Circuit: Jurnal Ilmiah Pendidikan Teknik Elektro

Volume 8 No. 2 March 2024 - August 2024

ISSN: 2549-3698; E-ISSN: 2549-3701
DOI: 10.22373/crc.v8i2.23124

Design of an Arduino-Based Boat Roll Stabilizer System Prototype

Using MPU6050 Sensor
Erwin Sitompula, Andri Setia Prabowoa, Hendra Tariganb, Antonius Suhartomoa
aPresident University, Indonesia
bMississippi College, US
E-mail: [email protected]

Submission: 19-04-2024 Accepted: 16-07-2024 Published: 13-08-2024

Abstract
In the midst of lakes or oceans, boats used for recreational fishing that are left immobile
are more vulnerable to rolling waves or wind. This study was conducted to develop a
prototype of the boat's stabilization system and control scheme in order to address the
stability issue with the craft. Based on the prototype concept, a small boat was
constructed with a control loop that included two DC motors, a motor driver, a gyroscope
MPU6050, and an Arduino Uno microcontroller. A prototype stabilizer was used to
evaluate two different sets of control parameters, and in both situations the stabilizer was
able to provide the boat with critical interference before bringing it back to a vertical
position. PWM 1 level change rate (Pulse-Width-Modulation) yields better results than 4
level change rates (PWM), with faster stabilization achievement. (1,12 seconds versus
2,46 second). In the development of the prototype, it is necessary to test the
implementation of the system on the boat to the actual size.
Keywords: Boat Stabilizer, Digital Control, MPU6050, Roll Angle

Abstrak
Perahu yang dipakai untuk rekreasi memancing saat berada dalam kondisi stasioner di
tengah danau atau laut, lebih rentan mengalami gulingan akibat angin atau ombak. Untuk
memecahkan masalah kestabilan perahu ini, peneltiian ini dilakukan untuk menciptakan
suatu prototipe sistem stabilisator perahu beserta skema kendalinya. Pada rancangan
prototipe, dibuat sebuah perahu miniatur yang dilengkapi dengan lup kendali yang terdiri
dari mikrokontroler Arduino Uno, giroskop MPU6050, dua motor DC, dan driver motor.
Kontrol digital berdasarkan tanda kesalahan (Error-Sign-based Controller) dengan
pembatas laju perubahan digunakan untuk mengontrol stabilisator. Dua set parameter
kontrol diuji pada prototipe stabilisator dan dalam kedua kasus tersebut stabilisator
berhasil membawa perahu kembali ke posisi tegak setelah diberi gangguan kritis. Laju
perubahan 1 level PWM (Pulse-Width-Modulation) memberikan hasil lebih baik
dibandingkan laju perubahan 4 level PWM, dengan pencapaian stabilisasi yang lebih
cepat (1,12 detik berbanding 2,46 detik). Pada pengembangan prototipe, perlu dilakukan
uji implementasi sistem pada perahu dengan ukuran sebenarnya.
Kata kunci: Stabilisator Perahu, Kontrol Digital, MPU6050, Sudut Guling

Introduction
Moving people or products from one location to another is known as
transportation, and it has become essential to economic expansion. Transportation that is
safe and secure is a necessary component of industry and commerce-related activities.
The National Transportation Safety Committee (KNKT) was established by the

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 217
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

government of the Republic of Indonesia in 1999 [1] with the goal of achieving
transportation safety and enhancing transportation services. One example of how the
government recognizes the value of safety in all forms of transportation, including land,
sea, and air, is this committee. Water transportation is of strategic significance for
Indonesia as an archipelagic state. This mode of transportation is instrumental in order to
reach and connect the frontier, outermost, and least developed regions and to unify
Indonesia [2]. Furthermore, water transportation using boats and small ships is also the
most economical and efficient means for short-distance occasional trips (for recreational
purposes) or regular commutes (between work and home) across shallow water such as
rivers and lakes. However, water transportation through rivers and lakes is still accident-
prone and requires more safety precautions. There have been several incidents of capsized
boats due to waves and wind [3][4]. In such incidents the turbulence caused water to enter
the boat, causing panic among the passengers, worsening the stability of the boat before
capsizing it, and inflicting fatalities. In another situation, a traditional riverboat used by
recreational fishers rolled and collapsed after being unable to maintain its stability due to
the wash of a passing vessel [5].
A boat stabilizer is a system that is necessary to lessen the rolling motion of a
boat. A stabilizer, such a bilge keel or anti-roll water ballast, can be passive, meaning that
no external power is needed. By raising the hydrodynamic resistance, the bilge keel
lessens the roll of the boat but also impedes its forward propulsion [6][7]. However, the
mass of the water ballast can be changed to accommodate the boat's present load.
However, compared to other types of ballast, water has a low density, hence the tank
needs more room. Fins and gyroscopes are examples of active stabilizers that require
additional power and a unique control mechanism. They are favored because they are
more efficient. Fins can be made to function at zero speed when the boat is at anchor or
when it is moving forward [8][9]. Underway fins are designed to stabilize the boat while
it moves, utilizing the movement of the water over the fins underwater to create the
stabilizing force [10][11]. Fins at zero speed constantly move to adapt to the waves,
current, and wind. A gyroscope is probably the most compact and effective type of
stabilizer, not only for boats but also for robots [12][13]. Its disadvantages include being
the priciest kind of stabilizer, hefty, and needing intricate maintenance.
The aforementioned facts motivated the authors to design an active stabilizer with
a suitable control scheme for a boat at anchor. Boat stabilizer systems have been
investigated previously [14][15]. Santoso proposed the use of a flywheel and stepper
motor to provide a counterweight to a boat [14]. The rotational speed of the flywheel is
regulated by a controller module while its orientation is adjusted based on the angle read
by an MPU6050 gyroscope. A stabilization attempt using servo motors and fins was
proposed in [15]. An Arduino Uno microcontroller was used to regulate the fins'
locations, which were installed on both sides of the boat. It was demonstrated that the roll
angle had stabilized back to zero after the miniature boat was subjected to the disturbance.
Although it has never been studied before, the stabilizer design put forth by the
authors of this research is tenable. The suggested stabilizer is made up of motor-driven
propellers mounted on either side of a boat. The Arduino Uno controller, two sets of DC
motors, propellers serving as actuators, and an MPU6050 gyroscope serving as a sensor

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 218
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

make up the control loop. The control scheme was expected to regulate the propellers’
rotation so that the generated forces could counter the boat’s roll movement [16].
Additionally, a prototype boat that was scaled down was created, with dimensions that
provided sufficient susceptibility to roll disturbance. Aerodynamics was not taken into
account in the boat's design, which was a restriction. Furthermore, because the propeller
position was always set, topics related to propeller tilt were not covered. Here, the authors'
earlier proposal for an error-sign-based controller with a change rate limiter is
implemented with various changes. After that, the boat is equipped with the prototype of
the active stabilizer system. Ultimately, a number of tests are conducted to determine how
well the complete system works to stop the boat from capsizing and lessen roll
movements.

Method
a. Some Nautical Terms and Boat Stability
Some essential terms related to the position around a boat and its rotational motion
are introduced here. Figure 1(a) shows the position around a boat, where starboard means
the right-hand side of the vessel and port refers to the left-hand side. A boat moves in 3
degrees of rotational motion as shown in Figure 1(b). Roll is a rotation around the
longitudinal x-axis, pitch is a rotation around the transverse y-axis, and yaw is a rotation
around the vertical z-axis. The sign of a rotational motion is taken as positive when it is
clockwise (as indicated by the grey circles), and negative otherwise.
+z axis
Bow Yaw +x axis

Direction of Direction of
sailing sailing
Pitch
Roll

Port +y axis

Starboard −y axis

Stern −x axis
−z axis

(a) (b)
Figure 1. Some Nautical Terminologies of a Boat
(a) Directions; (b) Rotations
The stability of a boat is defined as its ability to return to an upright position after
being tilted by external forces such as waves, wind, or change in weight distribution [17].
Examining 0(a), two forces act on a boat as it emerges on water: the downward
gravitational force (Fg) acting on the boat’s center of gravity (blue dot), and the upright
buoyance force (Fb) acting on the center of buoyancy (red dot), which is the centroid of
the displaced volume of water.

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 219
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

Righting Center of gravity

moment Center of bouyancy
Heeling
moment

Fg Fg
Fb Fb Fb Fg

(a) (b) (c)

Figure 2. A Boat at Different Orientations due to Rotation:
(a) Upright Position, (b) Righting Moment; (c) Heeling Moment
If the boat is rolled to one side, the center of gravity remains in the same position
but the center of buoyancy moves at the low side of the boat because the displaced volume
of water is larger on the low side [18]. This situation is shown in 0(b). Both forces Fg and
Fb now create a righting moment that will bring back the boat to an upright position. If
the boat is further rolled over a critical angle, the center of buoyancy moves to a position
where the so-called heeling moment will increase the roll angle of the boat, as depicted in
02 (c). Such a critical role angle is technically termed the angle of vanishing stability.
When the boat’s roll angle reaches this value, it cannot go back to its upright stable
position anymore. On the other hand, it continues moving in the direction of increasing
roll angle magnitude and capsizes.
b. Boat Roll Stabilizer and Control Scheme
In this paper, a roll stabilizer to create a counter moment and prevent a boat from
capsizing. Two propellers are installed on the starboard and the port side of the boat. The
motors that rotate the propellers are activated based on the measured roll angle. As
illustrated in Figure 3, the boat is rolled in a positive direction (clockwise rotation). Then,
to avoid capsizing due to the heeling moment, the stabilizer creates a counter moment Fm
through the starboard unit by turning on the starboard motor and rotating the starboard
propeller. On the contrary, the port unit is activated if the boat’s roll is negative
(counterclockwise rotation).
Port Heeling Stabilizer’s
unit moment counter
moment

Fm

Fg
Fb Starboard
Starboard
unit
unit

Figure 3. The Proposed Boat Roll Stabilizer Viewed from The Stern of the Boat
In this paper, an Error-Sign-based Controller (ESC) with a change rate limiter is
employed as the controller of the boat roll stabilizer. This controller was first proposed in
[19] and was proven to be able to avoid undesirable sudden changes in the control output
that may lead to oscillations of system output or even hardware malfunction [20]. A
modification was made to the controller, by integrating θ, in order to give a threshold
angle for the controller. The control law of ESC is given in Eqs. (1) and (2):

u (k ) + c, sign(e(k ))  0 and e(k )  

u1 (k + 1) =  1 ……………………… (1)
0, otherwise

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 220
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

u (k ) + c, sign(e(k ))  0 and e(k )  −

u2 (k + 1) =  2 ……………………… (2)
0, otherwise

Where u1 is the controller output for the starboard motor, u2 is the controller output
for the port motor, e is the control error, c is the change rate, and k is the discrete time
instant. The control error e is the deviation between the 0° roll angle as a set point and the
measured roll angle. In the beginning, u1 and u2 are set to zero. Based on the reading of
the roll angle, the value of u1 or u2 is increased whether with a small change rate c = Δa
or a large change rate c = Δb. If the roll angle returns to a safe range of the threshold
angle, the outputs of both motors are reset to zero. The control scheme will be further
explained using a flow chart in the following subsection.
c. Miniaturized Boat Prototype
For the purpose of testing the proposed boat roll stabilizer system prototype, a
miniaturized boat prototype was built. The schematic design of the boat can be seen in
04. The boat was made of lightweight extruded polystyrene foam. The dimensions of the
boat are set to allow the boat to roll easily, while adequately stable in pitch and yaw. The
stabilizer consists of two 6 V DC motors with 5 W power and a maximum rpm of 10,000.
Plastic propellers are connected to the motors using 5-cm metal shafts, as indicated in
Figure 4(b). Two battery compartments were prepared, as indicated in 04(c).
7.0 cm
Motor 1 Motor 2 Port Starboard

3.0 cm
M1 M2 M1

Stern Bow
5.0 cm

5.0 cm

5.0 cm
19.5 cm
Shaft Propeller

(a) (b)
Port
Battery Battery
compartment compartment
M2
7.0 cm

9.0 cm

Stern Bow

M1

Starboard
19.5 cm

(c)
Figure 4. Boat Prototype Design:
(a) side view, (b) rear view, and (c) top view

d. Block diagram, flowchart and wiring diagram

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 221
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

The block diagram of the proposed boat roll stabilizer system prototype is shown
in 05. The red arrows signify a high-current power flow, while the blue arrows represent
a low-current power flow or an information flow. Although they are not shown, the
diagram includes the sensor, the processor, the actuators, and the power source.

Gyroscope Microcontroller Battery 1

Battery 2

Motor Driver

DC Motor 1 DC Motor 2

Figure 5. The Block Diagram of the Proposed System

The roll angle of the boat is measured by the MPU6050, as depicted in 06, which
is a three-axis accelerometer and three-axis gyroscope with a Micro-Electro-Mechanical
System (MEM). An MPU6050 IMU chip makes up the MPU6050 module. Its dimensions
are 4 mm × 4 mm × 0.9 mm. It is packaged in a 24-pin QFN format. I2C pull-up resistors,
bypass capacitors, and an AP2112K 3.3V regulator are among the components in the
module. In addition, a power LED that shows the module's power status is included.
Subsequently, the roll angle is processed by the Arduino Uno microcontroller,
where the control scheme program is uploaded. An Arduino Uno is an ATmega328P-
based microcontroller board. It features a 16 MHz ceramic resonator, 6 analog inputs, 14
digital input/output pins (six of which can be used as PWM outputs), a USB port, a power
jack, an ICSP header, and a reset button.
+z axis

−x axis +y axis

+x axis
−y axis
−z axis

Figure 6. The MPU6050

The microcontroller calculates the voltage to be inputted to the DC motors and
transmits it to the L298N motor driver in the form of a Pulse-Width-Modulation (PWM)
signal. The 8-bit PWM values from 0 to 255 are the mapping of 0% to 100% of the motor
voltage. The motor driver L298N is a DC motor driver module used to control the speed
and direction of the motor. This module finds a wide range of applications in electronics
and is frequently connected to an Arduino microcontroller. The IC L298N is an H-bridge
type integrated circuit that can control inductive loads in devices such as solenoids, relays,
DC motors, and stepper motors. The IC L298N utilizes a transistor-transistor logic (TTL)
with NAND gates that serves to change the direction of a motor, either a DC motor or a
stepper motor.
Furthermore, the L298N driver ensures adequate voltage and current for the DC
motors, since the microcontroller operates at 5V with a maximum continuous current of

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 222
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

20 mA, whereas the DC motors operate with a maximum of 6V and 40 mA. The DC
motors used by the prototype generated a torque of 0.40 mN·m. Its shaft diameter is 2
mm while the rotor diameter is 8 mm, with a maximum rotational speed of 10,000 rpm at
the power rate of 5 W. The whole system is powered by two 9 V, 280 mAh batteries. The
flowchart of the proposed system is presented in 07. Once the program starts, the
gyroscope of the MPU6050 is to be initialized. The roll angle is measured when the
connection between the sensor and the microcontroller is established. The Error-Sign-
based Controller (ESC) was implemented with 2 values of change rate, small increment
c = Δa and large increment c = Δb. One of the values is applied to a certain motor based
on the sign and the range of the roll angle, as shown in the flowchart. Afterward, the
change rate value is added to the respective motor voltage. The designated motor voltage
is applied to the motor by the motor driver. If the roll angle is back to the threshold region
−10° ≤ x ≤ 10°, then the voltages of both motors is reset to zero. The discrete-time instant
of the controller is represented by an integer k.
Start

Initialize the
gyroscope

No Is the
gyroscope
connected?

Yes

k=0

Read
the roll angle x

Yes Set the voltage of both Send the designated

−10° ≤ x ≤ 10° ? voltage settings to k=k+1
motors to zero
the motor driver

No

Yes Set small increment Δa Has the No

10° < x ≤ 20° ? to the current voltage sampling time
of motor 1 elapsed?

No Yes

Yes Set large increment Δb Is the No

x > 20° ? to the current voltage reset button
of motor 1 pressed?

No Yes

Yes Set small increment Δa

−20° ≤ x < −10° ? to the current voltage End
of motor 2

No

x < −20° Set large increment Δb

to the current voltage
of motor 2

Figure 7. A flow Chart of the Proposed System

The interconnection among the parts of the boat roll stabilizer system prototype
is shown in 08 in the form of a wiring diagram.

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 223
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

Battery Microcontroller Battery

DC motor

Motor driver

DC motor
Gyroscope

Figure 8. A Wiring Diagram of the Proposed System

e. Gyroscope calibration
Before the implementation, the gyroscope was calibrated in two steps: zero offset
adjustment and sample roll angle measurements. In zero offset adjustment, the gyroscope
is placed on a flat and non-rotating surface. Ideally, the respective angles of rotation for
the roll, pitch, and yaw are zero on such a surface. It turns out that when the gyroscope is
turned on and is placed on the same surface the roll, pitch and yaw angles of rotation are
different from zero. The sensor was initialized and the angles were read for 5 seconds
with a sampling time of 8 ms. Afterward, the average angle of rotation measurements
representing roll, pitch and yaw were calculated giving the zero offset constants.
Subsequently, the three constants were integrated into the control program of the system.
In a sample roll angle measurement, the gyroscope was tested to measure the angles
between 0° and 90° with an increment of 15°.

Results and Discussion

a. Construction of the miniaturized boat prototype
The components that are described in the previous section were assembled to build
a miniaturized boat prototype as depicted in 09. The Arduino microcontroller was placed
in the stern area, while the motor driver was located in the bow area. The motor driver
was mounted close to the port motor.

(a) (b)

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 224
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

(c) (d)
Figure 9. The Miniaturized Boat Prototype: (a) Side View, (b) Rear View,
(c) Top View, and (d) the Shaft and the Propeller
b. Gyroscope calibration
As was mentioned previously, ideally a flat and non-rotating surface must have a
zero angle of rotation associated with each of: roll x-axis, pitch y-axis and yaw z-axis. A
zero offset adjustment was conducted successfully by initiating the gyroscope and
allowing it to measure the angles associated with the three coordinate axes, respectively,
on a flat and non-rotating surface. The average of the respective angles corresponding to
measurement duration of 5 s with a sampling time of 8 ms were 5.60° (roll x-axis), 0.11°
(pitch y-axis), and 1.12° (yaw z-axis). Thus, the rotation angle equaling 5.60° was an
offset value, which was subtracted from all sample roll angle measurements.
The sample roll angle measurements were conducted from 0° to 90° at an
increment of 15° with 5 trials for each angle. In order to achieve the objective of this
sample roll calibration, the gyroscope was oriented to a certain angle of tilt. Such an angle
of tilt was measured with the help of a protractor, as can be seen in 010. In this case, the
protractor was glued to a wall to make the calibration process easy. The roll angles
representing the protractor- and gyroscope-based angle of tilt measurements; and percent
difference can be seen in Table 1. The overall average angle error is 0.534°. This
corresponds to the average of the percent differences for roll angles between 15° and 90°
was 1.21%.

Figure 10. Gyroscope Sample Roll Angle Calibration Utilizing a Protractor

Table 1. The Results for Seven Sample Angle Measurement

Measurement Measurement
Error Error
Protractor Gyrosc Protractor Gyrosc
(°) (°)
(°) ope (°) (°) ope (°)
0 0.10 0.10 0 0.25 0.25
0 0.57 0.57 15 15.03 0.03
0 0.42 0.42 15 15.21 0.21
0 −0.23 0.23 15 15.54 0.54

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 225
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

Measurement Measurement
Error Error
Protractor Gyrosc Protractor Gyrosc
(°) (°)
(°) ope (°) (°) ope (°)
15 14.57 0.43 60 60.54 0.54
15 15.17 0.17 75 74.34 0.66
30 30.62 0.62 75 75.05 0.05
30 30.55 0.55 75 75.24 0.24
30 30.41 0.41 75 75.78 0.78
30 30.24 0.24 75 76.72 1.72
30 29.87 0.13 90 90.54 0.54
45 44.22 0.78 90 90.65 0.65
45 45.05 0.05 90 90.52 0.52
45 45.70 0.70 90 91.60 1.60
45 45.81 0.81 90 89.12 0.88
45 45.24 0.24 Average error 0.534
60 60.12 0.12
60 60.84 0.84
60 60.41 0.41
60 61.65 1.65
c. Prototype testing
The capability of the boat roll stabilizer system prototype was evaluated. It was
tested if it could prevent a miniaturized boat prototype from capsizing if a roll angle
exceeds the critical angle. The dimension of the miniaturized boat itself already leads to
low roll stability, making it easy to capsize at a small roll angle. 0 shows the small
prototype boat while floating undisturbed on water. To find the critical roll angle, an
experiment was conducted. With both motors turned off, the roll angle of the boat was
recorded via the serial monitor with a sampling time of 8 ms. Then the boat was manually
given a roll angle by tilting it to the side and then released. If the boat returned to the
upright position, the experiment was repeated with a greater role angle. If the boat rolled
further and was on the verge of capsizing, the boat’s further rotation was halted and the
respective roll angle was taken as the critical roll angle.

Figure 10. A Prototype Boat is Floating on Water

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 227
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

The results of the critical roll angle experiment are shown in 0. 0(a) represents the
case where the initial roll was given to the starboard direction, while in 0(b) to the port
direction. This experiment shows that the critical roll angle was approximately 30°, as
indicated by the red dashed lines. When the boat was given a 30° initial angle of role and
released, the angle continued to increase to ≥ 60° (as marked by the green dashed lines)
or higher and had to be promptly halted by hand to prevent the boat and the whole
hardware from sinking.

(a)
(b)
Figure 11. The Experiment to Find the Critical Roll Angle:
(a) Roll to Starboard Direction and (b) Roll to Port Direction
At this stage, the functionality of the proposed boat roll stabilizer system prototype
was tested. The setting for the first experiment was Δa = 4 PWM levels and Δb = 8 PWM
levels. The mapping of the PWM levels from 0-255 to 0-100% of motor voltage implies
that an increase of 4 PWM corresponds to an increase of 0.094 V and 8 PWM to 0.188 V.
In the experiment, the boat was given a tilt to the starboard direction, and the work of the
stabilizer was observed. The result of the stabilizer system prototype with the first setting
is shown in 0. The voltage fed to the motors is depicted in 0(a) while the roll angle is
presented in 0(b). The vertical green dashed line in 0(b) indicates the time when the roll
angle leaves the critical value of 30°. As can be inferred from 0(a), each of the starboard
and port motors was turned on alternately four times between k = 18 and k = 325. There
are four peaks for each color. This took approximately 307 time-instants or 2.46 s.
Afterward, the roll angle could be fully maintained between −10° and 10° and both motors
were never activated again.

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 229
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

.
(a)

(b)
Figure 12. Test of the Roll Stabilizer System Prototype with the First Control
Setting: (a) Voltage of DC Motors and (b) Roll Angle
The second experiment was conducted with Δa = 1 PWM level and Δb = 8 PWM
levels. The result of the stabilizer system prototype with the second setting is shown in 0.

(a)

(b)
Figure 13. Test of the Roll Stabilizer System Prototype with the Second Control
Setting; (a) Voltage of DC Motors and (b) Roll Angle
The voltage fed to the motors is depicted in 0(a) while the roll angle is presented
in 0(b). The vertical green dashed line in 0(b) indicates the time when the roll angle leaves
the critical value of 30°. The small value of Δa means that the change of motor voltage
was made more gradually when the roll angle was within the range of 10° < |x| ≤ 20°. In
total, there are only three sizable actions of the motors, compared to six in the previous

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 230
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

test which used Δa = 4 PWM levels. The starboard and port motor were active between
k = 30 and k = 170, which corresponds to 1.12 s.
The calibration process of zero offset adjustment was conducted smoothly. The
correction constant of 5.60° for the roll angle helped significantly in improving the
accuracy of the gyroscope. The obtained average angle error is 0.534°. The percent
difference between the protractor-based roll angle measurement and that of the gyroscope
for roll angles between 15o and 90o was 1.21%. With such a small percent difference
between the two set of measurements, the gyroscope was found to be accurate to measure
the roll angle in the boat roll stabilizer system prototype. In the subsequent test, the critical
roll angle was found to be approximately 30°. After the initial roll angle of 30° was
determined, the roll angle was increased further to 60°. The roll angle jumped up to
around 90° at k = 190 before going down to a small negative value at k = 194 (0(a)). This
jumping can be considered as an error due to incidental measurement outliers. This test
thus confirmed that a boat roll stabilizer is required to avoid the capsizing of the boat.
During the stabilizer test with the first control setting, after giving the initial
critical disturbance to the starboard direction, the proposed system needed 4 control
actions from the starboard motor and 3 control actions from the port motors, as can be
seen in 0(a). In 0(a), while the stabilizer worked with the second control setting, two
consecutive actions of the starboard motors occurred, as indicated by the arrows. The
reason for this is the bottom spike as pointed by the arrow in 0(b). The roll angle was
measured to be within −10° ≤ x ≤ 10° at this point and the controller instantly reset the
starboard motor voltage to zero before afterward increasing the voltage again as the roll
angle reentered the region of > 10°.
Based on the test results, it can be concluded that the stabilizer system prototype
worked well in both settings. After the boat was given a tilt of which the tilt angle
exceeded the critical angle, the controller was able to bring back the boat to the upright
position. The second setting with a smaller change rate Δa = 1 PWM took 1.12 s to bring
the boat back to the stable region. This duration (1.12s) is less compared to 2.46 s
corresponding to Δa = 4 PWM in the first setting. Indeed, the contribution of the change
rate Δb remains to be investigated. Its value was set to 8 PWM in both cases.

Conclusion
This study presented the design of a boat roll stabilizer with a change rate limiter
and an Error-Sign-based Controller (ESC). Propellers situated on the boat's port and
starboard sides make up the stabilizer design. The goal is twofold: to bring the boat back
to an upright position and to apply additional force to counter any potential heeling
moment. The suggested stabilizer system prototype was tested using a scaled-down boat
prototype. The boat is powered by two 6V DC motors, two propellers, an Arduino
microprocessor, an MPU6050 gyroscope, a motor driver, and polystyrene foam. The
critical roll angle of the boat prototype was found to be approximately 30°, confirming
that the boat needs a stabilizer to prevent it from capsizing. Subsequently, the stabilizer
system prototype was implemented in two different control settings. The control setting
with a smaller change rate (1 PWM level per sampling time) performed better than the
one with a larger change rate (4 PWM levels per sampling time). The time interval

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 231
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

between the start and the end of motor activity was reduced from 2.46 s to 1.12 s.
Nevertheless, the stabilizer system prototype was successful in performing its task to
maintain the boat’s stability in the two cases of control settings.
The proposed stabilizer system prototype is especially well-suited for protecting
boats at anchor from wind and waves when a stationary boat position is necessary, like
recreational diving, snorkeling, or fishing. In the future, the stabilizer may be more suited
for these kinds of uses, in which the propellers can be retractable to submerge them only
when necessary. It is necessary to carry out additional tests to verify that the suggested
design method and control strategy can be implemented in practice. A scaled-down boat
and actuators with an accurate model-to-real-life ratio for the boat's dimensions, mass,
motor torque, and motor mass will be needed for this. Finally, the implementation of the
proposed system in a real-life boat will be the ultimate goal of the research. For this
accomplishment, motors with adequate torque and testing in the real environment are
necessary.

References
[1] Komite Nasional Keselamatan Transportasi, “Sejarah KNKT,” Komite Nasional
Keselamatan Transportasi, 2022. [Online]. Available: https://knkt.go.id/
profile/sejarah. [Accessed: April 16, 2024].
[2] Kementerian Perhubungan Republik Indonesia, “Transportasi laut penghubung
daerah terdepan, terluar, dan tertinggal,” Kementerian Perhubungan Republik
Indonesia, 2020. [Online]. Available: https://dephub.go.id. [Accessed: April 1,
2024].
[3] JPNN, “Diterjang ombak tinggi, kapal pembawa 30 penumpang terbalik di
Kepulauan Seribu,” JPNN, March 11, 2024. [Online]. Available:
https://www.jpnn.com. [Accessed: April 1, 2024].
[4] F. Arnaz, “Perahu Tenggelam di Waduk Kedung Ombo, Dua Jadi Tersangka,”
Berita Satu, May 19, 2021. [Online]. Available: https://www.beritasatu.com.
[Accessed: April 10, 2024].
[5] M. Widiyanto, “Kronologi perahu tenggelam di Waduk Riam Kanan Banjar,
Anggota TNI AU tewas tenggelam, “ Tribun News Solo, May 24, 2021. [Online].
Available: https://solo.tribunnews.com. [Accessed: April 10, 2024].
[6] C. S. Kwon, J.-S. Kim, and H. J. Kim, “Effect of bilge radius and bilge keel height
on roll damping performance of floating structures,” in Proc. of International
Conference on Offshore Mechanics and Arctic Engineering, October 2021.
doi:10.1115/omae2021-61222
[7] H. Armono, K. Suastika, A. Ridlwan, and T. Biaperi, “Numerical study of bilge
keel length variations of floating breakwater to optimize transmission
coefficient,” in Proc. of International Seminar on Ocean and Coastal Engineering,
Environmental and Natural Disaster Management (ISOCEEN), 2019, pp.191-197.
doi:10.5220/0010060501910197
[8] Y. Wei, Y. Yang, S. Fan, Z. Liu, and P. Feng, “Model test study on anti-rolling of
ship at zero speed,” in Proc. of International Conference on Ocean, Offshore and
Arctic Engineering, December 2020. https://doi.org/10.1115/OMAE2020-19343
[9] M. Ji, H. Miao, and P. Zhao, “Design of multi-mode switch controller for fin
stabilizer,” in Proc. of IEEE International Conference on Mechatronics and
Automation (ICMA), August 2017. doi:10.1109/icma.2017.8015992
[10] H. Li, S. Lu, C. Guo, and X. Li, “Adaptive fuzzy sliding mode controller design
for ship fin stabilizer under rough sea conditions,” in Proc. of International

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 232
https://jurnal.ar-raniry.ac.id/index.php/circuit
ISSN 2549-3698 Erwin Sitompul, et al
e-ISSN 2549-3701

Conference on Information and Automation (ICIA), July 2014.

doi:10.1109/icinfa.2014.6932718
[11] Y. Wei and Y. Yang, “Research on the hydrodynamic performance of zero speed
fin stabilizer,” in Proc. of International Conference on Offshore Mechanics and
Arctic Engineering, October 2021. doi:10.1115/omae2021-62889
[12] D. Roshandel, A. Tavakolpour-Saleh, and A. Esmaeilpour, “Optimal fuzzy
control of a two-wheel mobile robot equipped with a control moment gyroscope
(CMG) stabilizer: Simulation and experiment,” Research Square, September
2023. doi:10.21203/rs.3.rs-3239206/v1
[13] C. W. Lumoindong and E. Sitompul, “A prototype of an IOT-based pet robot with
customizable functions (CoFiBot V2),” International Journal of Mechanical
Engineering and Robotics Research, pp. 510–518, 2021,
https://doi.org/10.18178/ijmerr.10.9.510-518
[14] D. T. Santoso, R. P. Sari, and F. F. Mudzakir, “Rancang bangun gryroscope
stabilizer untuk stabilisasi perahu,” Jurnal Rekayasa Mesin (JRM), vol. 16, no. 1,
pp.62-68, April 2021.
[15] M. S. Pratomo,”Rancang bangun sistem stabilizer rolling kapal penumpang
menggunakan fin stabilizer dengan kontroler PID,” Bachelor thesis, Department
of Electrical Engineering, Universitas Brawijaya, Malang, Indonesia, 2018.
[16] X. Zheng, J. Jiao, and C. Li, “Study on the influence of artillery recoil on the
lateral stability of patrol boat,” Journal of Physics: Conference Series, vol. 2381,
no. 1, p. 012112, December 2022. doi:10.1088/1742-6596/2381/1/012112
[17] M. G. Morabito, B. Brier, and S. Greene, “Preliminary stability and resistance
analysis of the cheops boat,” Journal of Ship Production and Design, vol. 36,
no. 01, pp. 14–40, February 2020. doi:10.5957/jspd.2020.36.1.14
[18] T. Rahmaji et al., “Design of fast patrol boat for improving resistance, stability,
and seakeeping performance,” Designs, vol. 6, no. 6, p. 105, November 2022.
doi:10.3390/designs6060105
[19] E. Sitompul, R. M. Putra, H. Tarigan, A. Silitonga, and I. Bukhori,
“Implementation of Digital Feedback Control with Change Rate Limiter for Water
Flow Rate Regulation Using Arduino,” Buletin Ilmiah Sarjana Teknik Elektro,
vol. 6, no. 1, pp. 72-82, 2024. doi: 10.12928/biste.v6i1.10234
[20] L. Wu and M. Sun, “A novel attracting-law based digital control strategy with
improved convergence rate and steady-state error,” in Proc. of IEEE Data Driven
Control and Learning Systems Conference (DDCLS), November 2020.
doi:10.1109/ddcls49620.2020.9275137

CIRCUIT: Jurnal Ilmiah Pendidikan Teknik Elektro, Vol.8, No.2, August 2024 | 233
https://jurnal.ar-raniry.ac.id/index.php/circuit