http://pro.sagepub.com/Ergonomics Society Annual Meeting

Proceedings of the Human Factors and

http://pro.sagepub.com/content/58/1/2038The online version of this article can be found at:

 DOI: 10.1177/1541931214581425

2014 58: 2038Proceedings of the Human Factors and Ergonomics Society Annual MeetingDrew M Morris and June J Pilcher

The Impact of Cold Stress on Driving Performance  

Published by:

http://www.sagepublications.com

On behalf of: 

  Human Factors and Ergonomics Society

can be found at:Proceedings of the Human Factors and Ergonomics Society Annual MeetingAdditional services and information for    

  http://pro.sagepub.com/cgi/alertsEmail Alerts:

 

http://pro.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

http://pro.sagepub.com/content/58/1/2038.refs.htmlCitations:  

What is This? 

- Oct 17, 2014Version of Record >>

at CLEMSON UNIV on October 24, 2014pro.sagepub.comDownloaded from at CLEMSON UNIV on October 24, 2014pro.sagepub.comDownloaded from

The Impact of Cold Stress on Driving Performance

Drew M Morris, & June J Pilcher

Clemson University

Department of Psychology

Exposure to cold environments can impact complex task performance due to increased error from cognitive

and physiological stress. Few studies have examined the effects of cold stress on driving performance or

the potential for advanced driver safety systems to detect error. The current study aims to examine the

effects of cold stress by way of skin cooling on driving simulator performance, and evaluate vehicle

behavior metrics for possible dangerous driving detection systems. Driving under cold stress is expected to

result in systematic vehicle behavior and driving performance error, which can be utilized for future safety

system research and development.

INTRODUCTION

Healthy humans require thermal stability and comfort to

perform optimally. In cold environments, physical

performance and cognitive performance can suffer as a result

of thermal stress. Complex tasks like driving require the use of

calibrated mental and physical actions to be conducted safely,

and may be especially susceptible to thermal stress.

Furthermore, driving tasks require accurate perception,

sustained attention, and muscle control – elements most

vulnerable to body cooling (Mäkinen, 2006).

Most Americans use a personal vehicle to travel daily,

and the majority of drivers live in regions that experience

uncomfortably cold temperatures during several months of the

year. Modern climate control systems and vehicle preheating

help to moderate cabin temperatures, but take a considerable

amount of time to fully alter the environment in winter

conditions. Leaving the driver exposed to debilitating thermal

stress capable of immediately impacting performance. Despite

this, few studies have looked at the impact of thermal stress on

driving performance, and none have researched detection

technology associated with cold stress impaired driving. The

practical relevance of such research in transportation safety is

indisputable, and would provide a scientific foundation for

future thermal stress driving studies.

Cold Stress and Cognitive Performance

Studies and reviews have shown the effects of thermal

stress on cognitive performance across a variety of tasks. By

definition, thermal stress involves either heating or cooling the

body to the point of stress. Temperature has a negative impact

on cognitive performance because of the body’s response to

this stress. It can be assumed then that driving tasks requiring

effective cognitive performance may be curtailed by cold

stress. Paradigms attempting to best explain the relationship

between cognitive performance and thermal stress have been

proposed throughout the years. Perhaps the most prominent

and applicable of these is the Yerkes and Dodson inverted-U

theory, discussed at length in a meta-analytic review by

Hancock, Ross, and Szalma (2007), suggesting that increased

stress is met with decreased human performance. Work by

Hancock and Warm (1989) expanded on this model with their

own extended-U model, which focuses on attentional resource

capacity and its homeostatic nature until a threshold of stress

is crossed. On performance measures during cold stress,

research has consistently shown that performance is affected

by several independent elements, such as temperature prior to

and during task, the duration of the cold exposure, as well as

the type of task being performed. As such, vehicle operation is

considered a complex task, and requires several of the

cognitive elements that are impacted by temperature.

Temperature prior to and during the task is often the

primary manipulation in cold stress research. These prior

temperatures are known to influence the perception of later

temperature, due to the rapid adaptability of thermoreceptors

(Parsons, 2002). Accordingly, temperatures may feel colder

than they really are because of recent exposures to less cold

environments (e.g., a vehicle cabin after leaving a warm

home), and impact attention. Looking at the effect of cold

exposure intensity, temperatures below thermal neutral (i.e.,

cold) have been correlated with generally reduced cognitive

performance. Meta-analytic findings have shown that

participants performing cognitive tasks in a cold conditions

(50-64.9ºF) showed a 7.81% decrement in performance, while

participants in colder conditions (<50ºF) showed an even

worse 13.91% decrement in performance – as compared to a

thermal neutral condition (65-75ºF) (Pilcher, Nadler, & Busch,

2002). Temperatures commonly found in vehicle cabins

during winter months.

When reviewing duration of cold stress exposure, Pilcher

et al. (2002) had also shown that exposures of less than one

hour have a greater negative impact on cognitive performance

than exposures of more than one hour, perhaps due to the

aforementioned effects of prior temperature exposure.

However, this does not mean that long term repeated exposure

to cold results in more fitness to drive. Participants involved in

a ten day repeated exposure study at 50ºF did not show

significant improvement on cognitive tasks over time when

compared to a control group (Mäkinen et al., 2006). Long

exposure still resulted in increased response time, decreased

accuracy and decreased efficiently. All of which are vital

components of proactive and safe driving.

The impact of thermal stress on cognitive performance is

also dependent on the task being performed. When analyzing

task performance by categories, faculties used to operate a

vehicle seem to be most at risk. Pilcher et al. (2002) found that

tasks based on reasoning, learning, or memory saw a severe

Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 2038

Cop

yrig

ht 2

014

Hum

an F

acto

rs a

nd E

rgon

omic

s S

ocie

ty. D

OI 1

0.11

77/1

5419

3121

4581

425

at CLEMSON UNIV on October 24, 2014pro.sagepub.comDownloaded from

28.05% decrement in performance during cold exposure, and a

moderate 7.81% decrement in attentional and perceptual tasks.

Other reviews of thermal stress and cognitive performance

have also found that cold stress has the largest impacts on

perceptual based tasks and psychomotor performance

(Hancock et al., 2007).

Loss of attention from distraction is often cited as the

primary reason for poor cognitive performance during cold

stress. Importantly, decreased attention is also one of the

leading causes of automotive accidents. Cheung, Westwood,

and Knox, (2007) found that skin temperature cooling

impaired attention by acting as a distractor, and cooling core

body temperature did not further impair cognitive

performance. Additionally, research by Muller et al. (2011)

showed that using exercise to increase body core temperature

during cold stress did not improve attention while the

participants were subjected to a cold environment, once again

due to distraction. This gives support to the theory that cold

environments impact cognitive performance through simple

skin temperature cooling, which can be altered very quickly if

a vehicle cabin is not yet heated.

Complimentary to the distracting temperature theory is

the arousal paradigm, in which cold environments introduce a

temporary arousal response from stress, limiting cognitive

performance and accuracy (Hancock, 1986). One study found

that cold air was enough to induce adrenergic stress in a

command and control task (Benoit, van Orden, & Osga, 1996).

The result of this arousal suggested a speed-accuracy tradeoff,

whereby reaction time improved but error rate increased in

repose to the stress (Enander, 1987). This change in response

and error rate may translate to dangerous driving behavior

when considering a cold vehicle, and is a primary question in

this study.

Dangerous Driving Detection Technology

The potential for technology to inform a driver of their

impairment is of great benefit to transportation safety, as well

as advanced driver assistance systems (ADAS) literature

(Werneke, Kleen, & Vollrath, 2014). Physical and cognitive

impairment results in variability in vehicle behavior. Current

automotive research has shown that driver impairment often

result in systematic and predictable variability, and can be

used as an indicator of dangerous driving.

Currently, ADAS such as drowsy driving monitoring

systems are available in modern vehicles, and have been

designed to detect driver impairment due to drowsiness.

Though the development of this technology is ongoing, there

have been two distinct methods of approaching the problem of

dangerous driving detection due to drowsiness. The more

invasive of the two is the operator behavior approach using

biometric indicators. Using this method, an intelligent system

monitors the behavior of the driver directly without any input

from the vehicle. These direct measures include monitoring

the driver’s eyes (PERCLOS) or using electroencephalogram

(EEG) to monitor brain activity, both of which are indicative

of a driver entering early stages of sleep (Kecklund &

Åkerstedt, 1999; Sandberg et al., 2011). Other systems have

also used various combinations of electromyogram (EMG) to

monitor loss of muscle tonus or atypical muscle movements

associated with vehicle correction, another indicator of sleep

onset (Lal & Craig, 2001). However, as is the issue with

biometric measures in a natural setting, they can be intrusive

and unreliable (Shuyan & Gangtie, 2009).

The second and more applicable of the two methods of

detecting dangerous driving is the vehicle behavior and lane

variability approach using vehicle indicators. Using this

method, an intelligent system monitors the driver indirectly

using input from the vehicles behavior. These measures

include monitoring the relative location and variability of a

vehicle within the lane, as well as tracking instances when the

vehicle leaves the lane completely (Forsman, Vila, Short,

Mott, & Van Dongen, 2013; Sandberg, Åkerstedt, Anund,

Kecklund, & Wahde, 2011). Additional measures also include

monitoring for atypical steering wheel movements or

accelerator and brake pedal forces (Liu, Hosking, & Lenné,

2009). Unlike the biometric based method, monitoring the

vehicle behavior may be directly applicable to cold stress

driving impairment detection. Vehicle behavior metrics are

generally nonintrusive, reliable and have been correlated with

an impaired driver.

In drowsy driving, cognitive and physical decline results

in systematic driving error following sleep deprivation. The

application of current drowsy driving detection technology to

thermal stress research is dependent on driving behavior under

cold stress being similar to driving behavior while drowsy. As

has been shown, cold stress impacts performance on cognitive

tasks by impairing perception, reasoning, learning and

memory. Cognitive performance decline is critically linked to

attentional deficit due to the distracting effects of cool skin

temperature (Cheung et al., 2007). Additionally, cold stress

impacts performance on physical tasks by impairing haptic

sensitivity, manual dexterity, and fine motor control. By

comparison, sleep deprivation has also been shown to impact

performance on cognitive tasks by impairing perception,

reasoning, learning and memory (Goel, Basner, Rao, &

Dinges, 2013). Drowsiness is also linked to attentional deficit

similar to cold stress (Langner & Eickhoff, 2013; Roca et al.,

2012). In further endorsement of the similarities, sleep

deprivation can impact performance on physical tasks by

impairing manual dexterity and fine motor control (Eastridge

et al., 2003).

Because of the significant overlap in task performance

decrement between drowsiness and cold stress, cold drivers

may demonstrate the same systematic errors as drowsy

drivers. Standard deviation of lateral lane position has long

been held as one of the most popular and reliable

measurements of drowsy driving detection, and indexes

dangerous driving through vehicle behavior. A single study on

the topic of cold environment driving by Daanen, van de

Vliert, and Huang (2003) found that cold stress can result in

significantly increased driving performance error on a low

fidelity simulator when measuring lateral lane position,

similarly to drowsy driving.

Purpose and Hypotheses

Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 2039

at CLEMSON UNIV on October 24, 2014pro.sagepub.comDownloaded from

The purpose of the current research is to study the effect

of cold stress on driving performance, and the potential to

detect driving error due to thermal stress impairment. Such

protocol would allow for a convenient assessment of driver

fitness where other measures are not practical, and employ

behavioral measures commonly seen in drowsy driving

literature.

Even at moderate temperatures, previous research has

shown that lowered skin temperature results in poorer

cognitive and physical performance that may negatively affect

driving ability. The introduction of a thermal stressor is

expected to result in poor driving performance due to both

limited attentional resources, and physical reaction to the cold.

Existing literature has shown that sleepiness has a detrimental

impact on driving performance, increasing driving errors and

limiting lane keeping ability. Research has also shown the

detrimental effects of cold stress on cognitive and physical

performance overlaps with that of sleepiness. Thus cognitive

and physical tolls of cold stress are expected to result in poor

driving performance that can be measured through traditional

drowsy driving detection measures. Characteristics of poor

driving performance include difficultly keeping a consistent

driving line and more variability in the operation of vehicle

controls such as pedals and steering wheel.

METHOD

Participants

Approximately 30 participants with a valid driver’s

license will participate in this study. Participants will be told

to report to the lab well rested, removed from caffeine and

alcohol, and wearing a standardized set of clothing which

gives 0.24 Clo of insulation. Clothing insulation values will be

generalized using a standardized Clo scale seen in Parsons,

2002.

Design

The study will use a between-subjects design with the

participant randomly assigned to one of two conditions. One

condition will be the Cold condition, in which the participant

will complete all tasks while wearing a cooling vest in a windy

environment. The second condition will be the Thermal

Neutral condition, in which the participant will complete all

tasks while wearing a room temperature version of the cooling

vest in a not windy environment.

In each condition, the participant will undergo an active

training and testing session with a driving simulator task,

followed by two short qualitative questionnaires to assess their

feelings of simulator sickness (SSQ) and perceived thermal

comfort (TCA). These qualitative scales will measure whether

the participant is experiencing any ill effects of using the

simulator, and to measure how the participant perceives the

effects of the thermal condition. As participant sleepiness will

be a confounding characteristic of performance, the Stanford

Sleepiness Scale will be used as an inclusion measure by

recording subjective feelings of tiredness prior to testing.

Accordingly, laboratory measures will take place during

traditional wakeful hours to avoid participant drowsiness.

During the driving task, data from the participant’s

performance will be logged in real time. The participant will

be told to inform the researcher if they feel nauseous from the

simulator. Next the participant will complete three additional

tasks (two cognitive performance measures and one physical

performance measure) in a randomized order, followed by

another administration of the TCA. Following the completion

of the TCA, the testing session will be finished.

The independent variable being manipulated will be the

presence of a thermal stressor. Quantitative cold stress will be

measured using mean skin temperature from four locations,

wind chill calculations, and internal temperature from two

locations. The dependent variable will be the participant’s

performance on the driving task, performance on the cognitive

and physical measures, as well as their qualitative thermal

comfort on the TCA. Of the driving measures, lane position,

lane crossings, steering wheel angle, vehicle velocity, and

difference in vehicle heading will be of primary interest.

Materials and Tasks

Participants will perform a driving task using a high

fidelity automotive driving simulator in a projected

environment. For both conditions, the simulated environment

will depict a clear daytime driving scenario through 10

minutes of rural straight roads, and 10 minutes of rural

winding roads. In total the participants will take approximately

20 minutes to complete the full circuit at 55 miles per hour.

The participants will be given a 5 minute training session to

become familiar with the simulator by completing a shortened

version of various road segments found in the testing circuit.

No other vehicles will be present during either the training or

the testing session. Participants will be told to drive as they

normally would, and to obey all traffic signage and laws.

During the testing session, the driving simulator will record

lane position, number of lane crossings, vehicle heading and

lane heading, and steering wheel angle at a sample rate of 10

Hz.

Participants will perform a reaction task using a hand held

Psychomotor Vigilance Testing device (PVT, Ambulatory

Monitoring, Inc.). The participants will be told to press a

response button as soon as the device visually prompts them,

where by the PVT unit will log the delay in the participant’s

response in milliseconds. The time elapse data will be used to

record differences in reaction times between conditions, and

users will be prompted randomly after a 2 to 10 second pause

between trials. The task will take 5 minutes to complete. The

participants will be given a training session to become familiar

with the device by completing a shortened 1 minute session of

the typical task.

In addition, participants will perform an attention task

using an additional PVT device. Similarly, the participants

will be told to press a response button as soon as the device

visually prompts them, where by the PVT unit will log the

delay in the participant’s response in milliseconds. The time

elapse data will be used to record differences in sustained

attention between conditions if there was a delay in response

Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 2040

at CLEMSON UNIV on October 24, 2014pro.sagepub.comDownloaded from

longer than 500 milliseconds (i.e., response lapse), as users

will be prompted randomly after an extended 45 to 60 second

pause between trials. The task will take 10 minutes to

complete. The participants will be given a training session to

become familiar with the device by completing a shortened 1

minute session of the typical task.

Participants will also perform a hand-eye dexterity test

using the Minnesota Manual Dexterity Placing Task (MMDT,

Lafayette Instruments). Participants will be told to place sixty

circular disks into sixty circular holes as quickly as possible

using one hand. The task has been shown to measure an

individual’s rapid eye-hand coordination and dexterity. The

task will take approximately 2 minutes to complete. The

participants will be given a training session to become familiar

with the device by completing a 2 minute walkthrough of the

task. The time required to complete the task will be used to

record loss of dexterity due to cold stress.

An adjustable nylon vest with large plastic packs filled

with a phase-change material will act as the cold stressor

during the cold conditions. The phase-change material is a

freezable liquid, which acts as a heat sink to remove heat from

the surface of the participant’s anatomical trunk. Additional

use of an in-simulator mounted fan will add a wind chill factor

to encourage skin cooling.

Internal body temperature will be estimated using both an

infrared tympanic thermometer, as well as an oral

thermometer. An infrared thermometer will be placed in the

participant’s ear and used to record the temperature of the

tympanic membrane. Additionally, a digital probe

thermometer will be placed under the participants tongue with

mouth closed for 90 seconds and used to record the

temperature near the lingual artery. Temperatures from both

methods will be used to extrapolate estimates of internal body

temperature.

Skin temperature will be logged using a portable four-

channel real-time data logger with insulated thermocouples

from multiple locations. The unit will log mean skin

temperature at 0.1 Hz using Type K chromel-alumel

thermocouples from four locations according to ISO 9886.

The thermocouples will be secured using breathable medical

tape, and remain on the participant throughout the entirety of

the laboratory testing session.

Procedure

Participants will report to the lab between the hours of

11:00 a.m. and 4:00 p.m. Once they arrive, participants will be

given the chance to have questions answered before they sign

the informed consent and start any procedures. Once signed,

information on health and basic biometrics will be taken, and

the thermocouples to be used for the skin temperature

measures attached. The basic biometrics will include height in

inches, weight in pounds, internal body temperature from two

locations, and body fat percentage. Skin temperature data from

the thermocouples will be logged as part of the initial

biometrics measure and continuously logged throughout the

entirety of the training and testing sessions. Participants will

then fill out a form which assesses sleep habits and their rating

on the TCA and SSQ qualitative scales.

Before any testing will occur, participants will undergo a

training session with each task and will not start the testing

session until they feel comfortable with the procedures

involved. This will include driving a training track with the

driving simulator, two shortened sessions with the PVT units,

and a slow walkthrough session with the MMDT.

Once training is completed, the participants will then start

the randomly assigned condition. Before the driving task, the

participants will put on the corresponding vest. The Thermal

Neutral condition will involve performing all tasks while

wearing a non-cooled cooling vest. Following this adjustment

period, participants will begin the driving task. After the

completion of driving circuit, internal body temperature will

be taken from the two locations and the SSQ and TCA

administered.

Participants will then be assigned to the three additional

tasks in a randomized order, two to measure attention and

reaction time, and one to measure dexterity. Following the

completion of one task, the participants will immediately start

the next task. Participants will complete the two cognitive

tasks with the handheld button box inside of the simulator

cabin. The dexterity task will be completed outside of the

cabin at a station with additional cooling fans for the Cold

condition. Following the completion of the three additional

tasks, the TCA will again be administered to qualitatively

measure perceived thermal stress, and internal body

temperature will be recorded. Once the participants complete

this assessment, the condition will be finished.

With the exception of the thermal stressors presence, the

Cold condition will be identical to the Thermal Neutral

condition. The moving air will come from a fan inside the

simulator which pushes room temperature air against the

participants’ lower body to exacerbate cold perception.

EXPECTED RESULTS

It is anticipated that cold stress will impact performance

on all tasks as compared to the Thermal Neutral condition. On

the driving performance task, the Cold condition is expected

to show increased error across all measures. Lane keeping is

expected to become more difficult during thermal stress,

resulting in increased lane crossings, and larger statistical

variability in lateral lane position and heading difference

metrics. In addition, this increased need to make corrections

due to lane variability should also be reflected in the steering

wheel reversal rate and vehicle velocity. These errors are

expected to resemble those seen in previous drowsy driving

studies, and allow for similar detection methods to be applied

to dangerous driving due to thermal stress.

Performance on the MMDT is expected to decrease, as

would be evident by an increase in the time to complete the

task due to cold induced vasoconstriction limiting dexterity.

Performance on the PVT attention lapse task is also expected

to decrease, as would be evident by an increase in the number

of responses that took longer than 500 milliseconds due to the

distracting effects of cold stress. However, performance on the

PVT reaction time task is expected to increase and decrease,

as would be evident by a decrease in the mean response time,

but an increase in the number of false start responses due to

Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 2041

at CLEMSON UNIV on October 24, 2014pro.sagepub.comDownloaded from

the physiological speed-accuracy tradeoff of cold stress. On

the subjective comfort measure, participants are expected to

show a decrease in their self-reported thermal comfort in the

Cold conditions due to the cold stressor.

DISCUSSION

Due to the dangerous nature of impaired driving, factors

that may curtail the ability to drive safely should be explored.

Cold stress has been shown to result in both cognitive and

physiological error in general task performance, but has not

been sufficiently explored in driving specific tasks.

Furthermore, the identification of a problem should be coupled

with a proposed solution. Drowsy driving detection

technology has been shown to reduce driving error by

monitoring vehicle behavior metrics and providing driver

feedback. These same metrics may be used for the detection of

impaired driving due to cold stress, and should also be

investigated. Because of the potential application of this

research, future studies should expand to include a wide range

of cold stress intensities and durations. In addition, other

metrics and methods of dangerous driving detection should be

explored to better understand how performance variability due

to cold stress could best be measured.

REFERENCES

Benoit, S. L., van Orden, K. F., & Osga, G. A. (1996). Effects of cold air

stress on the performance of a command and control task. Human Factors, 38(1), 130-141.

Cheung, S., Westwood, D., & Knox, M. (2007). Mild body cooling impairs

attention via distraction from skin cooling. Ergonomics, 50(2), 275-288.

Daanen, H. A., van de Vliert, E., & Huang, X. (2003). Driving performance in

cold, warm, and thermoneutral environments. Applied Ergonomics, 34(6),

597-602. Eastridge, B. J., Hamilton, E. C., O'Keefe, G. E., Rege, R. V., Valentine, R. J.,

Jones, D. J., . . . Thal, E. R. (2003). Effect of sleep deprivation on the

performance of simulated laparoscopic surgical skill. The American Journal of Surgery, 186(2), 169-174.

Enander, A. (1987). Effects of moderatecold on performance of psychomotor

and cognitive tasks. Ergonomics, 30(10), 1431-1445. Forsman, P. M., Vila, B. J., Short, R. A., Mott, C. G., & Van Dongen, H. P.

(2013). Efficient driver drowsiness detection at moderate levels of

drowsiness. Accident Analysis and Prevention, 50, 341-350. Goel, N., Basner, M., Rao, H., & Dinges, D. F. (2013). Circadian rhythms,

sleep deprivation, and human performance. Progress in Molecular

Biology and Translational Science, 119, 155-190. Hancock, P. A. (1986). Sustained attention under thermal stress.

Psychological Bulletin, 99, 263-281.

Hancock, P. A., Ross, J. M., & Szalma, J. L. (2007). A meta-analysis of performance response under thermal stressors. Human Factors, 49(5),

851-877.

Hancock, P. A., & Warm, J. S. (1989). A dynamic model of stress and sustained attention. Human Factors, 31(5), 519-537.

Kecklund, G., & Åkerstedt, T. (1999). Sleepiness in long distance truck

driving: An ambulatory EEG study of night driving. Ergonomics, 36(9), 1007-1017.

Lal, S. K., & Craig, A. (2001). The characteristics of sleepiness during real

driving at night — A study of driving performance, physiology and subjective experience. Biological Psychology, 55(3), 173-194.

Langner, R., & Eickhoff, S. B. (2013). Sustaining attention to simple tasks: A

meta-analytic review of the neural mechanisms of vigilant attention. Psychological Bulletin, 139(4), 870-900.

Liu, C. C., Hosking, S. G., & Lenné, M. G. (2009). Predicting driver

drowsiness using vehicle measures: Recent insights and future challenges. Journal of Safety Research, 40(4), 239-245.

Mäkinen, T. (2006). Human cold exposure, adaptation and performance in a

northern climate. (Unpublished Doctoral dissertation). Thule Institute, University of Oulu; Finnish Institute of Occupational Health, Oulu,

Finland. (D 876)

Mäkinen, T., Palinkas, L., Reeves, D., Pääkkönen, T., Rintamäki, H., Leppäluoto, J., & Hassi, J. (2006). Effect of repeated exposures to cold on

cognitive performance in humans. Physiology & Behavior, 87(1), 166-

176. Muller, M. D., Muller, S. M., Kim, C., Ryan, E. J., Gunstad, J., & Glickman,

E. L. (2011). Mood and selective attention in the cold: The effect of

interval versus continuous exercise. European Journal of Applied Physiology, 111, 1321-1328.

Parsons, K. C. (2002). Human thermal environments: The effects of hot,

moderate, and cold environments on human health, comfort and performance (Second Edition ed.). London: Taylor and Francis.

Pilcher, J. J., Nadler, E., & Busch, C. (2002). Effects of hot and cold

temperature exposure on performance: A meta-analytic review. Ergonomics, 45(10), 682-698.

Roca, J., Fuentes, L. J., Marotta, A., López-Ramón, M., Castro, C., Lupiáñez,

J., & Martella, D. (2012). The effects of sleep deprivation on the

attentional functions and vigilance. Acta Psychologica, 140(2), 164-176.

Sandberg, D., Åkerstedt, T., Anund, A., Kecklund, G., & Wahde, M. (2011).

Detecting driver sleepiness using optimized nonlinear combinations of sleepiness indicators. IEEE Transactions on Intelligent Transportation

Systems, 12(1), 97-108.

Sandberg, D., Anund, A., Fors, C., Kecklund, G., Karlsson, J. G., Wahde, M., & Åkerstedt, T. (2011). The characteristics of sleepiness during real

driving at Night—A study of driving performance, physiology and subjective experience. Sleep, 34(10), 1317-1325.

Shuyan, H., & Gangtie, Z. (2009). Driver drowsiness detection with eyelid

related parameters by support vector machine. Expert Systems with Applications, 36(4), 7651-7658.

Werneke, J., Kleen, A., & Vollrath, M. (2014). Perfect timing: Urgency, not

driving situations, influence the best timing to activate warnings. Human Factors, 56(2), 249-259.

Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 2042

at CLEMSON UNIV on October 24, 2014pro.sagepub.comDownloaded from