sensors

Article
Autonomous Water Quality Monitoring and Water Surface
Cleaning for Unmanned Surface Vehicle
Hsing-Cheng Chang, Yu-Liang Hsu * , San-Shan Hung, Guan-Ru Ou, Jia-Ron Wu and Chuan Hsu

Department of Automatic Control Engineering, Feng Chia University (FCU), No. 100, Wenhwa Road, Seatwen,
Taichung 40724, Taiwan; [email protected] (H.-C.C.); [email protected] (S.-S.H.);
[email protected] (G.-R.O.); [email protected] (J.-R.W.); [email protected] (C.H.)
* Correspondence: [email protected]; Tel.: +886-4-2451-7250 (ext. 3932)

Abstract: Water is one of the most precious resources. However, industrial development has made
water pollution a critical problem today and thus water quality monitoring and surface cleaning
are essential for water resource protection. In this study, we have used the sensor fusion technology
as a basis to develop a multi-function unmanned surface vehicle (MF-USV) for obstacle avoidance,
water-quality monitoring, and water surface cleaning. The MF-USV comprises a USV control unit, a
locomotion module, a positioning module, an obstacle avoidance module, a water quality monitoring
system, a water surface cleaning system, a communication module, a power module, and a remote
human–machine interface. We equip the MF-USV with the following functions: (1) autonomous
obstacle detection, avoidance, and navigation positioning, (2) water quality monitoring, sampling,
and positioning, (3) water surface detection and cleaning, and (4) remote navigation control and
real-time information display. The experimental results verified that when the floating garbage
located in the visual angle ranged from − 30 to 30 on the front of the MF-USV and the distances
between the floating garbage and the MF-USV were 40 and 70 cm, the success rates of floating
garbage detection are all 100%. When the distance between the floating garbage and the MF-USV
Citation: Chang, H.-C.; Hsu, Y.-L.; was 130 cm and the floating garbage was located on the left side (15 ~30 ), left front side (0 ~15 ),
Hung, S.-S.; Ou, G.-R.; Wu, J.-R.; Hsu,
front side (0 ), right front side (0 ~15 ), and the right side (15 ~30 ), the success rates of the floating
C. Autonomous Water Quality
garbage collection were 70%, 92%, 95%, 95%, and 75%, respectively. Finally, the experimental results
Monitoring and Water Surface
also verified that the applications of the MF-USV and relevant algorithms to obstacle avoidance,
Cleaning for Unmanned Surface
water quality monitoring, and water surface cleaning were effective.
Vehicle. Sensors 2021, 21, 1102.
https://doi.org/10.3390/s21041102
Keywords: unmanned surface vehicle; navigation; obstacle avoidance; water quality monitoring;
Academic Editor: Cheng-Chi Wang water surface cleaning; remote navigation control
Received: 21 December 2020
Accepted: 26 January 2021
Published: 5 February 2021
1. Introduction
Publisher’s Note: MDPI stays neutral Covering approximately 71% of the Earth’s surface, oceans and rivers are home to
with regard to jurisdictional claims in billions of types of aquatic life, yet humans have not treated the aquatic environment in a
published maps and institutional affil- friendly manner [ 1]. Because of human negligence, water pollution has accumulated for
iations.
decades. The pollutants in water include industrial waste, sewage, radioactive materials,
and plastic waste [ 2]. Therefore, it is necessary for researchers to develop unmanned sur-
face vehicles (USVs), autonomous surface vehicles (ASVs), and autonomous underwater
vehicles (AUVs) with lower development costs and higher resolution in time and space
Copyright: © 2021 by the authors. measurements for hydrographic measurement [ 3], path planning and navigation [ 4], obsta-
Licensee MDPI, Basel, Switzerland. cle avoidance [ 5], water quality monitoring [ 6], and water surface cleaning [ 7]. In terms of
This article is an open access article obstacle detection and avoidance, Villa et al. [ 5] designed a USV with a light detection and
distributed under the terms and ranging (LiDAR) model to provide the obstacle information in Cartesian coordinates and
conditions of the Creative Commons
proposed a line-of-sight (LOS)-based guidance, navigation, and control (GNC) algorithm
Attribution (CC BY) license (https://
to avoid obstacles. Wang et al. [ 8] used a navigation radar embedded on the USV to
creativecommons.org/licenses/by/
detect unknown obstacles and proposed an improved ant colony optimization (IACO)
4.0/).

Sensors 2021, 21, 1102. https://doi.org/10.3390/s21041102 https://www.mdpi.com/journal/sensors

Sensors 2021, 21, 1102 2 of 21

algorithm for avoiding collisions. Kim et al. [9] presented a deep learning-based obstacle
segmentation algorithm, called Skip-ENet, to recognize obstacles in real time by using
vision sensors embedded on the ASV. Steccanella et al. [10] utilized a U-net convolutional
neural network (CNN) based method to segment the image taken by the camera placed on
the low-cost ASV for waterline and obstacle detection. Guardeño et al. [11] used a LiDAR
sensor mounted on the USV to detect static obstacles and developed a new autotuning
environment for static obstacle avoidance (ATESOA) algorithm for reactive static obstacle
avoidance.
Recently, a number of researchers have developed novel ASV or AUV systems for long-
term continuous water quality monitoring and seawater sampling [6,12–15]. For example,
Li et al. [12] implemented a water color remote sensing-oriented USV (WC-USV), which is
composed of an unmanned surface vehicle platform with a ground control station, a data
acquisition module, and a transmission module, to execute autonomous navigation and
obstacle avoidance, water sample collection, water quality measurement, meteorological
information measurement, and remote control tasks. Ferri et al. [6] developed a small-sized
ASV, called a HydroNet ASV, to measure hydrocarbon and heavy metal concentrations
by using custom-made miniaturized sensors and collect water samples. In addition, a
VFH+ algorithm was utilized to avoid possible collisions based on the measurements
of the laser scanner, sonar, and altimeter sensors placed on the HydroNet ASV. Madeo
et al. [13] presented a low-cost USV, named the water environmental mobile observer
(WeMo), to provide a modular array of sensors for measuring the pH value, oxidation-
reduction parameter (ORP), salinity, and dissolved oxygen. Cao et al. [14] proposed an
intelligent wide-area water quality monitoring and online analysis system to analyze the
pH, total dissolved solids (TDS), and turbidity values for estimating the water quality
level, which comprises an automatic cruise intelligent USV, a water quality monitoring
system (WQMS), and a water quality analysis algorithm. Cryer et al. [15] presented an
ASV, named the C-Worker 4 (CW4), equipped with the conductivity, temperature, depth
(CTD), dissolved O2 (DO), pH, pCO2 , nitrate, chlorophyll fluorescence, coloured dissolved
organic matter (CDOM), turbidity, and dissolved organic carbon (DOC) sensors, to collect
high-resolution measurements in shallow coastal environments. Autonomous vehicles
expand the monitoring range in both time and space and increase the possibility of real-time
marine and continuous weather monitoring. They have compatible computing systems for
smart sampling (e.g., adaptive sampling [16] and optimal sampling [17]) with which not
all current vessels are equipped. Additionally, the monitoring efficiency can be enhanced
through coordination between autonomous vehicles with different features for different
uses [18]. These advantages are the reasons why autonomous vehicles are widely employed
in autonomous, complex, and distributed water quality monitoring systems. In particular,
ASVs, capable of remote tasks and safe navigation, can be regarded as low-cost assets for
water monitoring conducted using mobile sampling methods. They are also a cheaper and
simpler monitoring apparatus compared with AUVs [6].
With the extended application of robots in recent years, many researchers have at-
tempted to develop robots for cleaning regional environments [19–22]. However, those
designed to clear garbage from water surfaces require further development [7]. Although
semimanual garbage cleaning vessels are commonly used to collect plastic waste on water
surfaces, they are large and thus restricted to use in rivers with larger areas or heavier
waste. Therefore, using garbage cleaning vessels to remove low-density waste from small
rivers is impractical. Moreover, such vessels are disadvantaged by their inability to identify
which object to remove as well as by them potentially causing secondary pollution through
discharged exhaust. Some intelligent cleaner robots have been developed for collecting
floating garbage on water surfaces. For example, Kong et al. [7] developed an intelligent
water surface cleaner robot (IWSCR) system with a vision module, a motion control module,
and a grasping module, which utilized the YOLOv3 for garbage detection, the sliding-
mode controller for vision-based tracking and steering, and the feasible grasping strategy
for floating garbage grasping and collection. Wang et al. [23] proposed an autonomous
Sensors 2021, 21, 1102 3 of 21

robot to clean rubbish floating on a lake surface and utilized the Maneuvering Model
Group (MMG) model approach to model hydrodynamics of the robot. Ruangpayoongsak
et al. [24] developed a floating waste scooper robot for collecting plastic bottles floating on
a water surface. Li et al. [25] designed a vision-based water surface garbage capture robot
with the modified YOLOv3 algorithm for real-time garbage detection in dynamic aquatic
environments.
In this study, we have designed a multi-function unmanned surface vehicle (MF-USV)
with obstacle avoidance, water quality monitoring, and water surface cleaning functions.
The MF-USV is a robotic system which is composed of a USV control unit, a locomotion
module, a positioning module, an obstacle avoidance module, a water quality monitoring
system, a water surface cleaning system, a communication module, a power module,
and a remote human–machine interface. The MF-USV is equipped with the following
functions: (1) autonomous obstacle detection, avoidance, and navigation positioning:
navigating the vehicle by using the USV control unit to control the locomotion module,
adopting the obstacle avoidance module to detect and avoid obstacles, and employing the
positioning module to record the movement trajectory of the MF-USV. (2) Water quality
monitoring, sampling, and positioning: detecting real-time water pH values, collecting
water samples, and positioning the sampling points by using the water quality monitoring
system. (3) Water surface detection and cleaning: detecting and collecting floating garbage
on water surfaces by using the water surface cleaning system. (4) Remote navigation
control and real-time information display: enabling monitors to navigate the MF-USV
through the Bluetooth wireless module. In addition, the movement trajectory recorded by
the global positioning system (GPS), water pH values, and positions of sampling points
can be transmitted back through the communication module in real time and displayed
on the remote human–machine interface. Finally, functions, types of sensors, and related
obstacle avoidance, water quality monitoring, and water surface cleaning algorithms are
information extracted from the existing published papers and summarized in Table 1 for
straightforward comparison. Obviously, most of the existing USVs or ASVs focus on only
one function of obstacle avoidance, water quality monitoring, and water surface cleaning.
Hence, the purpose of this paper is to develop a MF-USV which can perform the obstacle
avoidance, water quality monitoring, and water surface cleaning functions simultaneously.

Table 1. Function comparison of the proposed multi-function unmanned surface vehicle (MF-USV) with some existing
USVs/autonomous surface vehicles (ASVs).

Reference Function Sensor Algorithm

Villa et al. [5] Obstacle avoidance LiDAR LOS-based GNC Algorithm
Wang et al. [8] Obstacle avoidance Navigation radar IACO Algorithm
Kim et al. [9] Obstacle detection Vision sensors Skip-ENet Algorithm
Waterline and obstacle
Steccanella et al. [10] Camera U-net CNN Algorithm
detection
Guardeño et al. [11] Obstacle avoidance LiDAR ATESOA algorithm
Obstacle avoidance Microwave radar, 4G Camera –
Li et al. [12]
Chl-a turbidity, water dissolved
oxygen, water conductivity,
Water quality monitoring –
oxidation-reduction potential, water
temperature, salinity, pH sensors
Obstacle avoidance Laser scanner, Sonar, Altimeter VFH+ algorithm
Ferri et al. [6]
Chemical sensors (Hg, Cr, Cd, and
Water quality monitoring –
dispersed oil)
pH sensor, ORP sensor, salinity
Madeo et al. [13] Water quality monitoring –
sensor, dissolved oxygen probe
pH sensor, TDS sensor, turbidity
Cao et al. [14] Water quality monitoring Ensemble learning algorithm
sensor
Sensors 2021, 21, 1102 4 of 21

Table 1. Cont.

Reference Function Sensor Algorithm

Sensors 2021, 21, x FOR PEER REVIEW CTD, DO, pH, pCO2, nitrate, 4 of 21
Cryer et al. [15] Water quality monitoring chlorophyll fluorescence, CDOM, –
turbidity, DOC sensors
Kong et al. [7] Water surface cleaning Vision module, Grasper YOLOv3 algorithm
sors
Wangetetal.
Kong al.[7]
[23] Water
Water surface
surface cleaning
cleaning Vision module,belt
Conveyor Grasper YOLOv3– algorithm
Wang et al. [23]
Ruangpayoongsak Water surface cleaning Conveyor belt --
Water surface cleaning Camera, waste scooper –
Ruangpayoongsak
et al. [24] et
Water surface cleaning Camera, waste scooper --
Lial.et[24]
al. [25] Water surface cleaning Camera, collection box YOLOv3 algorithm
Li et al. [25] Water surface cleaning Camera, collection box YOLOv3 algorithm
Threshold-based obstacle
Obstacle avoidance Ultrasonic Sensors Threshold-based obstacle
Obstacle avoidance Ultrasonic Sensors avoidance algorithm
Proposed MF-USV avoidance algorithm
Proposed MF-USV Water
Water quality
quality monitoring
monitoring pH sensor
pH sensor –--
Hue-based
Hue-based color
color filtering
filtering algo-
Water
Water surface
surface cleaning
cleaning Vision sensor, salvage net
Vision algorithm
rithm

The rest of
The rest of this
thispaper
paperisisorganized
organizedasasfollows.
follows.
InIn Section
Section 2, we
2, we introduce
introduce the the proto-
prototype
type design of the proposed USV. The water quality monitoring and
design of the proposed USV. The water quality monitoring and water surface cleaning water surface
cleaning
methods methods are then described
are then described in Sectionsin3Sections 3 and 4, respectively.
and 4, respectively. The experimental
The experimental results are
results are in
presented presented
Section 5.inFinally,
Sectionconclusions
5. Finally, conclusions
are given inare given6.in Section 6.
Section

Prototype Design
2. Prototype Design of of Multi-Function Unmanned Surface Vehicle (MF-USV) (MF-USV)
The overall
overall architecture
architecturedesign
designofofthe proposed
the proposed MF-USV
MF-USV is shown
is shownin Figure 1. The
in Figure MF-
1. The
USV weights
MF-USV 3.025
weights kg, kg,
3.025 andandits dimension
its dimension is 0.74 mm
is 0.74 (length)
(length) × 0.30.3
mm (width)
(width) × 0.280.28
m
(height).
m TheThe
(height). draft of the
draft of USV is 0.15
the USV is m and
0.15 mitsandmaximum cruise cruise
its maximum speed is 0.15 m/s.
speed is 0.15The MF-
m/sec.
USVMF-USV
The architecture is composed
architecture of: (1) a main
is composed of: (1)control
a main unit,
control(2) unit,
a locomotion module,mod-
(2) a locomotion (3) a
positioning module, (4) an obstacle avoidance module, (5) a water quality
ule, (3) a positioning module, (4) an obstacle avoidance module, (5) a water quality mon- monitoring
system,system,
itoring (6) a water
(6) asurface cleaningcleaning
water surface system,system,
(7) a communication
(7) a communication module,module,
(8) a power
(8) a
module, and (9) a remote human–machine interface, as shown
power module, and (9) a remote human–machine interface, as shown in Figure in Figure 2. The different
2. The
modules modules
different of the proposed MF-USV MF-USV
of the proposed are detailedarein Table 2.in Table 2.
detailed

Figure 1. Overall view of the positions of all components inside the MF-USV.
Figure 1. Overall view of the positions of all components inside the MF-USV.
Sensors 2021, 21, 1102 5 of 21

Sensors 2021, 21, x FOR PEER REVIEW 5 of 21

Figure 2.
Figure 2. System
System architecture of the
architecture of the MF-USV.
MF-USV.

Table 2.
Table 2. Components
Components of
of the
the MF-USV
MF-USV with
with their
their weight
weight and
and power
power consumption.
consumption.

Estimate of the
Estimate ofAver-
the Average
Components
Components Description
Description Weight (g)
Weight (g) age Electric Power
Electric Power Notes
Notes
Consumption (W) (W)
Consumption
2× ATmega2560 (Arduino Mega2560)
2 × ATmega2560 (Arduino
AVR® 8-Bit Microcontroller at 16 MHz
Mega2560)
Communication
AVR®8-Bitinterface: Universal at 16
Microcontroller
Main Controller AsynchronousMHz Receiver Transmitter 74 0.4
Communication Circuit
(UART), Inter-Integrated interface:
(I2C),
Main Controller Universal Asynchronous
Serial Peripheral Interface (SPI), Gen-Receiver 74 0.4
Transmitter (UART),
eral Purpose Input/Output (GPIO)2
Inter-Integrated Circuit (I C),
Propulsion: HMS-RF240 DC motor
Serial Peripheral Interface (SPI),
Rudder motor: MG996R Servo motor Maximum propeller speed: 15
Locomotion General Purpose Input/Output
Motor drive module: KEYES L298N
(GPIO) 135 5.52 cm/s
Module
Electric propeller Maximum rudder speed: 300°/s
Propulsion: HMS-RF240 DC motor
Rudder and actuation mechanism
Rudder motor: MG996R Servo
GY-GPS6MV2 Maximum propeller
Positioning Mod- motor
NEO-6M-5Hz speed: 15 cm/s
Locomotion
ule Module MotorGPS
drive module: KEYES 18 135 0.14 5.52
Antenna L298N
GPS Maximum rudder
speed: 300◦ /s
Electric propeller Microcontroller collects data
Rudder and actuation mechanism from the ultrasonic sensors,
Ultrasonic Sensor 2× HC-SR04 17 0.02
GY-GPS6MV2 execute the obstacle detection
Positioning Module NEO-6M-5Hz GPS 18 0.14 and avoidance algorithm
SEN0161Antenna GPS
pH Sensor 75 0.2
pH sampling probe Microcontroller
R385 Water pump collects data from
Water Sample Total volume of two sampling
2×Electromagnetic valve 387 1.7 the ultrasonic
Collection Device bottles: 60 ml
Ultrasonic Sensor
2×Sampling2 ×bottle
HC-SR04 17 0.02 sensors, execute the
Pixy CMUcam5 obstacle detection
Processor: NXP LPC4330 at 204 MHz and avoidance
algorithm
Communication interface: UART, I2C,
Vision Sensor SEN0161 27 0.7
pH Sensor SPI, USB 75 0.2
pH sampling probe
Image sensor: Omnivision OV9715
2×GS-9018 Servo motor
Water Surface 2×HN-35GBE-1640Y DC gear motor 1159 0.38
Sensors 2021, 21, 1102 6 of 21

Table 2. Cont.

Estimate of the Average

Components Description Weight (g) Electric Power Notes
Consumption (W)
R385 Water pump Total volume of two
Water Sample
2 × Electromagnetic valve 387 1.7 sampling bottles: 60
Collection Device
2 × Sampling bottle ml
Pixy CMUcam5
Processor: NXP LPC4330 at 204
MHz
Vision Sensor Communication interface: UART, 27 0.7
I2 C, SPI, USB
Image sensor: Omnivision OV9715
2 × GS-9018 Servo motor
2 × HN-35GBE-1640Y DC gear
motor
Water Surface Cleaning
Motor drive module: KEYES 1159 0.38
Device
L298N
Cleaning device: Salvage net
Communication 2×HC-08 Bluetooth wireless
7 0.09
Module module
Plastic boxes with waterproof
Cabling and Boxes 233 –
connectors and glands
Hulls and Deck 1073 –
3205
Total (without 9.15
batteries)

2.1. USV Main Control Unit

The main control unit used in the MF-USV are two Arduino MEGA 2560 microcon-
trollers which are based on an AVR®8-Bit Microcontroller at 16 MHz. The first Arduino
MEGA 2560 microcontroller is responsible for locomotion, obstacle avoidance, and water
surface cleaning tasks, while the second one is responsible for navigation positioning and
water quality monitoring tasks. The MF-USV control unit is directly connected to the
locomotion module, positioning module, obstacle avoidance module, water quality moni-
toring system, water surface cleaning system, communication module, and power module.
Moreover, the MF-USV can transmit the multi-sensing data to the remote human–machine
interface through two Bluetooth wireless modules for monitoring the movement trajectory
of the MF-USV, the water pH values, and the water sampling positions.

2.2. Locomotion Module

The locomotion module mounted on the MF-USV is composed of an electric propeller
controlled by a direct current (DC) motor (HMS-RF-240) and a rudder controlled by a
servo motor (MG996R), as shown in Figure 3. The DC and servo motors are both driven
by a motor drive module (L298N) through the Arduino MEGA 2560 microcontroller.
The DC motor connects with the electric propeller through the coupling for controlling the
propulsion speed. The servo motor controls the steering blade connected to the rudder for
controlling the propulsion direction.
 peller controlled by a direct current (DC) motor (HMS-RF-240) and a rudder controlled
by a servo motor (MG996R), as shown in Figure 3. The DC and servo motors are both
driven by a motor drive module (L298N) through the Arduino MEGA 2560 microcon-
troller. The DC motor connects with the electric propeller through the coupling for con-
Sensors 2021, 21, 1102
trolling the propulsion speed. The servo motor controls the steering blade connected to
7 of 21
the rudder for controlling the propulsion direction.

(a) (b) (c)

Figure 3.3. (a)
Figure (a)Electric
Electricpropeller
propellerand
andrudder.
rudder.
(b)(b) Computer
Computer aided
aided design
design (CAD)
(CAD) model
model of locomotion
of the the locomotion module.
module. (c) Lo-
(c) Locomo-
comotion module mounted on
tion module mounted on the MF-USV. the MF-USV.

2.3. Positioning Module

2.3. Positioning
The positioningModulemodule mounted on the MF-USV is the GY-GPS6MV2 GPS Module
The positioning
(as shown in Table 3),module
which mounted
is used toon the MF-USV
generate is the GY-GPS6MV2
the movement trajectory ofGPS Module
the MF-USV
Sensors2021,
Sensors 2021,21,
21,xxxFOR
FORPEER
PEERREVIEW
REVIEW 777of
of21
21
Sensors 2021, 21, FOR PEER (as
REVIEW shown in Table 3), which is used to generate the movement
during the voyage and record the position coordinates of the water sampling points. The
Sensors 2021, 21, x FOR PEER REVIEW trajectory of the MF-USV of 21
7 of 21
during
GPS module the voyageutilizesandtherecord
NEO-6M the position
GPS receivercoordinates of the water
manufactured sampling
by U-Blox points.
to provide
The GPS module
time-to-first-fix utilizes
(TTFF) the NEO-6M
acquisition of lessGPSthan receiver
1 second, manufactured
which is composedby U-Blox
of antoantennae
provide
time-to-first-fix
and
and aa radio
radio
and aa radio (TTFF)
frequency
frequency
radio frequency acquisition
frequency (RF) (RF)
(RF) of less
satellite
satellite
(RF) satellite than
signal
signal
satellite signal 1 second,
receiver,
receiver,
signal receiver,
receiver, an which
an
an is composed
interface
interface
an interface device
device
interface device of an antennae
equipped
equipped
equipped with
device equipped with aaa
with
and
andmicroprocessor,
a radio frequency (RF) satellite signal receiver, an interface device equipped with
with a
microprocessor,
microprocessor, and and
and a
a base
base
and aa base band
band
base band processor
processor
band processor
processor for for
for processing
processing
for processing
processing GPS GPS
GPS signal.
signal.
GPS signal. The
The
signal. The Arduino
Arduino
The Arduino
Arduino MEGA MEGA
MEGA
MEGA
microprocessor,
a microprocessor, and acanbase band
2560 microcontroller
2560
2560 microcontroller
microcontroller can
can read
read
read GPSprocessor
GPS
GPS coordinates
coordinates
coordinates forof processing
of
of the MF-USV
the
the MF-USV
MF-USV GPS signal.
and
and
and waterThe
water
water Arduino
sampling
sampling
sampling points
points
points
MEGA2560 microcontroller
2560 can
microcontroller read
can GPS
read coordinates
GPS coordinates of theof MF-USV
the MF-USV and water
and sampling
water samplingpoints
through the
through
through the UART
the UART protocol
UART protocol and
protocol and then
and then send
then send them
send them to
them to the
to the remote
the remote human–machine
remote human–machine interface
human–machine interface
interface
through
points through the UART
the protocol
UART and
protocol then
and in send
then them
send to
them the toremote
the human–machine
remote human–machine interface
through
through
through the the
the communication
communication
the communication
communication module module
module
module in in real
real
in real time.
time.
real time.
time. For For
For our
our propose,
propose,
For our propose,
propose, we we
we used
used
we used the
the Recom-
the Recom-
used the Recom-
through
interface through the communication module in real time.our Foroutput
our propose, weto used Recom-
the the
mended
mended
mended MinimumMinimum
Minimum Specific
Specific
Minimum Specific GPS/TRANSIT
GPS/TRANSIT
Specific GPS/TRANSIT
GPS/TRANSIT Data Data
Data ($GPRMC)
($GPRMC)
Data ($GPRMC)
($GPRMC) output output sentence
sentence
output sentence
sentence toto extract
extract
extract the
to extract the
mended
Recommended Minimum Specific (latitude
GPS/TRANSIT Data ($GPRMC) output sentence to the
travel time
travel
travel time and
time and
and local coordinates
local
local coordinates
coordinates (latitude and
(latitude and longitude).
and longitude).
longitude).
travel
extract thetime andtime
travel localand
coordinates (latitude (latitude
local coordinates and longitude).and longitude).
Table 3.
Table
Table 3. Parameters
3. Parameters of
Parameters of sensors
of sensors mounted
sensors mounted on
mounted on the
on the MF-USV.
the MF-USV.
MF-USV.
Table
Table 3. Parameters
3. Parameters of sensors
of sensors mounted
mounted on
on the the MF-USV.
MF-USV.
Working
Working
Working
Sensor Type
Sensor
Sensor Type
Type Image
Image
Image Range
Range
Range Working
Working Accuracy
Accuracy
Accuracy Producer
Producer
Producer
Sensor Type
Sensor Type Image
Image Range
Range Voltage (V)
Voltage
Voltage (V)
(V) Accuracy
Accuracy Producer
Producer
Voltage (V) (V)
Voltage
Position: 2.5
Position:
Position: 2.5 m
2.5 m
m
Position:
Position: 2.5
2.5 m m
GPS
GPS
GPS ≥−161 dBm
≥−161
≥−161 dBm
dBm 3.3~5.0
3.3~5.0
3.3~5.0 Velocity:
Velocity: 0.1
0.1
Velocity:0.1
0.1 m/s
m/s
m/s U-BLOX 111
U-BLOX
U-BLOX
1
GPS
GPS ≥−161
≥−161 dBmdBm 3.3~5.0
3.3~5.0 Velocity:
Velocity: 0.1
m/sm/s U-BLOX
U-BLOX 1
Heading:
Heading:
Heading:0.5 0.5°
0.5°
0.5°
◦
Heading:
Heading: 0.5°
Ultrasonic
Ultrasonic
Ultrasonic
Ultrasonic 2~400
2~400
2~400
2~400 cm
cm cm
cm 3.3~5.0
3.3~5.0
3.3~5.0
3.3~5.0 0.3cm
0.3
0.3
0.3 cm
cm
cm ELECFREAKS
ELECFREAKS
2
ELECFREAKS
ELECFREAKS 22 2
Ultrasonic 2~400 cm 3.3~5.0 0.3 cm ELECFREAKS2

pH
pH
pH
pH 0~14
0~140~14
0~14 5.0
5.05.0
5.0 ±0.1
±±0.1
±0.1 pH
pH
0.1 pH
pH DFROBOT
3
DFROBOT
DFROBOT
DFROBOT 333
pH 0~14 5.0 ±0.1 pH DFROBOT3

Lens
Lens
Lens
Lens field
fieldfield
field of view:
of
of view:
of view:
view:
Lens field of
◦ horizontal view: Resolution:
Resolution:
Resolution:
Resolution: 4
Vision
Vision
Vision
Vision ±
±37.5°
37.5
±37.5° horizontal
horizontal ±23.5° 6.0~10.0
horizontal ±23.5°
±37.5° horizontal ±23.5° 6.0~10.0
6.0~10.0
6.0~10.0 Resolution: PIXYCAM
PIXYCAM
PIXYCAM
PIXYCAM 444
Vision ±37.5°
±23.5◦ vertical
vertical ±23.5° 6.0~10.0 1280×××800
1280
1280 800dpi
800
1280 ×× 800 dpi
dpi
800 dpi
dpi PIXYCAM 4
vertical
vertical 1280
vertical
Note:1111Thalwil,
Note:
Note:
Note: Thalwil,Switzerland;
Thalwil,
Thalwil, Switzerland;
Switzerland;
Switzerland; Shenzhen,
222 Shenzhen,
2 Shenzhen, China;
China;
China;
Shenzhen, Shanghai,
333 Shanghai,
3 Shanghai,
China; China;
China;China;
Shanghai, 4 Texas,
China; Texas, USA.
444 Texas,
USA.
Texas, USA.
USA.
Note: 1 Thalwil, Switzerland; 2 Shenzhen, China; 3 Shanghai, China; 4 Texas, USA.
2.4.
2.4.2.4. Obstacle
Obstacle
Obstacle
2.4. Avoidance
Avoidance
Avoidance
Obstacle Avoidance Module
Module
Module
Module
2.4. Obstacle Avoidance Module
The
TheThe obstacle
obstacle
obstacle
The obstacle avoidance
avoidance
avoidance
avoidance module
module
module
module usedinin
used
used
used inthis
in thispaper
this
this paperisis
paper
paper iscomposed
is composed of
composed
composed of two
of two HC-SR04
HC-SR04 ul-
HC-SR04
HC-SR04 ul-
ul-
The obstacle avoidance module used in this paper is composed of two HC-SR04 ul-
trasonic
trasonic
ultrasonic sensors
sensors
sensors
trasonic sensors (as
(as
(as
sensors (as shown
shown
shown
(as shown
shown in in
in
in Table
Table
Table
in Table 3),
3),
3),
Table 3), which
which
which
3), which
which are are
are mounted
mounted
are mounted on on
on
on thethe
the left
left
left and
and
and
the left and right right
right
right front
front
front
right front
front of of
of
of the
the
of the
the
trasonic
MF-USV
theMF-USV
MF-USVand
MF-USV and
and provide
provideaaaasensor
andprovide
provide sensor
sensor detecting
sensordetecting
detecting15
detecting 15
15 degrees
15degrees
degreesin
degrees ininfront
in front
frontof
front ofofit,
of it,it,as
it, asasshown
as shown
shownin
shown ininFigure
in Figure
Figure4.
Figure 4.4.The
4. The
The
MF-USV and provide a sensor detecting 15 degrees in front of it, as shown in Figure 4. The
Arduino
TheArduino
Arduino MEGA
ArduinoMEGAMEGA2560
MEGA 2560
2560 microcontroller
2560microcontroller
microcontrollercan
microcontroller can
can obtain
canobtain
obtainthe
obtain the distance
thedistance
the between
distancebetween
distance betweenthe
between the vehicle
thevehicle
the vehicleand
vehicle and
andob-
and ob-
ob-
Arduino MEGA 2560 microcontroller can obtain the distance between the vehicle and ob-
stacles
stacles
obstacles
stacles inin
in real
real
inreal time,
time,
realtime, which
which
time,which
whichare are
are detected
detected
detectedbyby
aredetected by
by the
thethe
the ultrasonic
ultrasonic
ultrasonic
ultrasonic sensors.
sensors.
sensors.
sensors. OnceOnce
Once
Once the
thethe distance
distance
distance
the distanceis is is
is
stacles in real time, which are detected by the ultrasonic sensors. Once the distance is
greater
greater
greater or less
or
or less
greater or less
than
less than
than
than the
thethe threshold
threshold
threshold
the threshold values,
values,
values,
values, the
thethe MF-USV
MF-USV
MF-USV
the MF-USV executes
executes
executes
executes aaa threshold-based
threshold-based
a threshold-based
threshold-based obstacle
obstacle
obstacle
obstacle
greater or less than the threshold values, the MF-USV executes a threshold-based obstacle
avoidance
avoidance algorithm
algorithm
avoidance algorithm
algorithm to to
to avoid
avoid
to avoid obstacle
obstacle
avoid obstacle collision.
collision.
obstacle collision.
collision. The The
The proposed
proposed
The proposed threshold-based
threshold-based
proposed threshold-based
threshold-based obstacle obstacle
obstacle
obstacle
avoidance
avoidance
avoidance algorithm
algorithm
avoidance algorithm
algorithm is is
is shown
shown
is shown
shown in in
in Figure
Figure
in Figure
Figure 5. 5.
5. The
The
5. The
The first first
first of
of
first of the
the
of the threshold-based
threshold-based
the threshold-based
threshold-based algorithms algorithms
algorithms
algorithms cal- cal-
cal-
avoidance cal-
culates
culates the
culates the distances
the distances
distances of of the
of the obstacle
the obstacle
obstacle in in the
in the left
the left
left andand right
and right front
right front
front of of
of the the vehicle
the vehicle
vehicle (D (D and
(DLLL and
L
and D DDRRR)))
culates the distances of the obstacle in the left and right front of the vehicle (D and D R)
using
using the
using the ultrasonic
the ultrasonic sensors
ultrasonic sensors mounted
sensors mounted
mounted on on
on thethe
the left left and
left and
and rightright front
right front
front of of the
of the MF-USV.
the MF-USV.
MF-USV. Subse- Subse-
Subse-
using the ultrasonic sensors mounted on the left and right front of the MF-USV. Subse-
quently, there
quently, there are
are four
four situations
situations for for obstacle
obstacle collision
collision avoidance:
avoidance: (1) (1) go
go straight
straight ahead:
ahead:
Sensors 2021, 21, 1102 8 of 21

avoidance algorithm to avoid obstacle collision. The proposed threshold-based obstacle

avoidance algorithm is shown in Figure 5. The first of the threshold-based algorithms
calculates the distances of the obstacle in the left and right front of the vehicle (DL and
DR ) using the ultrasonic sensors mounted on the left and right front of the MF-USV.
Subsequently, there are four situations for obstacle collision avoidance: (1) go straight
ahead: when (DL > 65 cm or DL ≤ 10 cm) and (DR > 65 cm or DR ≤ 10 cm), the
algorithm determines that there is no any obstacle on the front of the vehicle, and then
the MF-USV goes straight ahead. (2) Turn right: when (10 cm < DL ≤ 65 cm) and
(DR > 65 cm or DR ≤ 10 cm), the algorithm determines that there is an obstacle on the
left front side of the vehicle, and then the MF-USV drives the servo motor (MG996R) in the
locomotion module to control the rudder to turn right. (3) Turn left: when (DL > 65 cm
or DL ≤ 10 cm) and (10 cm < DR ≤ 65 cm), the algorithm determines that there is an
obstacle on the right front side of the vehicle, and then the MF-USV drives the servo motor
(MG996R) in the locomotion module to control the rudder to turn left. (4) Turn around:
when (10 cm < DL ≤ 65 cm) and (10 cm < DR ≤ 65 cm), the algorithm determines that
Sensors
Sensors 2021, 21, xx FOR
2021, 21, FORPEER there is an obstacle on the front of the vehicle, and then the MF-USV drives the8 of
PEERREVIEW
REVIEW servo motor
21 8 of 21
(MG996R) in the locomotion module to control the rudder to turn around 180◦ .

Figure 4. Ultrasonic sensors mounted on the MF-USV.

Figure 4. Ultrasonic sensors mounted on the MF-USV.
Figure 4. Ultrasonic sensors mounted on the MF-USV.

Figure 5. Threshold-based obstacle avoidance algorithm.

2.5. Power Module

Figure and Communication
5. Threshold-based Module
obstacle avoidance algorithm.
Figure 5. Threshold-based obstacle avoidance algorithm.
The MF-USV is powered by a 6 V lead-acid battery with a nominal capacity of 2.3
2.5. Power Module
ampere-hour and communication
(Ah). The Communication Module
module includes two HC-08 Bluetooth wireless
modules,Thewhich
MF-USV haveis low-power
powered bydata
a 6 transmission propertywith
V lead-acid battery and aare operated
nominal in the of 2.3
capacity
V4.0
ampere-hour (Ah). The communication module includes two HC-08 Bluetooth the
BLE Bluetooth protocol. The Arduino MEGA 2560 microcontroller embedded on wireless
Sensors 2021, 21, 1102 9 of 21

2.5. Power Module and Communication Module

The MF-USV is powered by a 6 V lead-acid battery with a nominal capacity of 2.3
ampere-hour (Ah). The communication module includes two HC-08 Bluetooth wireless
modules, which have low-power data transmission property and are operated in the V4.0
BLE Bluetooth protocol. The Arduino MEGA 2560 microcontroller embedded on the MF-
USV transmits the GPS coordinates of the MF-USV and water sampling points, and the
water pH values in different water sampling points to the remote human–machine interface
through the Bluetooth wireless modules.

2.6. Remote Human–Machine Interface

Sensors 2021, 21, x FOR PEER REVIEW 9 of 21
The remote human-machine interface shown in Figure 6 is developed with the Lab-
VIEW graphical programming environment, which can receive the GPS coordinates of the
MF-USV and water sampling points, and the water pH values in different water sampling
human-machine
points interfacethrough
from the MF-USV is dividedtheinto the following
Bluetooth wirelessoperation
modules. interfaces:
The remote (1)human-
system
operating interface: the user can choose a series COM port to receive
machine interface is divided into the following operation interfaces: (1) system operating the measurement
signals and
interface: the set
usera can
storage
choose path for storing
a series COM port them. In addition,
to receive the travel signals
the measurement time ofand the
MF-USV
set is also
a storage pathdisplayed.
for storing (2)them.
Manual control interface:
In addition, thetime
the travel userofcan
themanually
MF-USV control
is also
displayed.
the MF-USV(2)toManual
go ahead, control interface:
turn left, the user
turn right, and can
turnmanually control the MF-USV
around, respectively. to
In addition,
go
theahead,
user canturn left,the
click turn right,
stop andto
button turn
stoparound, respectively.
the MF-USV sailing.In(3)addition,
Movement the trajectory
user can
click the stop
interface: thisbutton
interfaceto stop
canthe MF-USV
display thesailing. (3) Movement
movement trajectorytrajectory interface:
during sailing of this
the
interface can display the movement trajectory during sailing of the
MF-USV in real time via the position coordinates measured by the GPS positioning mod- MF-USV in real time
via
ule.the
(4)position coordinates
Water quality measured
monitoring by the GPS
interface: the pHpositioning
value and module. (4) Water quality
local coordinate of the
monitoring interface:
present position measuredthe pHbyvalue
the pHandsensor
local coordinate of the present
and GPS module positionInmeasured
are displayed. addition,
by
thethe pH sensor
position and GPS
coordinates ofmodule
the wateraresampling
displayed. In addition,
points are alsothe position
recorded coordinates
and displayedof in
the water sampling
this interface. points are also recorded and displayed in this interface.

Figure6.6.Remote
Figure Remotehuman–machine
human–machineinterface.
interface.

3. Water Quality Monitoring System

3. Water Quality Monitoring System
The water quality monitoring system consists of a SEN0161 pH sensor (as shown in
The water quality monitoring system consists of a SEN0161 pH sensor (as shown in
Table 3),aaGPS
Table 3), GPSmodule,
module,andandaawater
watersample
samplecollection
collectiondevice,
device, as
as shown
shown in
inFigure
Figure7.7.The
The
water sample collection device is designed for collecting water samples, which is
water sample collection device is designed for collecting water samples, which is composedcom-
posed of a water pump (R385), two electromagnetic valves, and two sampling bottles, as
shown in Figure 8. Figure 9 shows the circuit diagram of the water quality monitoring
system mounted on the MF-USV. The water quality monitoring process is shown in
Figure 10. The pH sensor mounted on the MF-USV is used to detect the pH value of the
water during sailing, which is then sent to the remote human–machine interface in real
 Sensors 2021, 21, 1102 10 of 21

of a water pump (R385), two electromagnetic valves, and two sampling bottles, as shown
in Figure 8. Figure 9 shows the circuit diagram of the water quality monitoring system
mounted on the MF-USV. The water quality monitoring process is shown in Figure 10.
The pH sensor mounted on the MF-USV is used to detect the pH value of the water
during sailing, which is then sent to the remote human–machine interface in real time.
According to the surface water classification and water quality standards of the Taiwan
Environmental Protection Administration (EPA), the pH values of the water quality should
be between 6.5 and 8.5, which comply with the class A terrestrial surface water quality
standards [26]. Once the water pH value is higher than 8.5 or lower than 6.5, that means
that the water on the present position is contaminated, the water pump is driven to extract
Sensors 2021, 21, x FOR PEER REVIEW the water for cleaning residual liquid in the pumping pipe. Subsequently, 10 of 21 the water in
Sensors 2021, 21, x FOR PEER REVIEW 10 of 21
the present position is sampled by the water pump through the pumping pipe. Then,
the first electromagnetic valve is turn on for collecting the sampling water into the first
sampling
present water sampling bottle. Simultaneously, by thethe position coordinate of theand
present water sampling
present
point water point
measured sampling
by
measured
the pointmodule
GPS measured GPS
is bymodule
also the GPSis module
recorded
also recorded
and sent isto
also
the
sent to
recorded
remote and sent to
human–machine
the remotethehuman–machine
remote interface through
human–machine interface the Bluetooth
through the wireless module.
Bluetooth wireless When
module.the When the
interface
present water sampling throughtask the Bluetooth
is finished, thewireless
water the module.
pump When
andpump
the firsttheelectromagnetic
present water sampling task
present
is waterthe
finished, sampling
water task isand
pump finished,
the first water
electromagnetic andvalve
the first
are electromagnetic
turned off, and the
valve are turned
valve off,turned
are and the off,MF-USV
and theleaves
MF-USV the leaves
presentthe water sampling
present water point to search
sampling point to search
MF-USV
the next contaminated leaves the
water point.presentOncewater
the pH sampling
value point
of the to
watersearch
in the the next contaminated
nextinposition water
the next contaminated water point. Once the pH value of the water the next position
is abnormal,point. Once
the water the the pH
sampling value of the water in the next position is abnormal, the water sampling
is abnormal, water procedure is the same
sampling procedure is as
thethe firstassampling
same the first procedure.
sampling procedure.
procedure
After the water quality is monitoring
the same astask the is
first samplingthe
completed, procedure.
MF-USV After
can the water quality
automatically re- monitoring
After the water quality monitoring task is completed, the MF-USV can automatically re-
task
turn to theturn is completed,
initial location. thecollected
MF-USVwater can automatically return to the initial location. The collected
to the initialThelocation. The collected samples
waterare later
samples analyzed
are later in specialized
analyzed in specialized
water samples are later analyzed in specialized laboratories.
laboratories.
laboratories.

Figure 7. The position of the water quality monitoring system mounted on the MF-USV.
Figure 7. The position of the water quality monitoring system mounted on the MF-USV.
Figure 7. The position of the water quality monitoring system mounted on the MF-USV.

(a) (a) (b) (b)

Figure
Figure 8. (a) CAD 8. (a)
design CAD design
model model of thecollection
water sample collection device. (b) Water sample collection
Figureof8.the
(a)water sample
CAD design model of thedevice. (b) Water
water sample sample
collection collection
device. device
(b) Water mounted
sample on the
collection
device mounted on the MF-USV.
MF-USV. device mounted on the MF-USV.
Sensors 2021, 21, 1102 11 of 21

Sensors 2021, 21, x FOR PEER REVIEW 11 of 21

Sensors 2021, 21, x FOR PEER REVIEW 11 of 21

Figure 9. Circuit diagram of the water quality monitoring system.

Figure 9. Circuit diagram of the water quality monitoring system.
Figure 9. Circuit diagram of the water quality monitoring system.

Figure 10. Water quality monitoring process.

Figure 10.
Figure 10. Water
Waterquality
qualitymonitoring
monitoringprocess.
process.
4. Water Surface Cleaning System
4. Water Surface Cleaning System
The water surface cleaning system shown in Figure 11 consists of a Pixy CMUcam5
visionThe water(as
sensor surface
showncleaning system
in Table shown
3) and in Figure
a water 11 consists
surface cleaningof device.
a Pixy CMUcam5
The Pixy
vision sensor (as shown in Table 3) and a water surface cleaning
CMUcam5 vision sensor is composed of a dual core processor NXP LPC4330 device. The Pixy
at 204MHz,
CMUcam5 vision sensor is composed of a dual core processor NXP LPC4330
an OmniVision image sensor, a 264K bytes RAM, and a 1M bytes Flash, with UART, at 204MHz,
SPI,
an OmniVision image sensor, a 264K bytes RAM, and a 1M bytes Flash, with UART, SPI,
 Sensors 2021, 21, 1102 12 of 21

4. Water Surface Cleaning System

The water surface cleaning system shown in Figure 11 consists of a Pixy CMUcam5
vision sensor (as shown in Table 3) and a water surface cleaning device. The Pixy CMU-
cam5 vision sensor is composed of a dual core processor NXP LPC4330 at 204MHz, an
Sensors 2021, 21, x FOR PEER REVIEW 12 of 21
OmniVision image sensor, a 264K bytes RAM, and a 1M bytes Flash, with 12
Sensors 2021, 21, x FOR PEER REVIEW
UART,
of 21
SPI,
2
I C, and universal serial bus (USB) communication interfaces. The Pixy CMUcam5 vision
sensor is utilized to detect objects by using a hue-based color filtering algorithm, which is
I2C, and universal serial bus (USB) communication interfaces. The Pixy CMUcam5 vision
called the color connected components (CCC) algorithm and implemented in the C/C++
I2C, and
sensor is utilized touniversal serialby
detect objects bus (USB)
using communication
a hue-based interfaces.
color filtering The Pixywhich
algorithm, CMUcam5 is vision
programming language and on the processor NXP LPC4330. Since the hue-based color
called the sensor is utilized
color connected to detect
components objects
(CCC) by algorithm
using a hue-based color filtering
and implemented algorithm,
in the C/C++ which is
filtering
programming called methods
the color
language and areonfast,
connected the efficient,
components
processor and
NXP relatively
(CCC) robust,
algorithm
LPC4330. Sinceand they
the are popular.
implemented
hue-based TheC/C++
in the
color hue-based
color filtering
programming
filtering methods algorithm
language
are fast, calculates
and on
efficient, the
andtherelativelyhue (color)
processorrobust, and
NXP LPC4330. saturation
they areSince of each
the hue-based
popular. The red, green,
color and
blue (RGB)
filtering pixel
methods derived
are from
fast, the
efficient,image
and sensor
relativelyand
hue-based color filtering algorithm calculates the hue (color) and saturation of each red, uses
robust, these
they as
arethe primary
popular. filtering
The
parameters.
hue-based The
color hue of
filtering the object
algorithm remains
calculates basically
the hue
green, and blue (RGB) pixel derived from the image sensor and uses these as the primary unchanged
(color) and under
saturation changes
of each in lighting
red,
filteringand green,
parameters.andThe
exposure blue
[27].(RGB)
hue Inthe
of pixel
this derived
paper,
object afrom
remains the image
redbasically
color objectsensor and
is selected
unchanged useschanges
under these
to asin
representthe primary
the floating
filtering
lightinggarbage
and toparameters.
exposure [27]. Inthe
perform The
this hue
paper,
water of the
a red
surface object
colorremains
cleaningobject basically
taskis selected
because unchanged
toofrepresentunder
its distinct changesfor
thecolour in it can
floatingbe lighting
garbage to and exposure
perform the [27].
water In this
surface paper,
cleaning a red
task color
becauseobject
easily distinguished from its surroundings. Hence, when the hue-based color filteringof is
its selected
distinct to represent
colour the
for it can floating
be garbage
easily distinguished
algorithm is utilized to perform
tofrom
detect thesurroundings.
its water
the surfaceHence,
floating cleaning
garbage, task
whenthethebecause of can
hue-based
MF-USV itscolor
distinct
detect colour
it correctly.
filteringThe for it can is
algorithm beutilized
easily distinguished
to detect the from its surroundings.
floating garbage, the Hence,
MF-USV when
can the hue-based
detect it color
water surface cleaning device is designed for cleaning and collecting the floating
filtering
correctly. The wateralgorithm is utilized
surface cleaning to detect
device the floating
is designed garbage,
for cleaning and thecollecting
MF-USVthe can detect it
garbage, which is composed of two DC gear motors (HN35GBE-1640Y) and a salvage net.
correctly.
floating garbage, The is
which water surfaceofcleaning
composed two DCdevice is designed
gear motors for cleaning and
(HN35GBE-1640Y) andcollecting
a the
Thefloating
DC gear motors
garbage,
are driven
whichare is driven
by a motor
composed
drive module (L298N) through the Arduino
salvage net. The DC gear motors by aofmotor
two DC drive gear motors
module (HN35GBE-1640Y)
(L298N) through and a
MEGA
the Arduino salvage
MEGA2560net.microcontroller,
Themicrocontroller,
2560 DC gear motors which areare
which areutilized
driven by a to
utilized tocontrol
motor drive
control themodule
the salvage
salvage nettoto rise
(L298N)
net and drop
through
rise andvia thevia
the
drop bobbins
Arduino for collecting
MEGA
the bobbins for2560 thethefloating
microcontroller,
collecting floating garbage.
which Figure
Figure 12
are utilized
garbage. to shows
12control
shows the thesalvage
circuit net
cir- diagram
to of
therise
cuit diagram water thesurface
ofand drop
watervia cleaning
the bobbins
surface system
cleaning mounted
forsystem
collecting theon
mounted the
floating
on theMF-USV.
garbage.
MF-USV.Figure 12 shows the cir-
cuit diagram of the water surface cleaning system mounted on the MF-USV.

Figure 11. The position of the water surface cleaning system mounted on the MF-USV.
Figure 11. The position of the water surface cleaning system mounted on the MF-USV.
Figure 11. The position of the water surface cleaning system mounted on the MF-USV.

Figure 12. Circuit diagram of the water surface cleaning system and locomotion module.
Figure 12. Circuit diagram of the water surface cleaning system and locomotion module.
Figure 12. Circuit diagram of the water surface cleaning system and locomotion module.
Sensors 2021,
Sensors 21, x1102
2021, 21, FOR PEER REVIEW 13of
13 of21
21

For
For thethewater
watersurface
surfacecleaning
cleaningtask, task, wewe utilize
utilizethethe
Pixy
PixyCMUcam5
CMUcam5 vision sensor
vision sensor to
detect the floating garbage (red color objects). When the
to detect the floating garbage (red color objects). When the vision sensor mounted on vision sensor mounted on the
front of the
the front of MF-USV
the MF-USV detects floating
detects garbage,
floating garbage,the MF-USV
the MF-USV can automatically
can automatically tracktrack
and
collect
and collectthe floating
the floating garbage
garbage through
through thethewater
watersurface
surfacecleaning
cleaningalgorithm.
algorithm.The The Pixy
Pixy
CMUcam5
CMUcam5 vision vision sensor
sensor defines that the X X and
and YYcoordinates
coordinatesof ofthe
theimage
imageranged
rangedfrom from0
0toto319
319andandfromfrom00toto199, 199,asasshown
shownininFigureFigure13. 13.TheThewater
watersurface
surface cleaning
cleaning algorithm
algorithm is is
shown
shownin in Figure
Figure 14. 14. First,
First, with
with the the water
water surface
surface cleaning
cleaning algorithm,
algorithm, the the floating
floating garbage
garbage
isisdetected
detectedby bythe thePixy
Pixy CMUcam5
CMUcam5 vision
vision sensor
sensor mounted
mounted on the
on front of the
the front ofMF-USV,
the MF-USV, and
its
and direction is determined
its direction is determined basedbased on theon x coordinate
the x coordinateof the ofimage extracted
the image by the by
extracted vision
the
sensor. Subsequently,
vision sensor. there arethere
Subsequently, five situations for the floating
are five situations garbage
for the floating cleaning
garbage and collec-
cleaning
tion: (1) turn left(1)
and collection: about
turn 30°: about 030◦ : when
left when x 0≤
63, ≤ 63, the algorithm
thex algorithm determinesdetermines
that there that is a
there is garbage
floating a floatingongarbage
the left onside theofleft
theside of the
vehicle, and vehicle,
then the and then thedrives
MF-USV MF-USV drivesmo-
the servo the
servo
tor motor (MG996R)
(MG996R) in the locomotion
in the locomotion module tomodule controltothe control
rudderthetorudder to turn
turn left about left30°.
about(2)
30◦ . (2)
Turn left Turn
aboutleft 15°:about
when1564 ◦ : when 64 ≤ x ≤ 127, the algorithm determines that there is
x 127, the algorithm determines that there is a floating
a floating
garbage ongarbage on the
the left front leftoffront
side side of the
the vehicle, andvehicle,
then theand then the
MF-USV MF-USV
drives drives
the servo motorthe
servo motor
(MG996R) in (MG996R)
the locomotion in themodule
locomotion module
to control thetorudder
controltothe rudder
turn to turn15°.
left about left(3)
about
Go
15◦ . (3) Go
straight straight
ahead: when ahead:
128 when x 128 190,≤thex ≤algorithm
190, the algorithm
determines determines
that therethat is athere
floatingis a
floating garbage on the front side of the vehicle, and then
garbage on the front side of the vehicle, and then the MF-USV goes straight ahead. (4) the MF-USV goes straight ahead.
(4) Turn
Turn rightright
about about 15◦ : when
15°: when 191 191 x ≤ x253, ≤ 253, the algorithm
the algorithm determines
determines that is
that there there is a
a float-
floating garbage on the right front side of the vehicle, and
ing garbage on the right front side of the vehicle, and then the MF-USV drives the servo then the MF-USV drives the
servo motor
motor (MG996R) (MG996R) in the locomotion
in the locomotion modulemodule
to control to control the rudder
the rudder to turnto turnabout
right right15°.
about (5)
15 ◦ . (5) Turn right about 30◦ : when 254 ≤ x ≤ 319, the algorithm determines that there is
Turn right about 30°: when 254 x 319, the algorithm determines that there is floating
floating on
garbage garbage
the righton the sideright
of thesidevehicle,
of the vehicle,
and then andthethen the MF-USV
MF-USV drives the drives themotor
servo servo
motor (MG996R) in the locomotion module to control the
(MG996R) in the locomotion module to control the rudder to turn right about 30°. In this rudder to turn right about 30◦ .
In this
way, theway, the floating
floating garbagegarbage
will be will on thebe front
on theoffront of the vehicle
the vehicle until thatuntil that is collected
is collected by the
water surface cleaning device. Finally, once the floating garbage is collectedcollected
by the water surface cleaning device. Finally, once the floating garbage is in the
in the salvage
salvage net, the water surface cleaning
net, the water surface cleaning task is completed. task is completed.

Figure
Figure13.
13.XXand
andYYcoordinates
coordinatesof
ofthe
theimage
imageextracted
extracted by
by the
the vision
vision sensor.
sensor.
Sensors 2021, 21, 1102 14 of 21
Sensors 2021, 21, x FOR PEER REVIEW 14 of 21

Figure 14.
Figure 14. Water
Watersurface
surfacecleaning
cleaningalgorithm.
algorithm.

5. Experimental Results
5. Experimental
The MF-USV Results
was designed to avoid the obstacles, monitor the water quality, and
cleanThe
the MF-USV was designed
floating garbage. to avoid thewere
Field experiments obstacles,
conductedmonitor the water
on 15–18 Januaryquality,
2019 inand
the
clean
shallowtheChia-Ming
floating garbage. Field experiments
Lake, Taichung, Taiwan, were conducted
to evaluate on 15–18 January
the performance of the2019 in the
MF-USV.
shallow
The testsChia-Ming Lake,into
were divided Taichung, Taiwan,
four parts: to evaluate
obstacle the performance
avoidance, of the
water quality MF-USV.
monitoring,
The tests were divided into four parts: obstacle avoidance, water quality
water surface cleaning, and autonomous navigation. These tasks could be performed monitoring, water
surface cleaning, and autonomous navigation. These tasks could be performed
separately or simultaneously. Figure 15 shows the boat body of the MF-USV which is separately
or simultaneously.
composed Figure
of the main 15 shows
control unit, the boat body
locomotion of the MF-USV
module, which
positioning is composed
module (GPS mod- of
the
ule),main control
obstacle unit, locomotion
avoidance module, positioning
module (ultrasonic module
sensors), water (GPSmonitoring
quality module), obstacle
system
avoidance
(pH sensormodule (ultrasonic
and water samplesensors),
collectionwater quality
device), monitoring
water system (pH
surface cleaning sensor
system and
(vision
water sample collection device), water surface cleaning system (vision
sensor and water surface cleaning device), communication module (Bluetooth wirelesssensor and water
surface
module), cleaning device), and
power module, communication module (Bluetooth
remote human–machine wireless module), power
interface.
module, and remote human–machine interface.
Sensors 2021,21,
Sensors2021, 21,1102
x FOR PEER REVIEW 1515ofof2121

(a)

(b)
Figure15.
Figure 15.View
Viewofofthe
theMF-USV.
MF-USV.(a)
(a)Side
Sideview.
view.(b)
(b)Rear
Rearview.
view.

5.1.
5.1.Obstacle
ObstacleAvoidance
Avoidance
The
Thetest
testwas
wasconducted
conductedtotocheckcheckthethecapability
capabilityofofthe theobstacle
obstacleavoidance
avoidanceofofthe the
MF-USV. The obstacle avoidance module of the MF-USV composed
MF-USV. The obstacle avoidance module of the MF-USV composed of two HC-SR04 ul- of two HC-SR04
ultrasonic sensorsisis designed
trasonic sensors designed forfor detecting
detecting andand avoiding
avoiding obstacles.
obstacles. Once
Oncethethedistances
distances
measured by the ultrasonic sensors are greater or less than the
measured by the ultrasonic sensors are greater or less than the threshold values,threshold values, the MF-the
USV executes a threshold-based obstacle avoidance algorithm to avoid
MF-USV executes a threshold-based obstacle avoidance algorithm to avoid obstacle col- obstacle collision.
When L > 65
lision.(DWhen (𝐷cm) and65 cm) > 65(𝐷cm), the
(DR and 65 threshold-based obstacle avoidance
cm), the threshold-based obstacle algorithm
avoidance
determines
algorithm determines that there is no any obstacle on the front of the vehicle,MF-USV
that there is no any obstacle on the front of the vehicle, and then the and then
goes straight ahead,
the MF-USV as shown
goes straight in Figure
ahead, 16a. Subsequently,
as shown in Figure 16a. when (DL > 65 cm)
Subsequently, and (10
when (𝐷 >cm 65
DR ≤
<cm) 65 cm), the threshold-based obstacle avoidance algorithm determines
and (10 cm <𝐷 ≤ 65 cm), the threshold-based obstacle avoidance algorithm deter- that there is
an obstacle
mines thaton the right
there is an front sideon
obstacle of the
the vehicle, as shown
right front side ofinthe
Figure 16b. as
vehicle, Then, the MF-USV
shown in Figure
drives the servo motor (MG996R) in the locomotion module to
16b. Then, the MF-USV drives the servo motor (MG996R) in the locomotion module control the rudder to turnto
left, as shown in Figure 16c,d. Once the MF-USV has avoided the obstacle
control the rudder to turn left, as shown in Figure 16c,d. Once the MF-USV has avoided successfully, the
distances are measured
the obstacle successfully, as (D > 65 cm) are
theL distances (DR > 65as
andmeasured cm),
(𝐷 the 65threshold-based
cm) and (𝐷 obstacle65 cm),
the threshold-based obstacle avoidance algorithm determines that there is no any obsta-
Sensors 2021, 21, 1102 16 of 21
Sensors 2021, 21, x FOR PEER REVIEW 16 of 21

avoidance algorithm determines that there is no any obstacle in front of the vehicle, and
cle in front of the vehicle, and then the MF-USV goes straight ahead, as shown in Figure
then the MF-USV goes straight ahead, as shown in Figure 16e,f.
16e,f.

Figure 16. Experimental

Figure 16. Experimental results
results for
forthe
theobstacle
obstacleavoidance
avoidancetask.
task.(a)
(a)When
When(D (𝐷L > 6565 cm)
cm)and
and(D(𝐷R > 65 65 cm),
cm), the
the MF-USV
MF-USV
goes straight ahead. (b) When (𝐷 (D L> >6565cm)
cm)and
and(10
(10cm
cm<𝐷<D≤ R65≤cm),
65 cm), there
there is anisobstacle
an obstacle on the
on the rightright front
front sideside of the
of the ve-
hicle. (c)(c)
vehicle. TheTheMF-USV
MF-USVturns
turnsleft.
left.(d)
(d)The
TheMF-USV
MF-USV keeps
keeps turning When (D
turning left. (e) When (𝐷L > 65
65 cm) and
and (D(𝐷R > 6565 cm), the
MF-USV goes straight ahead. (f) The MF-USV keeps going straight ahead.

5.2. Water Quality Monitoring

5.2. Water Quality Monitoring
The test was conducted to check the capability of water quality monitoring and
waterThe test was
sample conducted
collection to check
by the the capability
MF-USV. The water of water
qualityquality monitoring
monitoring and of
system water
the
sample collection by the MF-USV. The water quality monitoring
MF-USV composed of the pH sensor, GPS module, and water sample collection device is system of the MF-USV
composedfor
designed of monitoring
the pH sensor, waterGPSquality
module,andand water sample
collecting collectionThe
water samples. devicepH is designed
sensor and
for monitoring water quality and collecting water samples. The
GPS module mounted on the MF-USV were used to detect the pH values of the water pH sensor and GPS module
and
mounted on the MF-USV were used to detect the pH values of the water and position
position coordinates during sailing in real time, respectively. When the water quality was
coordinates during sailing in real time, respectively. When the water quality was normal
normal (6.5 pH 8.5), the MF-USV can automatically or manually sail on the lake
(6.5 ≤ pH ≤ 8.5), the MF-USV can automatically or manually sail on the lake and the sailing
and the sailing position coordinates measured by the GPS module was recorded and
position coordinates measured by the GPS module was recorded and sent to the remote
sent to the remote human–machine interface through the Bluetooth wireless module, as
human–machine interface through the Bluetooth wireless module, as shown in Figure 17.
shown in Figure 17. When the water pH value at the first location (as shown in Figure 17b)
When the water pH value at the first location (as shown in Figure 17b) was higher than 8.5
was higher than 8.5 or lower than 6.5, which represented that the present water field was
or lower than 6.5, which represented that the present water field was contaminated and the
contaminated and the water pump was driven to extract the water for cleaning residual
water pump was driven to extract the water for cleaning residual liquid in the pumping
liquid in the pumping pipe. Then, the water on the present position was sampled by the
pipe. Then, the water on the present position was sampled by the water pump through the
water pump through the pumping pipe and the first electromagnetic valve was turned on
pumping pipe and the first electromagnetic valve was turned on to collect the sampling
to collect the sampling water in the first sampling bottle. The water level of the first sam-
water in the first sampling bottle. The water level of the first sampling bottle was increased
pling bottle was
from Figure 18a increased
to Figure from Figure
18b. In 18a tothe
addition, Figure 18b. In
position addition,of
coordinate thetheposition
presentcoordi-
water
nate of thepoint
sampling present water by
measured sampling
the GPSpoint
modulemeasured by the GPS recorded
was simultaneously module was and simultane-
sent to the
ously
remoterecorded and sentinterface.
human–machine to the remote human–machine
Finally, the water pump and interface. Finally, the water
first electromagnetic valve
pump and first electromagnetic valve were turned off to
were turned off to finish the first water sampling task. The MF-USV leaved finish the first waterthesampling
present
task.
waterThe MF-USV
sampling leaved
point andthe present
went ahead water sampling
to search pointcontaminated
the next and went ahead to search
water point,the
as
next contaminated water point, as shown in Figure 17c,d. Once the
shown in Figure 17c,d. Once the contaminated water field at the second location (as shown contaminated water
field at the17e)
in Figure second location (as
was detected by shown
the pH in Figure
sensor, 17e) was
whose detected
pH value wasby the pH
higher thansensor,
8.5 orwhose
lower
pH
than 6.5, the residual liquid in the pumping pipe was cleaned by the water extracted bypipe
value was higher than 8.5 or lower than 6.5, the residual liquid in the pumping the
was
watercleaned
pump. by the water extracted
Subsequently, the waterby onthe
the water
secondpump.
water Subsequently,
sampling position the was
water on the
sampled
second waterpump
by the water sampling position
through was sampled
the pumping pipe byandthethewater
second pump through thevalve
electromagnetic pumping
was
pipe and the second electromagnetic valve was turned on to collect
turned on to collect the sampling water in the second sampling bottle. The water the sampling water in
level
the second
of the secondsampling
sampling bottle.
bottleThe
was water level of
increased the Figure
from second18c sampling
to Figure bottle
18d.was increased
Additionally,
from Figure coordinate
the position 18c to Figure of 18d. Additionally,
the second the position
water sampling pointcoordinate
measuredof bythethesecond water
GPS module
sampling point measured
was simultaneously recordedby the
andGPS
sentmodule was simultaneously
to the remote human–machine recorded and Finally,
interface. sent to
the remote human–machine interface. Finally, the water pump and second electromag-
Sensors 2021, 21, 1102 17 of 21

Sensors 2021, 21, x FOR PEER REVIEW 17 of 21

Sensors 2021, 21, x FOR PEER REVIEW 17 of 21

the water pump and second electromagnetic valve were turned off to finish the second
water sampling
netic valve were task.
turnedAfter
off tothe water
finish thequality
secondmonitoring task was
water sampling task.completed, the water
After the water qual-
netic
samples valve
were were turned
conveyed to off
the to finish
initial the second
location and water
later sampling
analyzed in task. After
specialized
ity monitoring task was completed, the water samples were conveyed to the initial the water qual-
laboratories,
loca-
as ity monitoring
shown in task
Figure was
17f. completed, the water samples were conveyed
tion and later analyzed in specialized laboratories, as shown in Figure 17f. to the initial loca-
tion and later analyzed in specialized laboratories, as shown in Figure 17f.

Figure 17.
Figure Sailing movement
17. Sailing movement trajectory
trajectory of
of the
theMF-USV.
MF-USV.(a) (a)The
TheMF-USV
MF-USVwas waslocated
locatedatatthe
thestart
startpoint.
point.(b)
(b)When
Whenthe
thewater
wa-
ter Figure
pH value
pH 17.
value Sailing
at the
at the movement
firstfirst
location was
location trajectory
higher
was higher ofthan
than the MF-USV.
8.58.5
ororlower
lower(a) The
than
than MF-USV
6.5, the
6.5, waswas
thewater
water located
was at theby
sampled
sampled start
by thepoint.
the water(b)
water When
pump.
pump. (c)the
Thewa-
ter pH leaved
MF-USV
MF-USV value at
leaved thepresent
the first location
present water was higherpoint.
sampling
sampling than (d)
point. 8.5
(d) or lower
The
The than 6.5,
MF-USV
MF-USV went
wenttheahead
watertowas
ahead sampled
search by the
next
the next water pump.water
contaminated
contaminated (c) The
MF-USV
point. (e) leaved the present water sampling point. (d) The MF-USV went ahead to search the next contaminated water
point. (e) The
The contaminated
contaminated water water field
field at
at the
the second
second location
location waswas detected
detected byby the
the pH
pH sensor
sensor and sampled by
and sampled the water
by the water
point.
pump. (f) (e) The contaminated water field at the second location was detected by the pH sensor and sampled by the water
pump. (f) The
The water
watersamples
sampleswerewereconveyed
conveyedtotothe theinitial
initiallocation.
location.
pump. (f) The water samples were conveyed to the initial location.

(a) (b) (c) (d)

(a) (b) (c) (d)
Figure 18. Experimental results for the water quality monitoring task. (a) The initial water level of the first sampling bot-
Figure
Figure
tle. 18.18.
(b) The Experimental
Experimental
water results
level of results
the forthe
firstfor thewater
sampling water quality
quality
bottle monitoring
aftermonitoring
the task.
task.
first water (a) The
(a) The initial
sampling. (c) Thewater
initial initiallevel
water levelof
ofthe
water thefirst
level of sampling
first secondbot-
sampling
the
tle.
sampling (b) The
bottle. (b)bottle.water level
(d) The
The water of
water
level the first
level
of the sampling
firstofsampling
the second bottle after
sampling
bottle the first
bottle
after the water
firstafter
water sampling.
thesampling.
second water (c) The initial
sampling.
(c) The water level of the second
initial water level of the second
sampling bottle. (d) The water level of the second sampling bottle after the second
sampling bottle. (d) The water level of the second sampling bottle after the second water sampling. water sampling.
5.3. Water Surface Cleaning
5.3. Water Surface Cleaning
The test
5.3. Water was Cleaning
Surface conducted to check the capability of the floating garbage detection and
The test was
water surface cleaning conducted to checkThe
of the MF-USV. thewater
capability of the
surface floating
cleaning systemgarbage
of thedetection
MF-USV and
The test
water surface was conducted to check the capability of the floating garbage detection and
composed of thecleaning of theand
vision sensor MF-USV. The water
water surface surface
cleaning cleaning
device system for
is designed of the MF-USV
detecting
water surfaceof
composed cleaning
the of the
vision MF-USV.
sensor and The water
water surface surface cleaning
cleaning device system
is of thefor
designed MF-USV
detecting
and cleaning the floating garbage. The vision sensor and its hue-based color filtering al-
composed of thethe
and cleaning vision sensor and water
floating surface cleaning device is designed for detecting
gorithm were used to detect garbage.
the floating Thegarbage
vision sensor
on waterand its hue-based
surface in real time.color filtering
When theal-
and cleaning theused
floating garbage. The vision sensor and its hue-based color filtering
floating garbage was detected, the MF-USV can automatically track and collect the floatingthe
gorithm were to detect the floating garbage on water surface in real time. When
algorithm
floating were used to detect thethe
floating garbage on water surface in and
real collect
time. When the
garbage bygarbage
using the was detected,
water surface MF-USV
cleaning can automatically
algorithm. When the track
vision the floating
sensor mounted
floating garbage was detected, the MF-USV can automatically track and collect the floating
ongarbage
the frontbyofusing the water
the MF-USV surfacethe
detected cleaning algorithm.
first floating Whenwhich
garbage, the vision
is onsensor
the leftmounted
front
garbage
on the by using
front of the water
the MF-USV surface cleaning
detected the algorithm.
first floating When the vision
garbage, which sensor
is on mounted
the left front
side of the vehicle, the MF-USV could drive the servo motor (MG996R) in the locomotion
onside
the of
front of the MF-USV detected the first floating garbage, which is on in the
the left front
module tothe vehicle,
control the the MF-USV
rudder could
to turn left drive
aboutthe 15° servo motor (MG996R)
for tracking the first floating locomotion
garbage,
side of the vehicle, the MF-USV could drive the servo motor (MG996R) in the locomotion
as module
shown intoFigure
control19a.theSubsequently,
rudder to turnthe leftwater
aboutsurface
15° forcleaning
trackingdevice
the first floating
turned on thegarbage,
DC
gear motors to control the salvage net for collecting the first floating garbage, as shown inDC
as shown in Figure 19a. Subsequently, the water surface cleaning device turned on the
gear motors to control the salvage net for collecting the first floating garbage, as shown in
Sensors 2021, 21, 1102 18 of 21

Sensors 2021, 21, x FOR PEER REVIEW

module to control the rudder to turn left about 15◦ for tracking the first floating garbage, 18 of 21
as shown in Figure 19a. Subsequently, the water surface cleaning device turned on the DC
gear motors to control the salvage net for collecting the first floating garbage, as shown
in Figure
Figure 19b.19b.
Then,Then, the vision
the vision sensorsensor
mountedmounted
on theon the MF-USV
MF-USV detecteddetected the floating
the second second
floating garbage on the left front side of the vehicle, the MF-USV could
garbage on the left front side of the vehicle, the MF-USV could drive the servo motor drive the servo
motor (MG996R)
(MG996R) in the locomotion
in the locomotion module module
to controltothe
control
rudder thetorudder to about
turn left turn left
15° about 15◦
for track-
for tracking
ing the second thefloating
second garbage,
floating garbage,
as shown asin
shown
Figurein19c.
Figure 19c. Subsequently,
Subsequently, the waterthesurface
water
surface cleaning device was used to collect the second floating garbage, as
cleaning device was used to collect the second floating garbage, as shown in Figure 19d. shown in Figure
19d.
Then,Then, the vision
the vision sensorsensor mounted
mounted on theon MF-USV
the MF-USV detected
detected the the third
third floating
floating garbage
garbage on
on the left side of the vehicle, the MF-USV could drive the servo motor
the left side of the vehicle, the MF-USV could drive the servo motor (MG996R) in the lo- (MG996R) in
the locomotion module to control the rudder to turn left about 30 ◦ for tracking the third
comotion module to control the rudder to turn left about 30° for tracking the third floating
floating garbage,
garbage, as shown asin
shown
Figurein19e.
Figure 19e. Finally,
Finally, the thirdthe third floating
floating garbagegarbage was collected
was collected by the
by the water surface cleaning device, as shown
water surface cleaning device, as shown in Figure 19f. in Figure 19f.

Figure 19. Experimental

Figure 19. Experimental results
results for
for the
the water
water surface
surface cleaning
cleaning task.
task. (a)
(a) When
When the
the first
first floating
floating garbage
garbage isis on
on the
the left
left front
front
side of the vehicle, the MF-USV turns left about 15° 15◦ for tracking
tracking the first floating garbage. (b) The first floating garbage is
collected by the water surface cleaning device. (c) (c) When
When thethe second
second floating
floating garbage
garbage is is on the left front side of the vehicle,
the MF-USV
the MF-USV turns
turns left
left about 15◦ for
about15° fortracking
trackingthe
thesecond
secondfloating
floatinggarbage.
garbage.(d)(d)
The
The second
second floating garbage
floating garbage is collected
is collected by
the water surface cleaning device. (e) When the third floating garbage is on the left side of the vehicle,
by the water surface cleaning device. (e) When the third floating garbage is on the left side of the vehicle, the MF-USV the MF-USV turns
left about
turns 30° for30tracking
left about the third
◦ for tracking the floating garbage.
third floating (f) The(f)third
garbage. The floating garbage
third floating is collected
garbage by theby
is collected water surfacesurface
the water clean-
ing device.
cleaning device.

Table
Table 4 4 shows
shows the
the success
success rates
rates ofof floating
floating garbage
garbage detection
detection by by the
the Pixy
Pixy CMUcam5
CMUcam5
vision sensor with its hue-based color filtering algorithm. When
vision sensor with its hue-based color filtering algorithm. When the floating garbage the floating garbage lo-
cated in the visual angle ranged from –30° ◦to 30° ◦ on the front
located in the visual angle ranged from –30 to 30 on the front of the MF-USV and the of the MF-USV and the
distances
distances between
between the
the floating
floating garbage
garbage and and thethe MF-USV
MF-USV were were 40 and 70
40 and 70 cm,
cm, the the success
success
rates
rates of
of floating
floating garbage detection were
garbage detection were all all 100%.
100%. When
When the the floating
floating garbage
garbage was was located
located
in
in front
front of of the
the MF-USV (visual angle
MF-USV (visual angle = 0◦ ), we
= 0°), we obtained
obtained better
better success
success rates
rates regardless
regardless thethe
distance between the
distance between the floating
floatinggarbage
garbageand andthe theMF-USV.
MF-USV. OnOnthethe other
other hand,hand, worseworse suc-
success
cess
ratesrates
were were obtained
obtained when when the floating
the floating garbage garbage was located
was located in avisual
in a larger largerangle visualandangle
at a
and at a longer distance from the front of the MF-USV. Table 5 shows
longer distance from the front of the MF-USV. Table 5 shows the success rates of floating the success rates of
floating garbage collection
garbage collection when the when distance thebetween
distance thebetween
floating the floating
garbage and garbage
the MF-USV and was
the
MF-USV
130 cm. When was 130thecm. When
floating the floating
garbage garbage
was located onwasthe located
left sideon ◦ ~30
(15the ◦ ), side
left (15°~30°),
left front side
(0◦ ~15
left ◦ ), side
front side (0◦ ),front
front(0°~15°), rightside (0°),
front (0◦ front
right
side ~15◦ ),side
and(0°~15°),
the rightand sidethe ◦ ~30side
(15right ◦ ), the
(15°~30°),
success
the
ratessuccess rates of garbage
of the floating the floating garbage
collection werecollection
70%, 92%, were
95%,70%,
95%, 92%,
and95%,75%,95%, and 75%,
respectively.
respectively.
Sensors 2021, 21,
Sensors 2021, 21, 1102
x FOR PEER REVIEW 19 of
19 of 21
21

Table
Table 4. The
4. The success
success rates
rates of floating
of floating garbage
garbage detection
detection by the
by the PixyPixy CMUcam5
CMUcam5 vision
vision sensor.
sensor.

Angle Angle
−30° −15° −30◦ 0° −15◦ 0◦15° 15◦ 30° 30◦
Distance Distance
40 cm 100% 40 cm 100% 100% 100% 100% 100%
100% 100% 100%
100%
70 cm 100% 70 cm 100% 100% 100% 100% 100%
100% 100% 100%
100%
100 cm 94% 100 cm 97% 94% 99% 97% 99%
99% 99% 96%96%
130 cm 89% 94% 98% 97% 93%
130 cm 89% 94% 98% 97% 93%

Table
Table 5. The
5. The success
success ratesrates of floating
of floating garbage
garbage collection
collection forwater
for the the water surface
surface cleaning
cleaning task.task.

Angle Left Left

Angle Left Front Left Right
Front Right Right
Right
Distance 15°~30° Distance 0°~15° 15◦ ~30◦ 0° 0◦ ~15◦ ◦
00°~15° 0◦ ~15◦ 15◦ ~30◦
15°~30°
130 cm 70% 130 cm 92% 70% 95% 92% 95%95% 95% 75%75%

5.4. Multi-Function Tests

5.4. Multi-Function Tests
The testwas
The test wasconducted
conducted to to check
check the the capability
capability of theofobstacle
the obstacle avoidance,
avoidance, water
water quality
quality
monitoring, and water surface cleaning of the MF-USV simultaneously. At first, thefirst,
monitoring, and water surface cleaning of the MF-USV simultaneously. At MF-
the
USVMF-USV
sailed onsailed on the
the lake lake automatically
automatically and executed
and executed the obstacle
the obstacle avoidance
avoidance task bytask by
using
using the threshold-based
the threshold-based obstacle obstacle
avoidanceavoidance algorithm,
algorithm, as shownas shown in Figure
in Figure 20a. Subse-
20a. Subsequently,
quently, when the quality of water was abnormal, the MF-USV
when the quality of water was abnormal, the MF-USV could collect the water sample could collect the waterin
sample in the water bottle and record the position coordinate at the present
the water bottle and record the position coordinate at the present water sampling point, water sam-
as
pling
shownpoint, as shown
in Figure in Figure
20b. Then, when 20b.
theThen, when
floating the floating
garbage garbagethe
was detected, was detected,
MF-USV the
could
MF-USV could collect and clean the floating garbage by using the water
collect and clean the floating garbage by using the water surface cleaning algorithm, as surface cleaning
algorithm, as shown
shown in Figure in Figure
20c. Finally, 20c.the
after Finally, after themonitoring
water quality water quality
and monitoring
water surface and water
cleaning
surface cleaning
tasks were tasks were
completed, completed,
the MF-USV couldtheexecute
MF-USV thecould execute
obstacle the obstacle
avoidance avoidance
task and return
task and
to the return
initial to the automatically,
location initial location as automatically, as shown
shown in Figure 20d,e.in Figure 20d,e.

Figure 20. Experimental results for the multi-function tasks of the MF-USV. (a) The MF-USV sailed on the lake automatically
Figure 20. Experimental results for the multi-function tasks of the MF-USV. (a) The MF-USV sailed on the lake automati-
and
callyexecuted the obstacle
and executed avoidance
the obstacle task.task.
avoidance (b) (b)
When thethe
When quality of of
quality water was
water was abnormal,
abnormal,the
theMF-USV
MF-USVcould
couldcollect
collect the
the
water
water sample in the water bottle. (c) When the floating garbage was detected, the MF-USV could collect and clean
sample in the water bottle. (c) When the floating garbage was detected, the MF-USV could collect and clean the
the
floating garbage. (d) The MF-USV executed the obstacle avoidance
avoidance task. (e) The MF-USV returned to the the initial
initial location
location
automatically.
Sensors 2021, 21, 1102 20 of 21

6. Conclusions
In this paper, a MF-USV has been creatively developed to avoid obstacles, collect con-
taminated water samples, and clean garbage floating on the water surface. The architecture
of the proposed MF-USV comprises the main control unit, locomotion module, positioning
module, obstacle avoidance module, water quality monitoring system, water surface clean-
ing system, communication module, power module, and remote human–machine interface
for the purposes of obstacle avoidance, water quality monitoring, and water surface clean-
ing. Our experimental results have successfully validated that the MF-USV can successfully
execute obstacle avoidance, water quality monitoring, and water surface cleaning tasks.
Once the contaminated water field is detected, the MF-USV can collect the water sample for
further analysis and send the pH value and GPS coordinates of the water sampling point
to the remote human–machine interface for recording and displaying. On the other hand,
when the floating garbage is detected, the MF-USV can automatically track and collect
the floating garbage through the water surface cleaning algorithm. The success rates of
floating garbage detection are all 100% when the floating garbage located in the visual angle
ranged from −30◦ to 30◦ on the front of the MF-USV and the distances between the floating
garbage and the MF-USV were 40 and 70 cm. When the distance between the floating
garbage and the MF-USV was 130 cm, the success rates of the floating garbage collection
were from 70% to 95% in different directions. Based on the above experimental results,
we believe that the proposed MF-USV and its associated algorithms will provide a novel
and effective contribution to unmanned surface vehicle design. The main contribution
of this paper is to develop a low cost USV with multi-functions and its threshold-based
algorithms for providing an effective tool for obstacle avoidance, water quality monitoring,
and water surface cleaning tasks.
In the future, the MF-USV will be tested in more complicated field environments
to develop more robust and stable methodology for water quality monitoring and water
surface cleaning. Moreover, extra sensors and algorithms will be added to the MF-USV
to improve the functions of obstacle avoidance, water quality monitoring, and water
surface cleaning.

Author Contributions: Conceptualization, Y.-L.H. and H.-C.C.; methodology, Y.-L.H., H.-C.C., S.-S.H.
and G.-R.O.; software, G.-R.O., J.-R.W. and C.H.; validation, Y.-L.H., S.-S.H., G.-R.O., J.-R.W. and C.H.;
formal analysis, Y.-L.H. and H.-C.C., G.-R.O.; investigation, Y.-L.H. and H.-C.C.; resources, Y.-L.H.,
H.-C.C. and S.-S.H.; writing—original draft preparation, Y.-L.H. and H.-C.C.; writing—review and
editing, Y.-L.H. and H.-C.C.; visualization, G.-R.O.; supervision, Y.-L.H., H.-C.C. and S.-S.H.; project
administration, Y.-L.H. and H.-C.C.; funding acquisition, Y.-L.H. and H.-C.C. All authors have read
and agreed to the published version of the manuscript.
Funding: This work is supported by the Ministry of Science and Technology of the Republic of
China, Taiwan, under Grant No. MOST 108-2628-E-035-002-MY3, MOST 109-2221-E-035-023, MOST
107-2221-E-035-080, and MOST 107-2813-C-035-135-E.
Institutional Review Board Statement: Not Applicable.
Informed Consent Statement: Not Applicable.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author. The data are not publicly available due to further study will be carried out
using the same data.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Jambeck, J.R.; Geyer, R.; Wilcox, C.; Siegle, T.R.; Perryman, M.; Andrady, A.; Narayan, R.; Law, K.L. Plastic waste inputs from
land into the ocean. Science 2015, 347, 768–771. [CrossRef] [PubMed]
2. Lebreton, L.C.M.; Van der Zwet, J.; Damsteeg, J.W.; Slat, B.; Andrady, A.; Reisser, J. River plastic emissions to the world’s oceans.
Nat. Commun. 2017, 8, 1–10. [CrossRef] [PubMed]
Sensors 2021, 21, 1102 21 of 21

3. Specht, M.; Specht, C.; Mindykowski, J.; Dabrowski, P.; Masnicki, R.; Marker, A. Geospatial modeling of the tombolo phenomenon
in sopot using integrated geodetic and hydrographic measurement methods. Remote Sens. 2020, 12, 737. [CrossRef]
4. Wang, N.; Xu, H. Dynamics-constrained global-local hybrid path planning of an autonomous surface vehicle. IEEE Trans.
Veh. Technol. 2020, 69, 6928–6942. [CrossRef]
5. Villa, J.; Aaltonen, J.; Koskinen, K.T. Path-following with LiDAR-based obstacle avoidance of an unmanned surface vehicle in
harbor conditions. IEEE/ASME Trans. Mechatron. 2020, 25, 1812–1820. [CrossRef]
6. Ferri, G.; Manzi, A.; Fornai, F.; Ciuchi, F.; Laschi, C. The HydroNet ASV, a small-sized autonomous catamaran for real-time
monitoring of water quality: From design to missions at sea. IEEE J. Ocean. Eng. 2015, 40, 710–726. [CrossRef]
7. Kong, S.; Tian, M.; Qiu, C.; Wu, Z.; Yu, J. IWSCR: An intelligent water surface cleaner robot for collecting floating garbage.
IEEE Trans. Syst. Man Cybern. Syst. 2020. [CrossRef]
8. Wang, H.; Guo, F.; Yao, H.; He, S.; Xu, X. Collision avoidance planning method of USV based on improved ant colony optimization
algorithm. IEEE Access 2019, 7, 52964–52975. [CrossRef]
9. Kim, H.; Koo, J.; Kim, D.; Park, B.; Jo, Y.; Myung, H.; Lee, D. Vision-based real-time obstacle segmentation algorithm for
autonomous surface vehicle. IEEE Access 2019, 7, 179420–179428. [CrossRef]
10. Steccanella, L.; Bloisi, D.D.; Castellini, A.; Farinelli, A. Waterline and obstacle detection in images from low-cost autonomous
boats for environmental monitoring. Robot. Auton. Syst. 2020, 124, 1–11. [CrossRef]
11. Guardeño, R.; López, M.J.; Sánchez, J.; Consegliere, A. AutoTuning environment for static obstacle avoidance methods applied to
USVs. J. Mar. Sci. Eng. 2020, 8, 300. [CrossRef]
12. Li, Y.; Tian, L.; Li, W.; Li, J.; Wei, A.; Li, S.; Tong, R. Design and experiments of a water color remote sensing-oriented unmanned
surface vehicle. Sensors 2020, 20, 2183. [CrossRef] [PubMed]
13. Madeo, D.; Pozzebon, A.; Mocenni, C.; Bertoni, D. A low-cost unmanned surface vehicle for pervasive water quality monitoring.
IEEE Trans. Instrum. Meas. 2020, 69, 1433–1444. [CrossRef]
14. Cao, H.; Guo, Z.; Wang, S.; Cheng, H.; Zhan, C. Intelligent wide-area water quality monitoring and analysis system exploiting
unmanned surface vehicles and ensemble learning. Water 2020, 12, 681. [CrossRef]
15. Cryer, S.; Carvalho, F.; Wood, T.; Strong, J.A.; Brown, P.; Loucaides, S.; Young, A.; Sanders, R.; Evans, C. Evaluating the sensor-
equipped autonomous surface vehicle C-worker 4 as a tool for identifying coastal ocean acidification and changes in carbonate
chemistry. J. Mar. Sci. Eng. 2020, 8, 939. [CrossRef]
16. Ferri, G.; Jakuba, M.V.; Yoerger, D.R. A novel trigger-based method for hydrothermal vents prospecting using an autonomous
underwater robot. Auton. Robot. 2010, 29, 67–83. [CrossRef]
17. Alvarez, A.; Mourre, B. Optimum sampling designs for a glider-mooring observing network. J. Atmos. Ocean. Technol. 2012, 29,
601–612. [CrossRef]
18. Leonard, N.E.; Paley, D.A.; Davis, R.E.; Fratantoni, D.M.; Lekien, F.; Zhang, F. Coordinated control of an underwater glider fleet
in an adaptive ocean sampling field experiment in monterey bay. J. Field Robot. 2010, 27, 718–740. [CrossRef]
19. Palacin, J.; Salse, J.A.; Valganon, I.; Clua, X. Building a mobile robot for a floor-cleaning operation in domestic environments.
IEEE Trans. Instrum. Meas. 2004, 53, 1418–1424. [CrossRef]
20. Kang, M.C.; Kim, K.S.; Noh, D.K.; Han, J.W.; Ko, S.J. A robust obstacle detection method for robotic vacuum cleaners. IEEE Trans.
Consum. Electron. 2014, 60, 587–595. [CrossRef]
21. Zhang, H.; Zhang, J.; Zong, G.; Wang, W.; Liu, R. Sky cleaner 3: A real pneumatic climbing robot for glass-wall cleaning. IEEE
Robot. Autom. Mag. 2006, 13, 32–41. [CrossRef]
22. Bai, J.; Lian, S.; Liu, Z.; Wang, K.; Liu, D. Deep learning based robot for automatically picking up garbage on the grass. IEEE
Trans. Consum. Electron. 2018, 64, 382–389. [CrossRef]
23. Wang, Z.; Liu, Y.; Yip, H.W.; Peng, B.; Qiao, S.; He, S. Design and hydrodynamic modeling of a lake surface cleaning robot. In
Proceedings of the 2008 IEEE/ASME (AIM) International Conference on Advanced Intelligent Mechatronics, Xian, China, 2–5
July 2008; pp. 1343–1348.
24. Ruangpayoongsak, N.; Sumroengrit, J.; Leanglum, M. A floating waste scooper robot on water surface. In Proceedings of the 2017
17th International Conference on Control, Automation and Systems (ICCAS), Jeju, Korea, 18–21 October 2017; pp. 1543–1548.
25. Li, X.; Tian, M.; Kong, S.; Wu, L.; Yu, J. A modified YOLOv3 detection method for vision-based water surface garbage capture
robot. Int. J. Adv. Robot. Syst. 2020, 17, 1–11. [CrossRef]
26. Environmental Protection Administration Executive Yuan, R.O.C. (Taiwan). Available online: https://oaout.epa.gov.tw/law/
LawContent.aspx?id=FL015490 (accessed on 4 February 2021).
27. Lee, Y.S.; Juang, J.G. Color identification for quadcopter flight control and object inspection. Adv. Mech. Eng. 2019, 11,
1–16. [CrossRef]