Construction of a Virtual Environment to Measure the Evolution of

Kendo Athletes

Francisco Marcelino A. de Ara

ujo

1,5,6,7 a

, Antonio Kau

e C. Ferreira

1 b

, Matheus Araujo Dantas

1 c

Halyson Itallo C. Pimentel

1 d

, Paulo Roberto A. Leal

1 e

, S

ergio Lu

ıs B. de Carvalho

2 f

N. M. Fonseca Ferreira

3,4 g

, Ant

onio Valente

3,5 h

and Salviano F. S. P. Soares

5,6,7 i

LABIRAS, Federal Institute of Piau

ı, Teresina, Brazil

LABIRAS, Federal University of Piau

ı, Teresina, Brazil

INESC-TEC, Technology and Science, Porto Portugal

ISEC, Institute of Engineering of Coimbra, Coimbra, Portugal

University of Tr

as-os-Montes and Alto Douro, School of Sciences and Technology, Engineering Department,

Vila Real, Portugal

Intelligent Systems Associate Laboratory (LASI), Portugal

Institute of Electronics and Informatics Engineering of Aveiro (IEETA), University of Aveiro, Aveiro, Portugal

Keywords:

Reaction Time, Sports, Virtual Reality, Kendo.

Abstract:

The use of technology applied in sports comes each year becoming a great tool to help athletes train. Moreover,

the post-pandemic world is undergoing dramatic changes in the way of thinking and acting, with new ways of

exercising emerging, but without leaving home. Thus this paper describes the development of a platform for

training, focusing on Kendo practitioners (Japanese fencing) using virtual reality tools to allow athletes and

training the distance. Through the use of a HMD (Head Mounted Device), kend

okas will be able to practice

blows and improve their reﬂex by a gamiﬁed experience in a virtual environment.

1 INTRODUCTION

The application of technology in sports is an area that

has seen constant renovation in the past few years and

has since become a great auxiliary tool in the training

of athletes. In relation to this market, some of the

technologies used include computer modeling, data

acquisition and analysis, mobile computers and in-

formation technology networks (Baca et al., 2009).

Currently, high-level athletes have adopted the use of

technological devices that collect data on their phys-

ical and mental abilities, with the aim of improving

their abilities and sports performance (Liebermann

et al., 2002) (Petri et al., 2018).

https://orcid.org/0000-0001-8928-0077

https://orcid.org/0009-0004-1869-4984

https://orcid.org/0000-0002-5718-9137

https://orcid.org/0000-0003-4019-6848

https://orcid.org/0000-0002-6173-3123

https://orcid.org/0009-0000-5651-1801

https://orcid.org/0000-0002-2204-6339

https://orcid.org/0000-0002-5798-1298

https://orcid.org/0000-0001-5862-5706

Among the technical resources used in sports

training, virtual reality (VR) is a technology capa-

ble of imitating real scenarios in a way that seems

familiar to human cognition, creating a virtual envi-

ronment. The user is inserted into this environment

and their movements are represented virtually, caus-

ing an enhanced feeling of immersion, a perfect setup

for creating training environments.

Martial arts and combat sports can also beneﬁt

greatly from this kind of technology. Kendo, a martial

art of Japanese origin and the main style represented

in this paper, can be described as a match between

two opponents. Its setup resembles Olympic fenc-

ing, however practitioners wield bamboo swords and

wear a special set of armor and clothing that is settled

down by traditional customs. Moreover, Kendo is a

sport that has high-performance, world-class cham-

pionships, but a very small amount of research in the

West about it, mainly in the ﬁeld of technology. When

compared to Football, which is an Olympic sport and

relies on generous technological investments, it is es-

timated that Kendo has a great deﬁcit in scientiﬁc

works devoted to it. Even so, in the years 2016 to

2020, there has been a remarkable increase in experi-

162

A. de Araújo, F., Ferreira, A., Dantas, M., Pimentel, H., Leal, P., B. de Carvalho, S., Ferreira, N., Valente, A. and Soares, S.

Construction of a Virtual Environment to Measure the Evolution of Kendo Athletes.

DOI: 10.5220/0012189000003587

In Proceedings of the 11th International Conference on Sport Sciences Research and Technology Support (icSPORTS 2023), pages 162-168

ISBN: 978-989-758-673-6; ISSN: 2184-3201

ments that employ technology to improve the training

of Kendo practitioners (also regarded as “kendoka”).

Some of the themes found were a study of the kine-

matics and dynamics of Kendo kata (Konyukhov and

Yeniavci, ), a study to determine the factors to reduce

the attack time of the men (Murase et al., 2017), the

development of a “kote” Kendo glove to analyse pa-

rameters (Jeong et al., 2018), a Kendo support system

(Takata et al., 2019), and detection of attack activity

(Torigoe et al., 2020). Most of the recent articles are

related to the ﬁeld of health sciences or the study of

Kendo as a martial art in general.

In view of this, virtual reality will be used to carry

out movement execution training for Kendo athletes.

The use of this technology in sports is applied in dif-

ferent modalities, several studies have already tested

and observed whether the use of virtual reality in

sports helps the practitioner’s performance, conclud-

ing that the use of this technology in parallel with

the usual physical activity is beneﬁcial for the practi-

tioner. (Neumann et al., 2018),(Stinson and Bowman,

2014),(Bideau et al., 2009).

2 JUSTIFICATION

Since the start of the COVID-19 pandemic, the public

was faced with an increasing need to ﬁnd new ways

to exercise, since gyms were temporarily closed all

over the world. In addition to that, due to safety

precautions, a substantial portion of the population

had to stay in their homes. Ahead of this situation,

the Brazilian Kendo Confederation (Confederac¸

Brasileira de Kendo – CBK) prepared a calendar of

online events for training and refereeing directed at

the Brazilian public, highlighting the ”Online Chil-

dren’s Suburi Championship” of Kendo Jidai Interna-

tional magazine, which had the participation of ath-

letes from all over the world. This event united sports

practitioners in an extremely difﬁcult time for human-

ity. In this perspective, in the post-pandemic context,

the practice of physical exercises with Internet tools

remains a widely adopted modality, becoming a new

way of practicing exercises and sustain social contact,

albeit remotely. Thus, performing these activities in

online settings facilitates training and makes knowl-

edge in different areas of Kendo much easier to share.

Following this line of reasoning, it seems fair to

state that people in the whole world should reconsider

the conditions in which, until quite recently, many

commonplace activities, such as physical exercise,

occurred. When governing bodies of many countries

throughout the world understood that the institution

of ”lockdown” measures was necessary, it suddenly

seemed clear that health authorities needed to come

up with public health strategies that must include the

creation and implementation of interventions in pop-

ular habits and customs, with a focus on promoting

safe physical activity and reducing sedentary lifestyle.

With these measures in mind, it is possible to pre-

vent major health crises should similar extreme situa-

tions be experienced by humanity ever again (Stock-

well et al., 2021).

Given the situation, virtual reality resources, espe-

cially Head Mounted Display (HMD) devices, would

offer an immersive and realistic environment. In con-

trast to the real environment, these devices allow free

manipulation, which provides greater autonomy for

instructors and developers, in addition to creating fa-

vorable frameworks for the progress of athletes. Con-

sequently, virtual reality technology helps improve

the accuracy of strikes and reduce the reaction time of

athletes, focusing on the growth and improvement of

Kendo practitioners, preparing them for competitions.

Furthermore, it is imperative that this technology can

be distributed easily and quickly, aiming at its dissem-

ination among different dojos and isolated practition-

ers, making it an integral part of the kendokas training

routine.

3 THEORETICAL REFERENCE

3.1 Kendo

Kendo, often translated as ”way of the sword”, is a

martial art of Japanese origin in which practitioners

(kend

oka) learn to ﬁght with bamboo swords (shinai)

while wearing traditional clothing and protective ar-

mor (kend

ogu or b

ogu) . It is also an elegant tradi-

tional combat sport that can be enjoyed from early

childhood well into late seniority. It emphasizes cor-

rect execution of techniques as well as correct de-

meanor from teachers and students alike, nurturing a

prospering environment of healthy and safe exercise

based on mutual respect and fair-play, owing much

to its martial origins and sportive leanings. Accord-

ing to the All Japan Kendo Federation (AJFK), the

concept and purpose of kendo are as follows: con-

cept - ”Kendo is way of disciplining human character

through the application of katana principles”; purpose

- ”To shape mind and body. To cultivate a vigorous

spirit and, through correct and rigorous training, to

strive to import the art of Kendo. To have esteem for

human courtesy and honor. To associate with other

in sincerity. And always seek self-cultivation. Thus

a person will be able to love his country and society;

contribute to the development of culture; and promote

Construction of a Virtual Environment to Measure the Evolution of Kendo Athletes

163

peace and prosperity among all peoples” (Salmon,

2013).

Futhermore, Kendo’s list of valid strikes is com-

posed of four basic attacks that can be used to score

points in a match. They are: men (attack to the top

of the head), kote (attack to the forearm), d

o (attack

to the torso) and tsuki (throat thrust). All of these at-

tacks are made on areas of the armor the competitors

wear, and are usually named after the part of the ar-

mor that is being hit. Competitors advance in a match

if they score two valid points (datotsu or ippon) that

are considered satisfactory by the three standing ref-

erees, or if they score one point and time runs out, or

yet if their adversary is disqualiﬁed for receiving two

fouls (hansoku).

Besides the competition setting, Kendo training

involves basic body exercise focused on developing

good movement and coordination, traditional exer-

cises such as suburi (swinging the sword reapeat-

edly, as if cutting the air), kirikaeshi (repeated cuts

or strikes in a set pattern against a training partner),

uchikomi keiko (training of strikes as allowed in com-

petition) and kata (predeﬁned forms of both classical

and modern inﬂuence that represent the most signiﬁ-

cant aspects of Japanese swordsmanship).

3.2 Virtual Reality in Training

Technology has become an amazing tool for high-

performance sports since its use in training for com-

petition comprehends much of the new methods for

improving athletes, especially in sports centered on

the constant monitoring of the evolution of their prac-

titioners (Fleming et al., 2010). Among the vari-

ous technologies used for this purpose, virtual reality

stands out for ”transporting” the athlete into the pos-

sible scenario that will be faced by him in the day of

the competition, bringing him closer to the real situa-

tion. In addition to this, there is also the possibility of

inserting unexpected variables that cannot be repro-

duced so easily in real-life training, enabling the ath-

lete to be prepared for any type of unforeseen event

that may occur during competition (Wang, 2012).

In view of this, athletes and technical committees

can take advantage of virtual scenarios and position-

ing technology as a sort of training partner, aiming at

improving and monitoring the evolution and perfor-

mance of practitioners, thus improving their skills and

allowing better performance in competitions (Craig,

2013).

Kendo, as a martial art and combat sport, requires

a constant training routine from the practitioner, as the

score depends on the execution of the strikes and the

accuracy of each blow against the opponent. In this

sense, similar to the present work, there is an article

that aims to use immersive virtual reality to improve

the training of Karate athletes, aimed at improving

their CRT. However, it seems that the way in which

the VR technology is used in that trial ends up mak-

ing it impractical for use in other Karate commissions,

as it does not use an standalone device. What that

means is that, for the project to work, it is necessary

to connect the VR device to a computer, making it de-

pendent of another component to work, reducing its

portability and ease of use (Petri et al., 2019).

4 MATERIALS AND METHODS

The choice for the HMD for this project was deter-

mined by the immersiveness that the equipment pro-

vides to the user. It also provides movement monitor-

ing capabilites and has several integrated sensors, in

this sense bringing the virtual environment closer to

the athlete, making the experience even more immer-

sive. Besides, it is a stand-alone device, which means

it does not require a cable connection with a com-

puter. It is a virtual world mirrored in the real world,

helping with the proximity of movement and train-

ing space. According to Doctor Cathy Craig (Craig,

2013) the structure of the HMD contains an inside-

out tracking technology, using a headset and cameras

while connected to a computer, without the need to

organize external towers in the room, since the Meta

Quest 2 allows the user to deﬁne the shape and size of

the limits of his virtual space.

For the development of this work, the hardware

device used was the Meta Quest 2. For the software

part, the Unreal Engine version 4.27, together with

the HTTP Request For Blueprints Plugin, a Web API

using Express framework and a Postgresql database.

The Meta Quest 2 was chosen in this work due to the

immersiveness that it provides since it allows follow-

ing the athlete’s movements within the virtual envi-

ronment, beyond the fact it is a stand-alone device,

which results in more freedom of movement for the

athlete. Unreal Engine 4 is a 3D application develop-

ment platform designed by Epic Games that offers a

wide range of tools that enable the creation of cutting-

edge content, interactive experiences, and immersive

virtual worlds. The version 4.27 of Unreal Engine

was chosen instead it latest launched version, the Un-

real Engine 5, its due the fact that at the beginning

of project development, the Unreal Engine 5 was not

fully stable, and also presented errors at Android OS

exporting. Although this engine is geared towards

games, its application goes far beyond this area. Thus

having prominence in research and interactive proto-

icSPORTS 2023 - 11th International Conference on Sport Sciences Research and Technology Support

164

typing. Among the options that Unreal Engine 4 al-

lows us, those that were used by the team are support

for VR, importing 3D models, remote control proto-

col and external plugins, which can be found on the

ofﬁcial Unreal Engine 4 website.

HTTP Request For Blueprints is a plugin devel-

oped by King Wai Mark that makes it possible to per-

form REST requests to a server efﬁciently, used to

communicate between the virtual environment and a

remote API. To create a project with VR interaction,

we used a template from Unreal Engine 4 called “Vir-

tual Reality.” This template provides a character who

has already implemented the movement mirrored with

reality.

With this character in the project, we created a

new map where the virtual training environment will

be created. To create the environment, we used an as-

set package from Synty Store called Samurai Pack.

We also used a shinai 3D asset model, a bamboo

sword used in Kendo, to be handled by the player,

as shown below in the ﬁgure 1.

Figure 1: Holding Shinai.

The shinai was programmed to have three col-

lision areas, as shown in the ﬁgure 2, representing

the three shinai parts: bo, satsu and sei. Depending

on the shinai area hit, the score changes. Addition-

ally, a socket has been added to the base of the shi-

nai’s skeletal mesh, to ensure that whenever the player

uses the gripping action on any part of the shinai, the

base is ﬁxed to the player’s hand, ensuring a correct

grip. However, a double grip was not implemented,

since the player is not holding a unique equipment,

the hands movement is totally free, which could con-

fuse the player when trying to use both hands grip on

shinai.

The training protocol was implemented placing

targets to be hit on a samurai 3D model, in the po-

sitions of the men, kote and do strikes, respectively,

the head, forearm and tummy positions, as shown in

Figure 2: Shinai areas.

the ﬁgure 3. As shown in the ﬁgure below, the targets

are highlighted areas over the samurai. These targets

light up one at a time, as well the samurai animation,

indicating where the athlete should hit simulating an

opening for a hit. Furthermore, a timer is shown to

indicate how much time the user is taking to hit the

area.

Figure 3: Highlighted area in red.

For the targets to start lighting, the athlete must

maintain a distance from the target similar to the ﬁght-

ing distance between kendokas in real life, which is

represented by a cylinder highlighted area. In Kendo,

all strikes are trained by both beginners and seniors,

with the exception of the Tsuki that only more seniors

practice because of its risk.

To ensure that the strike will be executed correctly,

we use a collider above the athlete’s head, which ver-

iﬁes if the athlete has raised the shinai to the cor-

rect position to execute the strike. Once the athlete

reaches that position, the message signals the possi-

bility of executing the strike. After the execution of

the strike, the highlight goes out and a short audio

track is played, representing audible feedback for the

user to know if the hit was successful or not, and the

samurai animation returns to the initial position, then

the reaction time is computed between the moment

Construction of a Virtual Environment to Measure the Evolution of Kendo Athletes

165

Figure 4: Initial position.

when the target lit and the end of the execution of the

strike. This process is repeated a predetermined num-

ber of times during a training session.

Figure 5: Message indicating permission to attack.

At the end of the session, we generate a JSON

with all the data acquired, storing the athlete id, the

session date, duration, score and information from all

the executed strikes, saving the shinai area hit, the

samurai part hit and the reaction time necessary for

the execution of that strike. After all that, the data of

the executed strikes are sent to a remote API through

the HTTP Blueprint plugin, to store the information

about the user training session on a database.

For the development of this API, the Node js was

used together with the Express framework and Prisma

Object-Relational Mapping (ORM) to develop a sys-

tem that can work online. Using Node js, Express

framework and Prisma ORM, you can create an API

that can communicate with a database, which allow

us to create, update and read data from it. This means

that we can abstract real data in a way that the system

understands and is free to manipulate. The Prisma

ORM is responsible for designing the entities, which

Figure 6: JSON data format.

are abstracted to objects in a model. Therefore, we

can think of a training session as an object model that

has relevant data. This object would have, for exam-

ple, an identiﬁcation number, the strokes given by the

student, his reaction time, among other relevant data.

Like the session, other entities need to be abstracted

before they can be stored. They also need to be related

to the aspect of athlete training. They are:

• User

• User Session

• User Session Statistics

In Node, to implement this abstraction, you need

to create a model for each of these entities. With

the models created in the ’schema.prisma’ ﬁle, Prisma

ORM itself can generate the database for immediate

use, by using a tool called Prisma Migrate. Models,

like their real entities, have relationships with each

other. For example,

An athlete can do a series of training sessions. A

training session is done when an athlete performs a

series of strikes. All of these relationships must be

represented and implemented according to real cases.

These relationships can be more easily seen in the

class diagram:

Now having all the models placed and associated

as necessary, you must now create and manipulate this

data. It is at this point that Unreal’s HTTP Request for

Blueprints plugin is used to make requests to the API,

as in the representation below:

Now with the data processed and sent to the API,

it is possible to observe, based on this database, some

aspects of the athlete’s performance in each training

and its progression.

icSPORTS 2023 - 11th International Conference on Sport Sciences Research and Technology Support

166

Figure 7: Class Diagram of API Models.

Figure 8: HTTP Blueprint Node.

5 CONCLUSION

The aims of this paper include the development of

a training platform focused on improving the perfor-

mance and technical improvement of Kendo athletes.

With the development of the platform and the ease

of distribution of the HMD technology (through the

Meta Quest 2 VR headset), it is possible to dissemi-

nate the use of this platform to other dojo practitioners

who are distant from the nearest teachers and schools,

providing a more accessible tool for their technical

growth and inclusion. Furthermore, during the con-

struction of the project, the team was faced with the

difﬁculty of representing the weight in the simulated

environment, some of the methods used do not meet

the needs of the project, leading to the conclusion that

it would be better to remove the weight simulation of

the application, and that this decision would not affect

the athlete’s practice.

6 FUTURE WORKS

In this way, once this work is concluded, it is intended

to add analysis of the athletes’ reaction time, with

the objective of following their evolution. That said,

this parameter will be used in a dashboard, composed

of several graphs aiming at monitoring the collected

information, allowing coaches to adapt training to the

needs of the athlete.

Funding: This work was also partially funded

by FCT—Fundac¸

ao para a Ci

encia e a Tecnolo-

gia (FCT) I.P., through national funds, within the

scope of the UIDB/00127/2020 project (IEETA/UA,

http://www.ieeta.pt/ (accessed on August 2nd 2023)).

REFERENCES

Baca, A., Dabnichki, P., Heller, M., and Kornfeind, P.

(2009). Ubiquitous computing in sports: A review and

analysis. Journal of Sports Sciences, 27(12):1335–

1346.

Bideau, B., Kulpa, R., Vignais, N., Brault, S., Multon, F.,

and Craig, C. (2009). Using virtual reality to analyze

sports performance. IEEE Computer Graphics and

Applications, 30(2):14–21.

Craig, C. (2013). Understanding perception and action in

sport: how can virtual reality technology help? Sports

Technology, 6(4):161–169.

Fleming, P., Young, C., Dixon, S., and Carr

e, M. (2010).

Athlete and coach perceptions of technology needs for

evaluating running performance. Sports Engineering,

13:1–18.

Jeong, K., Tan, A. M., and Fuss, F. K. (2018). Smart kote

glove for assessment of scoring parameters of dan and

kyu grade kendokas. In Multidisciplinary Digital Pub-

lishing Institute Proceedings, volume 2, page 225.

Konyukhov, A. and Yeniavci, H. Modeling strategy for the

sword mechanics: Kinematic and dynamic simulation

of the control motion for japanese swords in kendo

kata.

Liebermann, D. G., Katz, L., Hughes, M. D., Bartlett,

R. M., McClements, J., and Franks, I. M. (2002).

Advances in the application of information technol-

ogy to sport performance. Journal of sports sciences,

20(10):755–769.

Murase, N., Horiuchi, G., Sumi, K., Horiyama, K., and

Sakurai, S. (2017). Biomechanical factors to shorten

the movement time of men striking motion in kendo.

International Journal of Sport and Health Science,

page 201611.

Neumann, D. L., Mofﬁtt, R. L., Thomas, P. R., Loveday,

K., Watling, D. P., Lombard, C. L., Antonova, S., and

Tremeer, M. A. (2018). A systematic review of the ap-

plication of interactive virtual reality to sport. Virtual

Reality, 22(3):183–198.

Construction of a Virtual Environment to Measure the Evolution of Kendo Athletes

167

Petri, K., Bandow, N., and Witte, K. (2018). Using several

types of virtual characters in sports-a literature survey.

International Journal of Computer Science in Sport,

17(1):1–48.

Petri, K., Emmermacher, P., Danneberg, M., Masik, S.,

Eckardt, F., Weichelt, S., Bandow, N., and Witte, K.

(2019). Training using virtual reality improves re-

sponse behavior in karate kumite. Sports Engineering,

22:1–12.

Salmon, G. (2013). Kendo: A comprehensive guide to

Japanese swordsmanship. Tuttle Publishing.

Stinson, C. and Bowman, D. A. (2014). Feasibility of train-

ing athletes for high-pressure situations using virtual

reality. IEEE transactions on visualization and com-

puter graphics, 20(4):606–615.

Stockwell, S., Trott, M., Tully, M., Shin, J., Barnett, Y.,

Butler, L., McDermott, D., Schuch, F., and Smith, L.

(2021). Changes in physical activity and sedentary be-

haviours from before to during the covid-19 pandemic

lockdown: a systematic review. BMJ open sport & ex-

ercise medicine, 7(1):e000960.

Takata, M., Nakamura, Y., Torigoe, Y., Fujimoto, M.,

Arakawa, Y., and Yasumoto, K. (2019). Strikes-

thrusts activity recognition using wrist sensor to-

wards pervasive kendo support system. In 2019

IEEE International Conference on Pervasive Comput-

ing and Communications Workshops (PerCom Work-

shops), pages 243–248. IEEE.

Torigoe, Y., Nakamura, Y., Fujimoto, M., Arakawa, Y., and

Yasumoto, K. (2020). Strike activity detection and

recognition using inertial measurement unit towards

kendo skill improvement support system. Sensors and

Materials, 32(2):651–673.

Wang, J. (2012). Research on application of virtual reality

technology in competitive sports. Procedia Engineer-

ing, 29:3659–3662.

icSPORTS 2023 - 11th International Conference on Sport Sciences Research and Technology Support

168