Design and Implementation of a Stone Rotation Measurement System

with IMU Sensor and Stone Behavior Presentation System

Chen Guanyu

1 a

, Haruna Mori

1 b

, Yoshinari Takegawa

1 c

, Shimpei Aihara

2 d

and Fumito Masui

3 e

Graduate School of Systems Information Science, Future University Hakodate, Hokkaido, Japan

Department of Sport Science and Research, Japan Institute of Sport Sciences, Tokyo, Japan

Information Communication Group, Kitami Institute of Technology, Hokkaido, Japan

{g2123038, g2124052, yoshi}@fun.ac.jp, shimpei.aihara@jpnsport.go.jp, f-masui@mail.kitami-it.ac.jp

Keywords:

Curling, Visualization, Rotation Measurement.

Abstract:

In the strategic sport of curling, referred to as “chess on ice,” precise measurement and real-time presentation of

stone dynamics are critical for performance optimization. This study introduces a novel system that integrates

Inertial Measurement Units (IMUs) and real-time data processing to track and analyze the rotational dynamics

of curling stones accurately. Additionally, our system includes the Stone Behavior Presentation System, which

interactively displays real-time data on a tablet device, providing instant feedback to players and coaches.

This dual approach not only enhances the accuracy of data collection but also improves the immediacy and

applicability of the information for strategic decision-making in training and competitive environments.

1 BACKGROUND

Curling is a sport in which players slide stones on ice,

aiming to achieve a higher score than their opponents.

Curling is a strategic sport, also known as chess on

ice, and information sharing among players is essen-

tial. For example, it is no exaggeration to say that

who wins or loses a match is determined by how ac-

curately the behavior of a thrown stone, such as its

moving speed and rotation speed, can be collected.

Therefore, being able accurately and quickly to un-

derstand the ever-changing behavior of the stones is

one of the most important skills for curling players.

Takegawa et al. have developed a technique for es-

timating the position and moving speed of stones on

a curling sheet(Takegawa et al., 2023). Additionally,

curling requires precise control of a stone’s move-

ment on ice, where understanding its rotational dy-

namics is crucial for performance analysis and im-

provement(Maeno, 2014). Traditional methods for

measuring the rotation of a curling stone rely heavily

on manual observation and post-event video analysis,

https://orcid.org/0009-0007-8754-6979

https://orcid.org/0009-0001-6202-4150

https://orcid.org/0000-0003-1947-0021

https://orcid.org/0000-0002-8513-0204

https://orcid.org/0000-0001-9979-8734

which can be time-consuming and prone to inaccura-

cies. Particularly in the case of curling, players may

inadvertently obstruct the cameras, making it impos-

sible to accurately conﬁrm the rotation through slow-

motion replay. Additionally, in post-event analysis,

frequent replays are necessary, so good visualization

is a valuable factor for efﬁcient data interaction.

Accurate and real-time measurement of the curl-

ing stone’s rotation is imperative for coaches, play-

ers, and researchers to gain insights into the mechan-

ics of the sport. Such data is essential for optimizing

strategies, improving training methods, and advanc-

ing the scientiﬁc understanding of curling dynamics.

Moreover, real-time feedback during training sessions

can potentially enhance performance by allowing im-

mediate adjustments based on precise rotational data,

making this a worthwhile research consideration.

The primary objective of this study is to develop

a novel system that utilizes smart IMU sensors and

quaternion data to accurately measure and analyze the

rotational dynamics of a curling stone. Thus, we de-

velop “Tablet Stone,” a system that acquires real-time

stone behavior data (moving speed, rotation speed,

and rpm) and interactively displays the stone’s behav-

ior on a tablet device attached to the stone.

140

Guanyu, C., Mori, H., Takegawa, Y., Aihara, S. and Masui, F.

Design and Implementation of a Stone Rotation Measurement System with IMU Sensor and Stone Behavior Presentation System.

DOI: 10.5220/0012899700003828

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

In Proceedings of the 12th International Conference on Sport Sciences Research and Technology Support (icSPORTS 2024), pages 140-147

ISBN: 978-989-758-719-1; ISSN: 2184-3201

2 RELATED RESEARCH

2.1 Rotation Measurement

Previous studies (Barbour and Schmidt, 2001) have

explored various methods for tracking and analyzing

the movement of sports equipment, including the use

of video analysis and mechanical sensors. However,

these methods often fall short in terms of real-time

data acquisition and processing capabilities. Recent

advancements in Inertial Measurement Unit (IMU)

sensors have opened new possibilities for sports an-

alytics, providing high-resolution data on motion dy-

namics.

In particular, IMU sensors have been successfully

applied to real-time motion tracking in ﬁelds such

as biomechanics and robotics(Verdel et al., 2023).

The application of these sensors in curling, however,

remains relatively unexplored. This study aims to

bridge this gap by developing a system that leverages

smart IMU sensors to measure the rotational angle

and speed of a curling stone in real time, providing

immediate feedback and detailed post-analysis capa-

bilities.

2.2 Feedback Methods in Curling

There is, even now, no established theory regarding

what causes the stones to curve as they travel(Murata,

2022). Due to the complex and advanced element of

strategy and the fact that technique is dependent on

players, there are few scientiﬁc approaches to curling,

in comparison with other sports. Nevertheless, Masui

et al. have started research titled ‘Curling Science’,

which is an initiative to create new strategy support

that integrates information technology

. This project

involves research on digital curling(Ito and Kitasei,

2015), tactical analysis, measurement of stone be-

havior, and sweeping(Gwon et al., 2020; Won et al.,

2018). Digital curling refers to the proposal of a vir-

tual curling space, created using a computer’s physi-

cal simulator, that acts as a space to enable discussion

of curling strategy. This concept has been developed

by a large number of people(Yamamoto et al., 2015).

In addition, systems, such as the Portable Tactical

Support DB System, have been developed to record

information, e.g., shots, stone layout, match scores,

players taking part, and the condition of the ice, on a

tablet device (Masui et al., 2015; Masui et al., 2016;

Otani et al., 2016).

https://kaken.nii.ac.jp/en/grant/KAKENHI-

PROJECT-15H02797/

Figure 1: Work loop.

3 MEASUREMENT METHOD

3.1 System Proposal

3.1.1 System Overview

The proposed system utilizes the BNO055 IMU sen-

sor to monitor and analyze the rotational dynamics of

a curling stone in real-time. The sensor operates in

NDOF (Nine Degrees of Freedom) mode and trans-

mits data packets via UDP at approximately 9.2 Hz.

These packets include quaternions, raw accelerom-

eter, gyroscope, and magnetometer data. The host

computer, implemented using Python on a Surface

Pro 8, is responsible for receiving the quaternion data,

converting it to Euler angles, and calculating the ro-

tation angle and speed. A low-pass ﬁlter is applied

to the rotation speed data to reduce noise. The pro-

cessed data is displayed using a graphical user inter-

face (GUI) implemented with matplotlib and saved in

CSV format for post-experiment analysis. Figure 1 il-

lustrates the overall workﬂow of the system, detailing

the data ﬂow and processing steps involved in real-

time monitoring and analysis.

Design and Implementation of a Stone Rotation Measurement System with IMU Sensor and Stone Behavior Presentation System

141

Figure 2: Sensor device.

Figure 3: Results and analysis.

3.1.2 Hardware Components

BNO055 Sensor. The BNO055, developed by

Bosch, is a high-precision IMU sensor integrating a

triaxial 14-bit accelerometer, a high-precision triaxial

16-bit gyroscope, and a triaxial geomagnetic sensor.

This advanced integration reduces size and complex-

ity while minimizing potential errors.

BME280 Sensor. The BME280 sensor collects

barometric pressure data, which can be used for fu-

ture research. This additional data can provide in-

sights into environmental conditions that might affect

the curling stone’s performance.

ESP32 M5StickC PLUS. This microcontroller

handles the data acquisition from both the BNO055

and BME280 sensors, with all components connected

via the I2C interface. The combined weight is mini-

Information from Bosch’s ofﬁcial documentation.

https://www.bosch-sensortec.com/products/smart-sensor-

systems/bno055/

Figure 4: Collision analysis.

mized to only 19.8g, reducing the impact on the curl-

ing stone.

3.1.3 Data Transmission

Data packets, including quaternions and sensor read-

ings, are sent from the ESP32 M5StickC PLUS to the

host computer using the UDP protocol. This ensures

minimal latency and efﬁcient data transfer for real-

time processing.

3.1.4 Data Processing on the Host Computer

The host computer runs a Python-based program to

receive and process the data. Below is a detailed ex-

planation of the implementation steps:

Data Reception: A UDP socket is conﬁgured to lis-

ten for incoming data packets. The data is decoded

and parsed into individual sensor readings.

Quaternion to Euler Angle Conversion: Us-

ing the scipy.spatial.transform library, the received

quaternions are converted to Euler angles. This con-

version is crucial for calculating the rotation angle and

speed of the curling stone. Quaternions are used be-

cause they avoid the gimbal lock problem (Bernardes

and Viollet, 2022) that can occur with Euler angles,

providing a more stable representation of orientation.

Rotation Angle. The cumulative rotation angle is

computed from the Euler angles. The formula used

is:

Cumulative Rotation =

∑

i=1

(θ

− θ

i−1

) (1)

where θ represents the Euler angle around the z-axis.

The speed of rotation is calculated by differenti-

ating the Euler angles over time. The formula used

is:

Rotation Speed =

∆θ

∆t

(2)

icSPORTS 2024 - 12th International Conference on Sport Sciences Research and Technology Support

142

This measures how quickly the stone is rotating in de-

grees per second.

To reduce noise in the rotation speed data, a low-

pass ﬁlter is applied. The ﬁlter formula is as follows:

Filtered Speed = α ·Current Speed

Rotation Speed

+ (1 −α) · Previous Speed

Rotation Speed

(3)

where α is the ﬁlter coefﬁcient that determines the

amount of smoothing.

Data Visualization. A real-time GUI is imple-

mented using Matplotlib to display the cumulative ro-

tation and rotation speed. The GUI updates at a fre-

quency of approximately 9 Hz, providing immediate

visual feedback.

In Figure 3, the red line represents the curling

stone’s rotational angle(degree), which can be seen

to increase as the stone rotates. The blue line in-

dicates the rotational speed(degree per second), and

the yellow line represents the raw magnetometer

data(rotational angle).

Data Storage. All processed data, including times-

tamps, cumulative rotation angles, and rotation

speeds, are saved to a CSV ﬁle for later analysis. This

ensures that detailed records of each experiment are

maintained.

By following these steps, the system can provide

accurate and real-time measurements of the curling

stone’s rotational dynamics, offering valuable insights

for both training and research purposes.

3.1.5 Post-Experiment Analysis

The post-experiment analysis involves processing the

recorded data from the experiments to gain insights

into the rotational dynamics of the curling stone. This

section details the steps taken to analyze the data, in-

cluding data cleaning, quaternion to Euler angle con-

version, rotation calculation, and visualization. Load-

ing Data: The recorded data is loaded from a CSV

ﬁle, which contains timestamps and sensor readings,

including quaternions and magnetometer data.

Timestamp Conversion. Timestamps are con-

verted to a proper time format and adjusted to rela-

tive time from the start of the experiment. This allows

for accurate calculation of time intervals between data

points.

Offset Calculation. For each timestamp, offsets are

calculated based on the number of data points to dy-

namically determine intervals. This helps in accu-

rately plotting the data over time.

Quaternion to Euler Angle Conversion: Using

the scipy.spatial.transform library, quaternions are

converted to Euler angles. This conversion is essen-

tial for calculating the rotation angles of the curling

stone.

Cumulative Rotation Changes. The changes in ro-

tation angles (yaw) are calculated and accumulated to

obtain the cumulative rotation over time.

Magnetic Angle Calculation. The changes in the

magnetic ﬁeld angle are also calculated and accumu-

lated to provide additional insights into the stone’s ro-

tation dynamics.

Rotation Speed: The rotation speed is calculated

by dividing the rotation changes by the time inter-

vals. A Gaussian ﬁlter is applied to smooth the rota-

tion speed data.The alpha value of the low-pass ﬁlter

is adjusted during the post-experiment analysis to op-

timize the smoothness of the speed curves, ensuring

accurate representation of the curling stone’s dynam-

ics.

Visualization. The processed data is visualized us-

ing Matplotlib, displaying both cumulative rotation

and rotation speed over time. Annotations are added

to show total rotation in degrees and turns, as well as

cumulative magnetic angle changes.

3.2 Experiments

3.2.1 Equipment Setup and Procedure

To evaluate the feasibility and performance of the

prototype device, a series of experiments were con-

ducted. The lower unit, comprising the sensors and

M5StickC PLUS, was powered on, calibrated using a

ﬁgure-of-eight motion to ensure accurate orientation

tracking, and securely attached to the curling stone’s

top cover. Following this setup, the host computer’s

Python program was initiated to start data reception.

Throughout the experiments, a camera continuously

tracked and recorded the curling stone’s movements

to validate the data post-experiment. Various exper-

imental conditions were tested, including combina-

tions of low, medium, and high speeds, clockwise and

counterclockwise rotations, and collisions. The ob-

jective was to evaluate the prototype under different

scenarios.

Design and Implementation of a Stone Rotation Measurement System with IMU Sensor and Stone Behavior Presentation System

143

3.2.2 Data Collection and Initial Findings

During the experiments, data transmission, reception,

and parsing were successful. The angle calculation al-

gorithms performed correctly, and the GUI refreshed

as expected. Once the experiments were complete,

the data, including the computed and raw data, was

automatically saved.

3.2.3 Post-Experiment Analysis

The post-experiment analysis involved using the

saved raw data to recreate the experiments. Addition-

ally, the raw magnetometer data was used to verify

the results and mitigate potential errors from Micro-

Electro-Mechanical Systems (MEMS) algorithms.

3.2.4 Example and Observations

As illustrated in the accompanying Figure 2, a

medium-speed clockwise rotation experiment showed

that the curling stone’s rotational angle(degree), rep-

resented by a red line, increases as the stone rotates.

The blue line indicates the rotational speed(degree per

second), and the yellow line represents the raw mag-

netometer data(rotational angle). The smoothness of

the red line compared to the yellow line is evident,

highlighting the effectiveness of the data smoothing

process. The optimized UI from the post-experiment

analysis quickly displays the number of rotations and

speciﬁc degrees. The Matplotlib component allows

for zooming in on speciﬁc areas for detailed visual

analysis.

3.2.5 Additional Insights and Improvements

Interesting observations were made from the collision

data set, which is valuable for future collision detec-

tion functionality. In Figure 4, the rotational speed

showed an upward trend at the moment of collision

before decreasing, providing crucial information for

monitoring collisions. Additionally, the experiments

demonstrated the ability to distinguish between clock-

wise and counterclockwise rotations. As shown in

Figure 5, the positive and negative changes in rota-

tional magnitude correspond to clockwise and coun-

terclockwise rotations, respectively.

Furthermore, the experiments revealed several de-

sign improvement points for the prototype:

1. Increase Sampling Rate. Improving the sen-

sor sampling rate can provide more granular data,

leading to more precise analysis of the curling

stone’s rotational dynamics.

2. Enhance Program Stability. Ensuring the stabil-

ity of the Python software is crucial for long-term

Figure 5: clockwise and counterclockwise rotations.

experiments. This can be achieved by implement-

ing better error handling and optimizing the code.

3. Implement Asynchronous and Multithreaded

Processing. Utilizing asynchronous and multi-

threaded processing can reduce CPU load and im-

prove the system’s real-time capabilities. This

enhancement will enable smoother and more ef-

ﬁcient data handling, especially during high-

frequency data acquisition.

4. Improve Real-Time Visualization. Enhancing

the real-time visualization capabilities of the sys-

tem can provide immediate feedback to users.

This can be achieved by optimizing the GUI and

ensuring that it can handle high-frequency data

updates without lag.

5. Ensure Reliable Wireless Network Support.

Reliable wireless network support is essential to

maintain data transmission integrity, especially

during high-frequency updates.

6. Reduce Power Consumption. The cold envi-

ronment of the curling rink affects battery perfor-

mance. Reducing unnecessary power consump-

tion can extend battery life, ensuring longer oper-

ational periods during experiments.

4 TABLET STONE

The Tablet Stone is intended for players who have

competed in the All-Japan University Curling Cham-

pionships, or a higher-level competition, and is in-

tended to be used during practice, such as when prac-

tising delivery and sweeping. The requirements for

the development of the Tablet Stone are listed below.

1. Since the players are concentrating on delivery

and sweeping, it is appropriate to present infor-

mation on the behavior of the stones in the vicin-

ity where the players are looking.

2. Since the throwers and sweepers comprise differ-

ent numbers of people, who also assume differ-

ent postures and require different information, the

icSPORTS 2024 - 12th International Conference on Sport Sciences Research and Technology Support

144

IMU sensor

Thrower-oriented

presentation method

Pressure sensor

Tablet

Sweeper-oriented

presentation method

Figure 6: System conﬁguration.

Figure 7: Examples of thrower-oriented presentation

method.

Figure 8: Examples of sweeper-oriented presentation

method.

information presentation method must take these

factors into account. In particular, the sweepers

need to view information from a display on a ro-

tating stone, and it is necessary to consider a pre-

sentation method that is not affected by the rota-

tion of the stone.

The system conﬁguration of the proposed Tablet

Stone is shown in Figure 6. Tablet Stone displays the

stone’s behavior in real-time on the screen of an on-

board tablet device. The sensor data from the IMU

sensor mounted on the tablet and the data from the

pressure sensor mounted under the handle are con-

verted into moving speed, rotation speed, and rpm on

a server located outside the curling sheet. These data

are sent to the tablet and displayed. Data between de-

vices is sent and received via UDP communication.

The tablet was developed as an iOS application using

SwiftUI on an iPad mini.

We propose two presentation methods, one for

throwers and the other for sweepers. The pressure

sensor mounted on the stone’s handle is used to switch

between the two presentation methods. Each presen-

tation method is described in detail in subsections 4.1

and 4.2.

4.1 Thrower-Oriented Presentation

Method

The thrower-oriented presentation method is used in

situations where the thrower is holding the stone han-

dle and delivering the stone. The appearance of

the thrower-oriented presentation method is shown in

Figure 7. The posture of the throwers is low, and the

handle itself and the hand holding the handle obstruct

the display. These factors were taken into considera-

tion when deciding the placement of the presentation

content and the size of the text. Figure 7-(I) is a me-

ter that displays the stone’s behavior in real-time. The

values are, from the top, the moving speed and the ro-

tation angle. The moving speed is also indicated by an

orange circle meter. The maximum value of the cir-

cle meter is 10.0 m/s, which makes it easy to visually

understand how fast the stone is sliding. The rotation

angle is displayed to conﬁrm that the slide is straight.

Design and Implementation of a Stone Rotation Measurement System with IMU Sensor and Stone Behavior Presentation System

145

meter

Figure 9: Examples of meter rotation direction ﬁxation function and meter absolute position ﬁxation function.

fix

(Var X, VarY)

Δt

(Var X, VarY)

r*sin(-

)

r*cos(-

)

(CenterX, CenterY)

fix

(MeasureX, MeasureY)

Figure 10: Example of meter absolute position ﬁxation

function.

Figure 7-(II) is a button to start UDP communication.

Figure 7-(III) is a button to reset the orientation of the

meter. After resetting, the meter faces the thrower’s

side.

4.2 Sweeper-Oriented Presentation

Method

The sweeper-oriented presentation method is used in

the situation where a stone has been thrown and the

sweepers are sweeping the stone. The sweepers are

in pairs and are positioned on either side of the stone

as shown in Figure 8. The same information as in the

thrower-oriented presentation method is displayed on

the meter. Speciﬁcally, Figure 8-(I) is a meter that

displays the stone’s behavior in real-time. The values

are, from the top, moving speed, rotation speed, and

rpm. The moving speed is the same as that in the

thrower-oriented presentation method. The rotation

speed is displayed to allow the user to check how fast

the stone is rotating. Rpm is displayed to allow the

user to check how many rpm have been made. Figure

8-(II) is a button to start UDP communication. Figure

8-(III) is a button to reset the orientation of the meter.

After resetting, the meter will face the sweepers on

the left and right, respectively.

For the sweeper-oriented presentation method, we

propose a Meter Rotation Direction Fixation Function

and a Meter Absolute Position Fixation Function in

order to maintain readability of the meter when the

stone is rotating.

4.2.1 Meter Rotation Direction Fixation

Function

The assumed initial position of the stone is as in Fig-

ure 9-A. When the stone rotates, the meter seen by the

player also rotates (Figure 9-B), resulting in poor vis-

ibility of the meter. The meter rotation direction ﬁxa-

tion function estimates the rotation angle of the stone

from the rotation angle of the IMU sensor mounted

on the tablet and rotates the meter in the opposite di-

rection, as shown in Figure 9-C.

4.2.2 Meter Absolute Position Fixation Function

Even if the Meter Rotation Direction Fixation Func-

tion is applied, the meter will appear in front of the

handle or behind the handle depending on the rotation

of the stone, as shown in Figure 9-C. By applying the

Meter Absolute Position Fixation Function, the posi-

tion on the stone where the meter is displayed is ﬁxed

regardless of the rotation of the stone, as shown in

Figure 9-D.

This function ﬁxes the absolute position of the

displayed meter by rotating the meter by θ degrees

with respect to the center point. An illustration of

the function is shown in Figure 10. The left ﬁg-

ure shows the initial state without rotation, and the

right ﬁgure shows the state with rotation. In the ini-

tial state without rotation, the coordinates of the ref-

erence position are D

f ix

(VarX, VarY ) and the cen-

ter coordinates of the stone are C(CenterX, CenterY ).

With rotation, D

∆t

(VarX, VarY ) denotes the coordi-

nates of the reference position, C(CenterX, CenterY)

denotes the coordinates of the center of the stone, and

f ix

(MeasureX, MeasureY ) denotes the coordinates

of the position drawn after rotation by θ degrees. Let

r be the distance from C to D

f ix

. θ is the angle con-

verted from the rotation angle of the built-in sensor to

radians. The formulas for MeasureX and MeasureY

are described below.

MeasureX = CenterX + r × sin(−θ) (4)

MeasureY = CenterY − r × cos(−θ) (5)

icSPORTS 2024 - 12th International Conference on Sport Sciences Research and Technology Support

146

5 CONCLUSION

The developed system, featuring advanced IMU sen-

sors and robust data processing techniques, represents

a signiﬁcant breakthrough in the real-time analysis of

curling stone dynamics. This system not only mea-

sures rotational dynamics accurately, but also fea-

tures the innovative Stone Behavior Presentation Sys-

tem, which displays real-time data interactively on a

tablet device. This dual functionality enhances train-

ing effectiveness and strategic decision-making for

both coaches and players.

Future enhancements will focus on increasing the

sampling rate, improving system stability, and ex-

panding real-time processing capabilities, ensuring

the system remains cutting-edge. As for veriﬁcation,

we plan to verify if the tablets affect the trajectory of

the stones and the players, and to test the two pro-

posed methods to see if they affect the training effec-

tiveness. Additionally, we plan to reﬁne the Stone Be-

havior Presentation System to offer more customized

and user-friendly interfaces that can adapt dynami-

cally to different game scenarios and user preferences.

Through these advancements, our system promises to

revolutionize training methods and strategic planning

in the sport of curling.

ACKNOWLEDGEMENTS

This work was supported by the “The Enhancement of

HPSC Infrastructure through Technology Innovation

Project” of Japan Sports Agency.

REFERENCES

Barbour, N. and Schmidt, G. (2001). Inertial sensor tech-

nology trends. IEEE Sensors Journal, 1(4):332–339.

Bernardes, E. and Viollet, S. (2022). Quaternion to euler

angles conversion: A direct, general and computation-

ally efﬁcient method. PLoS One, 17(11):e0276302.

Gwon, J., Kim, H., Bae, H., and Lee, S. (2020). Path plan-

ning of a sweeping robot based on path estimation of

a curling stone using sensor fusion. Electronics, 9(3).

Ito, T. and Kitasei, Y. (2015). Proposal and implementa-

tion of ”digital curling”. In 2015 IEEE Conference on

Computational Intelligence and Games (CIG), pages

469–473.

Maeno, N. (2014). Dynamics and curl ratio of a curling

stone. Sports Engineering, 17:33–41. Published: 30

July 2013, Issue Date: March 2014.

Masui, F., Hirata, K., Otani, H., Yanagi, H., and Ptaszynski,

M. (2016). Informatics to support tactics and strate-

gies in curling. International Journal of Automation

Technology, 10(2):244–252.

Masui, F., Ueno, H., Yanagi, H., and Ptaszynski, M. (2015).

Toward curling informatics — digital scorebook de-

velopment and game information analysis. In 2015

IEEE Conference on Computational Intelligence and

Games (CIG), pages 481–488.

Murata, J. (2022). Study of curling mechanism by preci-

sion kinematic measurements of curling stone’s mo-

tion. Scientiﬁc Reports, 12(1).

Otani, H., Masui, F., Hirata, K., Yanagi, H., and Ptaszyn-

ski, M. (2016). Analysis of curling team strategy

and tactics using curling informatics. In Correia,

P. P. and Cabri, J., editors, icSPORTS, pages 182–187.

SciTePress.

Takegawa, Y., Sasaki, N., Aihara, S., and Masui, F. (2023).

Development of a curling stone tracking system using

infrared leds, and an accompanying application. In

Proceedings of the 11th International Conference on

Sport Sciences Research and Technology Support (ic-

Sports 2023), pages 136–143.

Verdel, N., Mohorcic, M., Drobnic, M., Supej, M., and De-

polli, M. (2023). Time synchronization in wireless

imu sensors for accurate gait analysis during running.

In 2023 IEEE International Workshop on Sport, Tech-

nology and Research (STAR), pages 126–129.

Won, D.-O., Kim, B.-D., Kim, H.-J., Eom, T.-S., Muller,

K.-R., and Lee, S.-W. (2018). Curly: An ai-based

curling robot successfully competing in the olympic

discipline of curling. In Proceedings of the Twenty-

Seventh International Joint Conference on Artiﬁcial

Intelligence, IJCAI-18, pages 5883–5885. Interna-

tional Joint Conferences on Artiﬁcial Intelligence Or-

ganization.

Yamamoto, M., Kato, S., and Iizuka, H. (2015). Digital

curling strategy based on game tree search. In 2015

IEEE Conference on Computational Intelligence and

Games (CIG), pages 474–480.

Design and Implementation of a Stone Rotation Measurement System with IMU Sensor and Stone Behavior Presentation System

147