannot not solely be attributed to task underload and lack of stimulation. In this case, it is likely that a combination of different types of sleepiness impair the driver, such as physical fatigue due to prolonged sitting as in the study of El Falou et al. (2003). Physical fatigue requires different countermeasures such as active micro-movements (Van Veen et al., 2014;Van Veen et al., 2015). Thus, we conclude that cooling is effective against early stages of fatigue that are attributed to task underload, but it is not suited to counteract physical or sleep-related fatigue experienced by professional drivers in prolonged driving. To encounter issues in logistics caused by driver fatigue, different types of countermeasures e with cooling being one of it e will be required to address different types of fatigue. The effect of short-term cooling on sleepiness explored in this study may be used in future automotive applications. As the results have shown, not only the car drivers' alertness increased but the drivers felt also positively activated. Other means of in-car sensory stimulation, for example olfaction or visual and auditory impulses might also serve as possibilities to reduce monotony while driving. However, the design of these types of intervention strongly depends on individual preferences. For thermal stimuli in contrast, the arousal-performance model (Hygge, 1992; Parsons, 2003) guides the design in terms of appropriate climate changes for increasing arousal. The arousing effect of thermal stimuli may be used as a part of a drivers’ assistance system which triggers cooling when sleepiness is detected. Such a system would not only require a precise sleepiness detection, but also contextual knowledge about the route and other environmental factors as well as information on the driver. Only by taking these contextual factors into account, it is possible to attribute sleepiness to task-underload in which case cooling can serve as an interim countermeasure. Acknowledgements The authors thank Stefan Wiedemann and Dr. Sebastian Hergeth who have provided support and expertise that greatly assisted our research. This research has profited from the insights provided by the three anonymous reviewers. We are grateful for their comments. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.apergo.2017.01.017. References Åkerstedt, T., Gillberg, M., 1990. Subjective and objective sleepiness in the active individual. Int. J. Neurosci. 52 (1e2), 29e37. Anund, A., Kecklund, G., Peters, B., Åkerstedt, T., 2008. Driver sleepiness and individual differences in preferences for countermeasures. J. sleep Res. 17 (1), 16e22. Beatty, J., 1982. Task-evoked pupillary responses, processing load, and the structure of processing resources. Psychol. Bull. 91 (2), 276e292. Bradley, M.M., Lang, P.J., 1994. Measuring emotion: the self-assessment manikin and the semantic differential. J. Behav. Ther. Exp. psychiatry 25 (1), 49e59. Bradley, M.M., Miccoli, L., Escrig, M.A., Lang, P.J., 2008. The pupil as a measure of emotional arousal and autonomic activation. Psychophysiology 45 (4), 602e607. Bradley, M.M., 2009. Natural selective attention: orienting and emotion. Psychophysiology 46 (1), 1e11. Briest, S., Karrer, K., Schleicher, R., 2006. Driving without awareness: Examination of the phenomenon, in: Gale, A. (Ed.), Vision in Vehicles XI, 159e167. Brown, I.D., 1994. Driver fatigue. Hum. Factors J. Hum. Factors Ergonomics Soc. 36 (2), 298e314.

18

E. Schmidt et al. / Applied Ergonomics 62 (2017) 9e18

Charlton, S.G., Starkey, N.J., 2011. Driving without awareness: the effects of practice and automaticity on attention and driving. Transp. Res. part F traffic Psychol. Behav. 14 (6), 456e471. Collins, K.J., Abdel-Rahman, T.A., Easton, J.C., Sacco, P., Ison, J., Dore, C.J., 1996. Effects of facial cooling on elderly and young subjects: interactions with breathholding and lower body negative pressure. Clin. Sci. 90 (6), 485e492. Daanen, H.A., van de Vliert, E., Huang, X., 2003. Driving performance in cold, warm, and thermoneutral environments. Appl. Ergon. 34 (6), 597e602. Egelund, N., 1982. Spectral analysis of heart rate variability as an indicator of driver fatigue. Ergonomics 25 (7), 663e672. Ekman, P., Rosenberg, E.L., 1997. What the face reveals: basic and applied studies of spontaneous expression using the Facial Action Coding System (FACS). Oxford University Press, New York. ^ne, J., Grabisch, M., Hewson, D., Langeron, Y., Lino, F., 2003. El Falou, W., Duche Evaluation of driver discomfort during long-duration car driving. Appl. Ergon. 34 (3), 249e255. EN ISO 14505e3, 2006. Ergonomics of the Thermal Environment - Evaluation of the Thermal Environment in Vehicles - Part 3: Evaluation of Thermal Comfort Using Human Subjects. Frith, C.D., Allen, H.A., 1983. The skin conductance orienting response as an index of attention. Biol. Psychol. 17 (1), 27e39. Gershon, P., Shinar, D., Ronen, A., 2009. Evaluation of experience-based fatigue countermeasures. Accid. Analysis Prev. 41 (5), 969e975. Gershon, P., Shinar, D., Oron-Gilad, T., Parmet, Y., Ronen, A., 2011. Usage and perceived effectiveness of fatigue countermeasures for professional and nonprofessional drivers. Accid. Analysis Prev. 43 (3), 797e803. Graham, F.K., Clifton, R.K., 1966. Heart-rate change as a component of the orienting response. Psychol. Bull. 65 (5), 305e320. Greco, A., Lanata, A., Valenza, G., Scilingo, E. P., Citi, L., 2014. Electrodermal activity processing: A convex optimization approach, in: Engineering in Medicine and Biology Society, 36th Annual International Conference of the IEEE, 2290e2293. Goldwater, B.C., 1972. Psychological significance of pupillary movements. Psychol. Bull. 77 (5), 340e355. Hallvig, D., Anund, A., Fors, C., Kecklund, G., Karlsson, J.G., Wahde, M., Åkerstedt, T., 2013. Sleepy driving on the real road and in the simulatorda comparison. Accid. Analysis Prev. 50, 44e50. Hayward, J.M., Holmes, W.F., Gooden, B.A., 1976. Cardiovascular responses in man to a stream of cold air. Cardiovasc. Res. 10 (6), 691e696. Heath, M.E., Downey, J.A., 1990. The cold face test (diving reflex) in clinical autonomic assessment: methodological considerations and repeatability of responses. Clin. Sci. 78 (2), 139e147. Hygge, S., 1992. Heat and performance. In: Jones, D.M., Smith, A.P. (Eds.), Handbook of Human Performance, vol. 1. The Physical Environment. Academic Press, London, pp. 79e104. Imhof, M., 1998. Erprobung der deutschen Version der Adjektiv-Checkliste nach Thayer (1989) zur Erfassung der aktuellen Aktiviertheit. Z. für Differ. Diagn. Psychol. 19, 179e186. Kaida, K., Åkerstedt, T., Kecklund, G., Nilsson, J.P., Axelsson, J., 2007. Use of subjective and physiological indicators of sleepiness to predict performance during a vigilance task. Ind. health 45 (4), 520e526. € hringer-Kuhnt, T., Baumgarten, T., Schleicher, R., 2005. Karrer, K., Briest, S., Vo Driving without awareness. In: Underwood, G. (Ed.), Traffic & Transport Psychology, Theory and Application, pp. 455e469. Lal, S. K., Craig, A., 2000. Driver fatigue: Psychophysiological effects. In International conference on fatigue and transportation, Fremantle, Australia. Lal, S.K., Craig, A., 2001. A critical review of the psychophysiology of driver fatigue. Biol. Psychol. 55 (3), 173e194. €m, U., Englund, K., Nordstrom, B., Stenudd, A., 1999. Laboratory studies on Landstro the effects of temperature variations on drowsiness. Percept. Mot. Ski. 89 (3), 1217e1229. €m, U., Englund, K., Nordstrom, B., Stenudd, A., 2002. Use of temperature Landstro variations to combat drivers' drowsiness. Percept. Mot. Ski. 95 (2), 497e506. LeBlanc, J., Dulac, S., Cote, J., Girard, B., 1975. Autonomic nervous system and adaptation to cold in man. J. Appl. Physiology 39 (2), 181e186. LeBlanc, J., Blais, B., Barabe, B., Cote, J., 1976. Effects of temperature and wind on

facial temperature, heart rate, and sensation. J. Appl. Physiology 40 (2), 127e131. LeBlanc, J., Cote, J., Dulac, S., Dulong-Turcot, F., 1978. Effects of age, sex, and physical fitness on responses to local cooling. J. Appl. Physiology 44 (5), 813e817. Lossius, K., Eriksen, M., Walløe, L., 1994. Thermoregulatory fluctuations in heart rate and blood pressure in humans: effect of cooling and parasympathetic blockade. J. Aut. Nerv. Syst. 47 (3), 245e254. May, J.F., Baldwin, C.L., 2009. Driver fatigue: the importance of identifying causal factors of fatigue when considering detection and countermeasure technologies. Transp. Res. Part F Traffic Psychol. Behav. 12 (3), 218e224. Niederl, T., 2007. Untersuchungen zu kumulativen psychischen und physiologischen Effekten des fliegenden Personals auf der Kurzstrecke. PhD thesis. University of Kassel. Oron-Gilad, T., Shinar, D., 2000. Driver fatigue among military truck drivers. Transp. Res. part F traffic Psychol. Behav. 3 (4), 195e209. Parsons, K., 2003. Human thermal environments. The Effects of Hot, Moderate and Cold Environments on Human Health, Comfort and Performance. CRC Press, New York. Patel, M., Lal, S.K.L., Kavanagh, D., Rossiter, P., 2011. Applying neural network analysis on heart rate variability data to assess driver fatigue. Expert Syst. Appl. 38 (6), 7235e7242. Philip, P., Sagaspe, P., Taillard, J., Valtat, C., Moore, N., Akerstedt, T., Charles, A., Bioulac, B., 2005. Fatigue, sleepiness, and performance in simulated versus real driving conditions. Sleep 28 (12), 1511e1516. Reyner, L.A., Horne, J.A., 1998. Evaluation of “in-car” countermeasures to sleepiness: cold air and radio. Sleep 21 (1), 46e50. Royal, D., 2003. National survey of distracted and drowsy driving attitudes and behavior: 2002 (vol. 1. Findings). U.S. Department of Transportation NHTSA DOT HS 809 566, Washington, DC (April). Russell, J.A., 1980. A circumplex model of affect. J. personality Soc. Psychol. 39 (6), 1161e1178. Schmidt, E.A., Schrauf, M., Simon, M., Buchner, A., Kincses, W.E., 2011. The shortterm effect of verbally assessing drivers' state on vigilance indices during monotonous daytime driving. Transp. Res. part F traffic Psychol. Behav. 14 (3), 251e260. Schmidt, E., Decke, R., Rasshofer, R., 2016. Correlation between subjective driver state measures and psychophysiological and vehicular data in simulated driving, in: IEEE Intelligent Vehicles Symposium, 1380e1385. Schwarz, J.F., Ingre, M., Fors, C., Anund, A., Kecklund, G., Taillard, J., Åkerstedt, T., 2012. In-car countermeasures open window and music revisited on the real road: popular but hardly effective against driver sleepiness. J. sleep Res. 21 (5), 595e599. Tejero Gimeno, P., Pastor Cerezuela, G., Choliz Montanes, M., 2006. On the concept and measurement of driver drowsiness, fatigue and inattention: implications for countermeasures. Int. J. Veh. Des. 42 (1e2), 67e86. Thayer, R.E., 1989. The Biopsychology of Mood and Arousal. Oxford University Press, New York. Thiffault, P., Bergeron, J., 2003. Monotony of road environment and driver fatigue: a simulator study. Accid. Analysis Prev. 35 (3), 381e391. Van Veen, S., Vink, P., Franz, M., Wagner, P.O., 2014. Enhancing the vigilance of car drivers: a review on fatigue caused by the driving task and possible countermeasures, in: Vink, P. (Ed.), Advances in Social and Organizational Factors, 516e525. Van Veen, S., Orlinskiy, V., Franz, M., Vink, P., 2015. Investigating car passenger wellbeing related to a seat imposing continuous posture variation. J. Ergonomics 5 (3), 140. Van Veen, S., 2016. Driver Vitalization: Investigating Sensory Stimulation to Achieve a Positive Driving Experience (PhD thesis, TU Delft). Watson, D., Tellegen, A., 1985. Toward a consensual structure of mood. Psychol. Bull. 98 (2), 219e235. Wyon, D.P., 1973. The role of the environment in buildings today: thermal aspects (factors affecting the choice of a suitable room temperature). Build. Int. 6 (1), 39. Wyon, D.P., Wyon, I., Norin, F., 1996. Effects of moderate heat stress on driver vigilance in a moving vehicle. Ergonomics 39 (1), 61e75.