A Smartwatch-Based Approach to Support and Analysis

of Driver Stress and Anxiety

Tiago Mota de Oliveira

1 a

, Andr

e Roberto Ortoncelli

2 b

, Claudemir Casa

1 c

Claudinei Casa

3 d

and Luciano Silva

1 e

Federal University of Paran

a, Polytechnic Center, Curitiba, Brazil

Federal University of Technology, Paran

a, Dois Vizinhos, Brazil

Pontiﬁcal Catholic University of Paran

a, Curitiba, Brazil

Keywords:

Smartwatch, Stress Detection, Driving Education, Human-Computer Interaction, Emotional Regulation.

Abstract:

This study presents a smartwatch-based solution for monitoring drivers’ stress and anxiety levels using heart

rate data, standing out for not requiring synchronization with other devices. The system captures heart rate

variations and GPS coordinates, offering real-time feedback to assist drivers while storing all data in a cloud

database for subsequent expert analysis. Additionally, a reporting tool is provided to help specialists (e.g.,

psychologists) evaluate drivers’ emotional states and offer appropriate support. A pilot study was conducted

with eleven students from a driver training center during practical lessons to assess the proposed system. The

results show that the application was positively received by all participants, with two expressing interest in

using it beyond the study. These ﬁndings suggest that the proposed solution could enhance driver well-being

and preparedness, particularly among new drivers.

1 INTRODUCTION

Trafﬁc accidents are one of the world’s leading causes

of death (and the primary cause for individuals aged

5 to 29) resulting in approximately 1.19 million fatal-

ities and between 20 and 50 million non-fatal injuries

worldwide each year (WHO, 2023). In addition to the

consequences for victims and their families, these in-

cidents entail high economic costs, such as hospital

expenses (Alhaek et al., 2024). In Brazil, in 2024,

there were more than 13 thousand deaths and more

than 1 million injuries in trafﬁc accidents (Minist

erio

dos Transportes, 2023).

Factors such as stress, anxiety and fear of driv-

ing contribute to the number of trafﬁc accidents and

cause other problems for drivers. For example, ex-

cessive stress generally leads to degraded driving per-

formance, which increases the risk of road accidents

(Zhao et al., 2024). Furthermore, driving anxiety cor-

relates with poorer quality of life (Stephens et al.,

https://orcid.org/0000-0001-7943-268X

https://orcid.org/0000-0001-9622-8525

https://orcid.org/0000-0002-1812-4701

https://orcid.org/0009-0002-9723-3865

https://orcid.org/0000-0001-6341-1323

2024), and ﬁnally, amazophobia (fear of driving), af-

fects about 10% of the world’s population (Craske

et al., 2009). In fact, individuals with amaxopho-

bia often avoid driving, which limits their mobility

and quality of life, in addition to generating socioeco-

nomic impacts (Vansteenwegen et al., 2014).

Different physiological parameters are explored in

the literature to detect drivers’ stress and anxiety, such

as pupil dilation (Marquart et al., 2015), blood pres-

sure (Li et al., 2021), galvanic skin response (Lee

et al., 2023), among other physiological signals (Lin

and Li, 2023), with emphasis on heart rate (Xu et al.,

2024), which are widely studied (Haque et al., 2024).

Based on drivers’ physiological signals, some

studies focus on improving driver safety in real-time

(Braun et al., 2019), exploring automotive user inter-

faces to assist drivers and help them during practi-

cal driving activities (Li et al., 2021). These meth-

ods commonly require different hardware resources

that may demand experts or have a high cost, making

some of these solutions inaccessible to most of the

population due to their cost or technical complexity.

Many studies propose intelligent solutions to sup-

port drivers by detecting physiological signals related

to stress and anxiety. However, few have focused

544

de Oliveira, T. M., Ortoncelli, A. R., Casa, C., Casa, C. and Silva, L.

A Smartwatch-Based Approach to Support and Analysis of Driver Stress and Anxiety.

DOI: 10.5220/0013717000003985

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

In Proceedings of the 21st International Conference on Web Information Systems and Technologies (WEBIST 2025), pages 544-552

ISBN: 978-989-758-772-6; ISSN: 2184-3252

on developing actual tools that summarize drivers’

behavior and emotional states into reports for later

expert analysis. Solutions that provide insights into

drivers’ behavior and stress/anxiety levels are essen-

tial for experts (such as psychologists) to offer appro-

priate support (Ortoncelli et al., 2020).

This work contributes to the literature by present-

ing a solution that monitors drivers’ stress and anxiety

levels using heart rate data, measured exclusively us-

ing a smartwatch without requiring synchronization

with any other device, such as a smartphone. The

proposed solution collects data from the driver’s heart

rate and the vehicle’s GPS coordinates. Based on vari-

ations in heart rate, the application provides feedback

to assist drivers in their driving activities. It stores all

collected data in a cloud database, allowing experts to

access the reports generated later and provide special-

ized support to drivers.

We decided to develop a solution that uses a

smartwatch as the only device since there are mod-

els of watches that collect heart rate and run An-

droid applications, making it possible to develop a

simple solution that does not require synchronization

with other devices and can be used quickly and eas-

ily. In addition, smartwatches have become popular

with applications for leisure, health and research pur-

poses (Barka and Politis, 2024; Pinge et al., 2024).

In the transport sector, smartwatches facilitate cost-

effective, unintrusive and dynamic monitoring of the

driver’s state and behavior while on the road.

In this context, the main contributions of this work

are highlighted below:

• An application for smartwatches with the Android

operating system monitors and supports drivers

by collecting heart rate data and GPS coordi-

nates during driving. It stores the data in a cloud

database in real time and provides instant feed-

back to help drivers manage stress and anxiety;

• A script to produce reports that summarize the

driver’s heart rate and vehicle movement, so that

specialists (such as psychologists) can analyze the

driver’s data and provide appropriate individual-

ized care;

• The source code

of all implemented solutions is

publicly available, allowing researchers and de-

velopers to access, review and enhance the sys-

tem;

• The pilot case study was conducted with eleven

students from a driving school in practical classes

to obtain their National Driver’s License, follow-

ing procedures previously approved by the Ethics

https://github.com/motaoliveiraufpr/CFCStress

Committee for Research with Human Beings. We

selected this public because we believe that pro-

viding adequate support to the stress and anxiety

levels of driving students is essential so that they

are better prepared to deal adequately with real

trafﬁc situations and can pass the practical driving

test in less time.

Regarding the experiments carried out, in the par-

ticipants’ opinion, the use of the application was pos-

itive. Two of the eleven participants (the two with the

highest stress level, according to the driving school

instructor) showed interest in purchasing the applica-

tion and the smartwatch to use in other driving activi-

ties, which is an indication that the proposed tool can

positively contribute to helping drivers with stress and

anxiety.

The remainder of this text is organized as follows.

Section 2 presents details of the proposed tools. Sec-

tion 3 details and analyzes the conducted case study.

Finally, Section 4 presents the ﬁnal considerations

and future works.

2 PROPOSED TOOLS

The proposed approach uses an App on an An-

droid smartwatch, which collects heart rate and GPS

data from the driver during a driving activity, syn-

chronously storing this data in a cloud database. The

app also displays feedback to drivers according to

their heart rate levels 2.1. We also developed a script

that can present summarized reports of the collected

data so that experts can analyze and assist the driver

if necessary. Details about the application developed

are in Subsection 2.2. The feedback provided to the

driver according to heart rate levels is in Subsection

2.3. Subsection 2.4 describes the scripts developed

to produce reports on the collected data. Subsection

2.5 outlines the data visualisation layer and the oper-

ational features of the dashboard.

2.1 Heart Rate Levels

Five stress and anxiety levels were considered based

on the driver’s heart rate. These levels were de-

ﬁned empirically and are used to provide feedback

to drivers. Additionally we established these levels

based on the number of beats per minute (BPM) of

the driver.

The levels of stress and anxiety used are presented

below:

• Z1: BPM lower than 116;

• Z2: BPM between 117 and 140;

A Smartwatch-Based Approach to Support and Analysis of Driver Stress and Anxiety

545

• Z3: BPM between 141 and 158 ;

• Z4: BPM between 159 and 167; and

• Z5: BPM greater than Z4.

Regarding the heart rate levels considered in this

study, it should be noted that there is no universally

accepted standard for comparing heart rate. However,

studies indicate that comparing the heart rate of ex-

perimental participants is a practical approach (Catai

et al., 2020). The heart rate ranges used can be ad-

justed in future research by comparing the BPM col-

lected in experiments.

2.2 Developed App

The application was developed for the Android Oper-

ating System, with the Samsung Health Sensor API

to ensure compatibility with Wear OS. Furthermore,

we used Kotlin

and Java

programming languages.

In addition, for real-time data storage and synchro-

nization, we chose the Firebase database.

Moreover, the MediaPlayer API of the Android

multimedia framework was used to play the audio

ﬁles which are in OGG format. To generate the au-

dio ﬁles, we used the Vidnoz online tool

, that pro-

vides different voice patterns – we chose the pattern

entitled Elza, a friendly and articulate young female

voice in Portuguese.

The following steps are followed for the app use:

1. The user needs to create an account in the app pro-

viding their name and login (Figure 1a).

2. The user must log in to the app (with their user-

name and password) by clicking the Start button

(Figure 1b).

3. If the user is logged into the app, it performs the

following activities:

• The app plays the audio: “Welcome, my name

is Juliana, I’m your virtual assistant” (this

audio is played in Portuguese). The presence of

a name aims to humanize the interaction, mak-

ing the experience more friendly and intimate;

• In intervals of three seconds, the app monitors

the heart rate and GPS coordinates, saving this

data asynchronously in the cloud database;

• The app displays the heart rate, as well as stress

and anxiety levels (Z1 to Z5) on the smartwatch

display (Figure 1c);

https://developer.samsung.com/health

https://kotlinlang.org/

https://www.oracle.com/br/java/

https://pt.vidnoz.com/text-to-speech.html

• Every 6 minutes, the app plays an audio record-

ing of the driver’s heart rate zone (Z1 to Z5).

The feedback texts/audios are in Subsection

2.3;

• If the user is at levels Z4 or Z5, a pop-up is also

displayed on the screen (Figure 1d); if the user

stops the car and presses yes, the unlatching au-

dio is played (described in Subsection 2.3);

• New feedback is presented to the driver every

six minutes, according to their stress level from

Z1 to Z5.

4. When the driving activity ends, the user must

press the stop button so that the app registers the

end of the activity, stopping data collection and

saving it in the database.

a - Login screen b - BPM measuring

c - Real-time heart rate d - Relaxation prompt

Figure 1: App Screens.

2.3 Feedback

The feedback explored is presented in this Subsection.

In this paper, the texts are in English, but as the exper-

iments were carried out with Brazilians, the app was

implemented with the texts in Portuguese (the audios

in Portuguese are in the project’s GitHub repository).

For the lower stress and anxiety zones (Z1 and

Z2), positive and motivational messages were used,

encouraging the user to maintain their state, empha-

sizing pleasure and the experience at the moment.

For the intermediate zone (Z3), the message alerts the

user about the state change and offers a practical so-

lution (deep breathing) to help control anxiety. For

the higher zones (Z4 and Z5), the message provides

instructions on breathing deeply to help the user re-

lax physically and mentally. The use of counting dur-

ing deep breathing helps to reinforce the practice and

maintain focus on the process. It is worth noting that

WEBIST 2025 - 21st International Conference on Web Information Systems and Technologies

546

breathing techniques in times of stress and anxiety are

commonly mentioned in the literature (Leyro et al.,

2021)

The texts used in the feedback audios are pre-

sented below:

• Feedback for Z1 level: “You are relaxed. Very

good! Great job.”

• Feedback for Z2 level: “You’re cool, that’s great!

Enjoy the ride and enjoy the moment.”

• Feedback for Z3 level: “You are entering a state

of anxiety. How about relaxing with deep breath-

ing? You can continue driving, but at this mo-

ment, take some deeper breaths; it will help you.”

• Feedback for Z4 level: “You’re in a state of

heightened anxiety. How about taking a break?

Stop the vehicle and let me know when you’re in

a safe place.”

• Feedback for Z5 level: “You are in a state of very

high anxiety. I advise you to stop the car as soon

as possible to begin the relaxation. You must stop

the vehicle and let me know when you are in a

safe place to start the relaxation.”

If the user is in the highest levels of stress/anxiety

(Z4 or Z5), when he/she presses yes on the app (with

the car stopped), he/she will receive the following

feedback: “Now we are going to start the relaxation

process, get into a comfortable position. You will in-

hale, ﬁll your lungs and exhale, releasing all the air

very deeply and slowly for ﬁve seconds. Let’s go.

One, two, three, four, ﬁve. Exhale one, two, three,

four and ﬁve. Excellent job; you can repeat this relax-

ation for 2 minutes or 5 times. With this, we ventilate,

strengthen the brain and have a more productive train-

ing session.”

2.4 Report

To generate a mapped report of the heart rate data

collected during each practical driving activity, we

created a Python code using GeoPandas

and the

Folium

libraries to visualize geospatial data interac-

tively using Google Maps features.

Our tool connects to the cloud database, allowing

us to obtain the data recorded by each driver during a

driving activity. The entire route is drawn on the map

based on the stored GPS coordinates, using markers

with different colors for the moment the user is in

each stress zone. The following markers colors were

used: level Z

in green, level Z

in beige, level Z

orange, level Z

in pink and level Z

in red.

https://geopandas.org/

https://python-visualization.github.io/folium/

Figure 2 shows an example of a report produced

for a driver who passed through all the deﬁned stress

zones during a driving activity. In this report, the user

can interact with the map by moving it and zooming

in and out.

Figure 2: Reports produced.

It is important to note that these reports are im-

portant for specialists to analyze the driver’s stress

and anxiety levels in driving conditions, providing

them with adequate support. The reports can be is-

sued while the user is driving, and accessed by a pro-

fessional in any other location with internet access.

If necessary, this professional can interact with the

driver immediately after the driver has ﬁnished driv-

ing or even during the driving activity (in extreme

cases).

2.5 Data Processing and Visualization

In addition to processing heart rate and geolocation

data, the system incorporates a more robust visualiza-

tion module that generates interactive reports to illus-

trate the emotional intensity experienced during each

driving session. These reports function as both ped-

agogical and psychological tools, supporting instruc-

tors and learners in identifying and reﬂecting on emo-

tionally signiﬁcant moments throughout the learning

process.

The visualization tool uses geospatial plotting

with color-coded stress indicators based on heart rate

variability. Developed with responsive web technolo-

gies, it enables real-time and post-session access via

desktop and mobile devices, promoting intuitive and

user-centered interaction.

By combining wearable data with accessible inter-

faces, the system supports emotional self-regulation

and learning in dynamic driving scenarios.

A Smartwatch-Based Approach to Support and Analysis of Driver Stress and Anxiety

547

This two-layer (desktop + mobile) architecture

enables the full cycle of monitoring → interpreta-

tion → intervention (Marceglia et al., 2015). Desk-

top dashboards address the longitudinal and multi-

user supervision needs of the professionals, whereas

the mobile app priorities real-time feedback and self-

regulation—an allocation pattern recommended in

multi-device human-computer interaction literature

(Grubert et al., 2016). The approach leverages ﬁnd-

ings that wearable biofeedback lowers acute driving

stress (Iqbal et al., 2022; Liu et al., 2023) and aligns

with recent systematic reviews on stress-aware inter-

action design (Perez and Ruiz, 2024).

The desktop interface (Figure 3), shown below,

was designed to allow specialists to monitor student

performance through real-time reporting and post-

session review of physiological and location-based

metrics.

Figure 3: Desktop visualization.

The mobile version (Figure 4) further extends ac-

cessibility and interactivity with the feedback in a

simpliﬁed and responsive environment, even during

the driving session itself.

3 CASE STUDY

To validate the proposed tools, we conducted a case

study using the hardware described in Subsection 3.1,

which was applied in a pilot case study with eleven

student drivers from a Driver Training Center (DTC),

as described in Subsection 3.2. Supplementary mate-

rial in Subsection 3.3 was provided in advance to fa-

miliarise participants with the underlying techniques

and their intended applications. To evaluate the expe-

rience using the proposed solution, we applied ques-

tionnaires to the students (drivers), DTC employees

and a psychologist specializing in drivers who fear

driving. The experiment’s results are in Subsection

3.4.

Figure 4: Mobile visualization.

We chose to run the experiments in partnership

with a DTC, because the audience of participants

in the experiment includes novice drivers (student

drivers), who may have high levels of stress and anxi-

ety due to their inexperience. Contributing to improv-

ing the training of these drivers can result in better

drivers who pass the ﬁnal exam to obtain their licenses

more quickly. Apart from that, the Ethics Commit-

tee for Research with Humans previously approved

all procedures performed.

3.1 Hardware Conﬁguration

We used a Samsung Galaxy Watch 6 smartwatch,

which was released in 2023, featuring the Wear OS

4 operating system with the One UI Watch 5 interface

while offering an integrated experience with Android

devices. This smartwatch has 2 GB of RAM, 16 GB

of storage and a 5-nanometer Exynos W930 proces-

sor. Figure 5 shows a photo of this smartwatch used

to run our application.

It is worth highlighting two possibilities for con-

necting to the internet with the smartwatch used: i)

via Bluetooth/Wi-Fi, being able to obtain the internet

routed from a nearby smartphone, and ii) its LTE ver-

sion, which allows direct connection to the internet

using an eSIM (Embedded SIM) without the need of

a smartphone, which was the option used in the ex-

periments.

WEBIST 2025 - 21st International Conference on Web Information Systems and Technologies

548

Figure 5: Smartwatch running the proposed app.

3.2 Case Study Steps

To perform the experiments, we executed the follow-

ing steps:

1. All students enrolled in the DTC who were taking

practical driving lessons were invited to partici-

pate in the experiment.

2. We presented all methodological procedures to

students interested in participating in the experi-

ment, explaining that they could withdraw from

participating at any time.

3. We consulted the DTC instructors to identify

which students, among those who expressed inter-

est in participating in the experiment, were most

likely to exhibit signs of stress and anxiety. Based

on their experiences in previous classes, the in-

structors identiﬁed eleven students.

4. Each selected student had to complete a free

and informed consent form, as established by the

ethics committee.

5. Participants were registered in the App.

6. Participants put the watch on their wrists, logged

into the app and started the driving lessons.

7. Participants completed their class with the watch

on their wrists, accompanied by a DTC instructor

in the DTC vehicle. Each class lasted an average

of 40 minutes of driving time.

8. At the end of the class, the students and the in-

structor answered a questionnaire with quantita-

tive questions about their experience with the app.

9. A psychologist specializing in helping drivers

with fear of driving accompanied the entire pro-

cess, answering a questionnaire about the exper-

iments and evaluating the reports produced with

the data collected (Subsection 2.4).

3.3 Supplementary Material

In addition to the audio guidance provided during mo-

ments of high stress (zones Z4 and Z5), the software

incorporates pre-training content based on mindful-

ness principles. Before the practical driving session

begins, the student is given access to visual mate-

rial with simple instructions on how to manage anx-

iety through techniques such as focused breathing,

observing the environment, and accepting thoughts

without judgment. This preparation aims to raise the

student’s awareness of the importance of emotional

balance even before a stressful situation occurs.

This prior resource is intended to make the real-

time feedback more effective. When the system

prompts the student to stop the vehicle and begin the

relaxation process, they are already familiar with the

strategies to be used, which increases engagement and

receptiveness to the exercise.

The preparatory content shown to the student be-

fore the driving session is based on simple and ef-

fective mindfulness strategies. Figure 6 presents the

content used as an introductory mindfulness guide to

the student before the start of the driving session. The

guidance provided includes:

• Present Moment: Be aware of the current situa-

tion.

• Observe Your Environment: Notice sounds,

smells, and physical sensations.

• Focus on Breathing: Focus on your breathing to

calm yourself.

• Acceptance Without Judgment: Accept

thoughts and emotions without criticism.

Repeat a few times to reduce symptoms of anxiety

and stress!

Figure 7 presents the diaphragmatic breathing

content, which is presented to drivers and instructors,

with the following step:

• Place one hand on your chest and the other on

your abdomen.

• Inhale deeply for 3 seconds.

• Hold your breath for 3 seconds.

• Exhale through your mouth for 6 seconds.

The user is encouraged to repeat the exercise a few

times to reduce symptoms of anxiety and stress.

3.4 Results

The eleven student drivers who participated in the pi-

lot study completed all the experiment activities and

A Smartwatch-Based Approach to Support and Analysis of Driver Stress and Anxiety

549

Figure 6: Introductory mindfulness guide.

Figure 7: Diaphragmatic breathing guide.

reported no problems wearing the watch. A question-

naire was administered to all participants at the end

of the driving activity. The questionnaire, alternatives

and participant responses are presented below:

1. Is the interface intuitive and easy to use?

• Yes: 11 answers (100%).

• No: 0 answers (0%).

2. Did the software help reduce your stress while

driving?

• Yes, a lot: 7 answers (63.6%).

• Yes, a little: 2 answers (18.2%).

• It made no difference: 2 answers (18.2%).

• No, it increased stress: 0 answer (0%).

3. Was the sound alert helpful in reminding you

to manage your stress?

• Yes, it was very helpful: 8 answers (72.7%).

• Yes, but it could be adjusted: 1 answer (9.1%).

• It made no difference: 2 answers (18.2%).

• I found it irritating: 0 answer (0%).

4. Was alerts frequency appropriate for you?

• Yes: 9 answers (81.8%).

• It is necessary to reduce the frequency of alerts;

they were very frequent: 1 answer (9.1%).

• It is necessary to increase the frequency of

alerts, there should be more: 1 answer (9.1%).

• I didn’t notice the warnings: 0 answer (0%).

5. Did the software help you better understand

your stress levels while driving?

• Yes, completely: 8 answers (72.7%).

• Yes, partially: 0 answer (0%).

• It made no difference: 3 answers (27.3%).

6. Did you experience any discomfort with the

software? If so, what was it?

• No, I didn’t feel any discomfort: 9 answers

(81.7%).

• The sound of the alerts was unpleasant: 1 an-

swer (9.1%).

• The alerts were very frequent: 1 answer (9.1%).

WEBIST 2025 - 21st International Conference on Web Information Systems and Technologies

550

7. Would you use this software regularly in your

driving lessons?

• Yes: 9 answers (81.7%).

• No: 1 answer (9.1%).

• Maybe: 1 answer (9.1%).

Survey results indicate that the app has an intuitive

and easy-to-use interface, with 100% approval rat-

ings. Most participants noted a signiﬁcant reduction

in stress while driving and found the audible alerts

useful, although a small percentage suggested adjust-

ments. Overall acceptance was high, with 81.7% say-

ing they would use the software regularly, despite

minor complaints about the sound and frequency of

alerts.

In addition to the questionnaire, qualitative feed-

back was collected after each practical driving ses-

sion monitored by the app developer, involving stu-

dent drivers, DTC instructors, and the psychologist

overseeing the experiment. According to the opin-

ion of the six instructors who monitored the practical

driving activities, with the use of the app, a slight in-

crease in the students’ conﬁdence could be observed,

presenting a positive impact on the learning experi-

ence of the students who did not experience any dis-

comfort with the use of the app. The instructors also

provided suggestions regarding the feedback feature,

noting that its text could be improved and that it was

triggered too frequently within a short period.

The specialist psychologist highlighted the im-

pacts of the software on students’ conﬁdence, indi-

cating that this is an extremely important factor since

emotional security is an essential factor for effective

learning to drive a vehicle. However, both the instruc-

tors and the psychologist point out the need to reeval-

uate the stress zones based on the data collected in the

experiments. Besides that, they also suggested testing

other patterns of sound alerts.

Regarding the students/drivers, it should be noted

that none of the students entered zones Z4 and Z5 dur-

ing the pilot case study. However, they still reported

feeling less anxiety/stress when using the application,

reporting that upon receiving feedback about their

emotional state through the software, they felt more

comfortable and demonstrated greater conﬁdence in

driving. It is also important to mention that 2 of the 11

drivers showed interest in purchasing the watch and

the application to continue using it in other classes.

New case studies are necessary with a broader audi-

ence, but the pilot study results highlight the impor-

tance of the proposed solution.

4 CONCLUSION

This study introduced a smartwatch-based solution

for monitoring and analyzing driver stress and anxi-

ety levels using heart rate data without requiring ad-

ditional synchronized devices. The proposed system

enables real-time data collection and feedback, assist-

ing drivers in managing their emotional states while

driving. The collected data is also stored in a cloud

database, allowing experts, such as psychologists, to

analyze reports and provide personalized support.

The pilot study was conducted with driving school

students, demonstrating the system’s feasibility and

potential to improve driver well-being. The positive

reception of the application among participants, in-

cluding two individuals who expressed interest in us-

ing it beyond the study, suggests that the tool can

contribute to safer and more conﬁdent driving expe-

riences, particularly for novice drivers.

Enabling data analysis about the driver with re-

ports is also one of the contributions of this work. In

particular, the maps produced can be a valuable tool

for specialized psychologists. Still, it is worth noting

that new functionalities should be implemented in fu-

ture work to improve the software that produces these

reports, allowing the analysis of drivers’ performance

on different days/moments.

Despite these promising outcomes, some limita-

tions must be addressed in future work. The current

study involved a small sample size, and further re-

search with a more extensive and diverse group of

participants is necessary. The tool can be interactively

adjusted based on the feedback received by experi-

menting with a larger group of drivers. It is worth not-

ing that it is crucial to conduct experiments not only

with student drivers but also with experienced drivers

who report high levels of stress and anxiety.

In future work, in addition to adjusting the func-

tionalities of the proposed tools, it is possible to ex-

plore other low-cost wearable devices that collect

other physiological signals, such as galvanic skin re-

sponse and respiration rate, to improve accuracy. Fu-

ture developments could also explore machine learn-

ing techniques to enhance the detection and classiﬁ-

cation of emotional states in real-time, aiming to pre-

dict the moments in which drivers will enter high-

stress and anxiety zones, and providing feedback in

advance.

ACKNOWLEDGMENTS

We would like to thank the Coordination for the Im-

provement of Higher Education Personnel (CAPES) -

A Smartwatch-Based Approach to Support and Analysis of Driver Stress and Anxiety

551

Program of Academic Excellence (PROEX) for the

ﬁnancial support provided through the scholarship

grant. We are also grateful to the Driver Training Cen-

ter Auto Escola Mil

enio, where the experiments were

conducted, and to psychologist Juliana Daga for her

continuous support and involvement throughout the

development and implementation of the project.

REFERENCES

Alhaek, F., Liang, W., Rajeh, T. M., Javed, M. H., and Li,

T. (2024). Learning spatial patterns and temporal de-

pendencies for trafﬁc accident severity prediction: A

deep learning approach. Knowledge-Based Systems,

286:111406.

Barka, R. E. and Politis, I. (2024). Driving into the future: A

scoping review of smartwatch use for real-time driver

monitoring. Transportation research interdisciplinary

perspectives, 25:101098.

Braun, M., Schubert, J., Pﬂeging, B., and Alt, F. (2019).

Improving driver emotions with affective strategies.

Multimodal Technologies and Interaction, 3(1):21.

Catai, A. M., Pastre, C. M., de Godoy, M. F., da Silva, E.,

de Medeiros Takahashi, A. C., and Vanderlei, L. C. M.

(2020). Heart rate variability: are you using it prop-

erly? standardisation checklist of procedures. Brazil-

ian journal of physical therapy, 24(2):91–102.

Craske, M. G., Kircanski, K., and Vervliet, B. (2009). Fear

of driving: Epidemiology, diagnosis, and treatment.

Oxford University Press.

Grubert, J., Kranz, M., and Quigley, A. (2016). Challenges

in mobile multi-device ecosystems. mUX—The Jour-

nal of Mobile User Experience, 5(1):1–17.

Haque, Y., Zawad, R. S., Rony, C. S. A., Al Banna, H.,

Ghosh, T., Kaiser, M. S., and Mahmud, M. (2024).

State-of-the-art of stress prediction from heart rate

variability using artiﬁcial intelligence. Cognitive

Computation, 16(2):455–481.

Iqbal, T., Simpkin, A. J., Roshan, D., Glynn, N., et al.

(2022). Stress monitoring using wearable sensors:

A pilot study and Stress-Predict dataset. Sensors,

22(21):8135.

Lee, M., Lee, S., Hwang, S., Lim, S., and Yang, J. H.

(2023). Effect of emotion on galvanic skin response

and vehicle control data during simulated driving.

Transportation research part F: trafﬁc psychology and

behaviour, 93:90–105.

Leyro, T. M., Versella, M. V., Yang, M.-J., Brinkman,

H. R., Hoyt, D. L., and Lehrer, P. (2021). Respiratory

therapy for the treatment of anxiety: Meta-analytic

review and regression. Clinical psychology review,

84:101980.

Li, W., Zhang, B., Wang, P., Sun, C., Zeng, G., Tang,

Q., Guo, G., and Cao, D. (2021). Visual-attribute-

based emotion regulation of angry driving behaviors.

IEEE Intelligent Transportation Systems Magazine,

14(3):10–28.

Lin, W. and Li, C. (2023). Review of studies on emotion

recognition and judgment based on physiological sig-

nals. Applied Sciences, 13(4):2573.

Liu, K., Jiao, Y., Du, C., et al. (2023). Driver stress detec-

tion using ultra-short-term HRV analysis under real-

world driving conditions. Entropy, 25(2):194.

Marceglia, S., Fontelo, P., Rossi, E., and Ackerman,

M. J. (2015). A standards-based architecture pro-

posal for integrating patient mhealth apps to electronic

health record systems. Applied Clinical Informatics,

6(3):488–505.

Marquart, G., Cabrall, C., and De Winter, J. (2015). Review

of eye-related measures of drivers’ mental workload.

Procedia Manufacturing, 3:2854–2861.

Minist

erio dos Transportes (2023). Painel es-

tat

ıstico de acidentes de tr

ansito. https:

//www.gov.br/transportes/pt-br/assuntos/transito/

arquivos-senatran/docs/renaest. Acessado em

11/11/2023.

Ortoncelli, A. R., Silva, L., Bellon, O. R. P., de Oliveira,

T. M., and Daga, J. (2020). Summarizing driving be-

havior to support driver stress analysis. In 2020 15th

IEEE International Conference on Automatic Face

and Gesture Recognition (FG 2020), pages 587–591.

IEEE.

Perez, M. d. l. C. and Ruiz, J. A. (2024). Detection and

monitoring of stress using wearables: A systematic

review. Frontiers in Computer Science, 6:1478851.

Pinge, A., Gad, V., Jaisighani, D., Ghosh, S., and Sen,

S. (2024). Detection and monitoring of stress using

wearables: a systematic review. Frontiers in Com-

puter Science, 6:1478851.

Stephens, A., Collette, B., Hidalgo-Munoz, A., Fort, A.,

Evennou, M., and Jallais, C. (2024). Help-seeking for

driving anxiety: who seeks help and how beneﬁcial is

this perceived to be? Transportation research part F:

trafﬁc psychology and behaviour, 105:182–195.

Vansteenwegen, A., Verhaeghe, S., Clercq, C., and Kir-

canski, K. (2014). The impact of driving anxiety on

mobility, quality of life and socio-economic participa-

tion: A systematic review. Clinical Psychology Re-

view, 34(8):673–688.

WHO (2023). Global status report on road safety

2023. http://www.abeetrans.com.br/abeetrans/?p=

4213. Acessado em 20/11/2023.

Xu, G., Qin, R., Zheng, Z., and Shi, Y. (2024).

An adaptive system for wearable devices to detect

stress using physiological signals. arXiv preprint

arXiv:2407.15252.

Zhao, Q., Yang, L., and Lyu, N. (2024). A driver stress

detection model via data augmentation based on deep

convolutional recurrent neural network. Expert Sys-

tems with Applications, 238:122056.

WEBIST 2025 - 21st International Conference on Web Information Systems and Technologies

552