853

1 INTRODUCTION

The most suitable solution for automatic surveilance

of water areas is using floating unmanned surface

vehicles - small-sized ship.

The USV could measure water parameters in-situ,

and make sampling for ongoing more detailed

laboratory research.

The main goal of current development is to design

complex solution including unmanned surface

vehicle (USV), custom autopilot firmware and

software for real-time depth visualization. In the

beginning of development we have pointed next

requirements for the USV:

− Easily accessible low-cost hardware parts

− Easy USV transportation

− Monohull

− Free software libraries and IDE

− Custom autopilot

− Custom base ground station

− Propulsion high reliability

− Maneuverability

− Minimal continuous USV work over 1 hour

− Custom software for collected data post-

processing

For making USV easy to transportation we have

determined the longest dimension is 1 meter, which is

depends of average size of car trunk. Monohull

design was chosen because of easy to proceed and

USV was targeted to work in calm waters as lakes

and internal bays where the high waves are rare.

2 HULL AND HARDWARE PART

USV was made of two main materials - fiberglass and

styrofoam as hull base. The hull base was CNC-

machined and glued together in one part (Figure 1).

Autonomous Unmanned Surface Vehicle for Water

Surface Monitoring

V. Nikishin, M. Durmanov & I. Skorik

Sevastopol State University, Sevastopol, Russia

ABSTRACT: The unmanned surface vehicle (USV) considered in this article is used for the purpose of

environmental monitoring of water areas. Small size of vehicle allows to move sensors to areas close to shore

line, under the low bridges where the manned boats difficult or danger to use because of damage. With its help,

measurements of the required parameters are carried out directly in the studied area, or water samples are

taken for further study in the laboratory. For this, it is necessary that the USV automatically move to

predetermined points of the route. This article describes one of the stages in the development of an automatic

control system for a given USV course. The problem of finding the steering angle for turning the USV to the

required points of the route is being solved. An algorithm for visualizing the depth measurements of the

exploration water area in the form of a heat map described.

http://www.transnav.eu

the International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 14

Number 4

December 2020

DOI: 10.12716/1001.14.04.09

854

Figure 1. CNC-machined styrofoam hull base

For minimizing interaction electronic parts with

water they are was placed over the top in separated

covered box mounted with PVC pipes glued to the

USV hull. Placing antennas as high as possible also

gives advantages for reliable connection with ground

station and easy access to electronic, also it worth

because of magnetometer is far enough from engine

containing strong magnets. Final results of the hull is

shown in Figure 2.

We have selected electrical propulsion system

because modern brushless and brushed motors has

enough power and they are has small size suitable for

using in USVs. Electrical propulsion system allows

fast changing battery and it does not pollute water by

any hazardous materials or exhaust gases. Our USV

uses 3S (12 Volts) battery, 4074 sized brushless motor

with 2000kV, total power over 1 kW. Motor is driven

by electric speed controlled (ESC). Because of ESC is

dissipates a lot of heat it is require cooling, cooling is

produced by water injected by water jet.

Figure 2. Unmanned surface vehicle

For steering and movement of USV we used water

jet. Water jet has advantages in safety and

maneuverability. Safety reason is very important if

USV works near the beaches with swimming people.

Because of water jet rotor located inside tube, so there

are almost impossible to injure someone by rotating

blades. Open propellers are more dangerous, also

there are big chances that floating garbage, seaweed

or fish lines could wound up on the shaft and

propeller blades. Even if water jet has no safety lattice

there are always problem of floating garbage and

seaweed (Figure 3).

Figure 3. Seaweed in water jet

For steering of USV was used rotating nozzle.

Nozzle has rotating angle near 30 degrees left and

right. This allows to get very good maneuverability.

3 AUTOPILOT

The main purpose of autopilot is to control USV by

rotating nozzle to steer it at given route point. This

requires sensors (gyro, accelerometer and

magnetometer) integration with microcontroller

(MCU). We selected Atmega2560 on ArduinoMega

board [1]. This MCU has enough interfaces for

connecting with sensors and does not require high

programming skill. For autonomous movement of

USV main measurements is current heading and

current location. Heading is measures by MEMS

magnetometer connected to MCU by I2C interface.

Magnetometer require calibration to minimize hard

and soft iron distortion. For measurement location of

USV uses GPS module U-Blox connected to MCU by

UART interface. For controlling nozzle and motor

ESC uses PWM signal generated by internal

hardware timer [2].

The paper is overviews control system algorithm

for heading hold of USV. Algorithm calculates turn

angle for rudder rotation for steering the USV. This

allow to steer USV in path-following mode to

checkpoints stored in autopilot memory.

For correct work of USV`s control system we have

to consider where USV have to be turn - left or right

when it moves on targeted path line. Thus, a

formalized description of the algorithm is needed,

according to which the calculated angle of rotation

will allow the USV to rotate along the smallest arc to

the azimuth required at the moment.

There are also external factors that influence the

movement of USV. For example, it can be a current or

wind that displaces the USV from the path line.

Therefore, here it is also necessary to know the angle

of rotation so that the USV can correct its movement.

But in order to make a turn, it is not enough to

know only the turning angle. You want to calculate

the steering angle. In the described BMA, such a

device is a rotating nozzle of a water jet.

αrud - steering angle of rudder from -20 ° to + 20 °.

In Figure 4 shows USV (1), its heading relative to

the north (2) is indicated, azimuth (3), i.e. the angle to

855

which the USV must turn in order to reach the path

line when moving to the given checkpoint.

α rudder.

Check

point

Path

line

Figure 4. USV testing

Control system consist of compass and GPS-

receiver. Compass in fact is 3-axis magnetometer.

Compass data uses for heading calculation.





180













arctg

(1)

X and Y – magnetic projection of Earth magnetic field.

GPS data uses for location of USV. Azimuth to

checkpoint calculates as:

)cos*cos*sin

sin*cos;cos*(sin2

212





−

−= arctg

(2)

where

Δλ= λ2 - λ1 – longitude difference;

φ1 and φ 2 – latitude difference.

First checkpoint is current USV location, second

checkpoint – next checkpoint of path-line.

If the azimuth value is negative, then 360 is

added to it.

In Figure 5 shows angle rotation calculation block

diagram.

GPS

Compass

Heading

Current USV location Checkpoint coordinates

Azimuth

Turn angle

Figure 5. Angle rotation calculation block diagram

Then USV heading and azimuth uses for turn

angle calculation:



−=

(3)

Figure 6 shows the difference of azimuth and

heading gives value of turn angle. But this is simplest

case. Some cases where additional correction

requires:

1 Δ – negative value;

2 Δ – more than 180. in this case USV will turn

ineffective at longest arc.

Check

point

Path

line

Figure 6. Navigational angles

1 – USV, 2 – current heading , θ - azimuth to checkpoint,

- turn angle, N – north heading

Calculation of rudder turn angle sown in Figure 7.

Δ > 180°

Δ < 0°

Δ < -180°

Нет

Да

Нет

Да

Нет

Δ = φ - θ

20*

180

360 −

рул



20*

180

360 +

−=

рул



20*

180



−=

рул



20*

180



−=

рул



Figure 7. Rudder angle calculation block-diagram

856

Given algorithm includes angle corrections. Since

the rudder of USV can rotates left and right for 20,

angle have to be divided to 180 and multiplies at 20.

Thus, rudder angle is calculates in allowed range

from -20 to 20. Negative values of rudder angle

means that USV steers to the right and vice versa

Figure 8.

20°

-11

20°

Figure 8. Rudder angle limits

The algorithm described in Figure 5 for calculating

the steering angle ensures the correct movement of

the USV along a path-line. During the movement, the

USV will turn to the checkpoint along the smallest

arc.

Also, the calculation of the steering angle allows

USV to automatically correct its movement relative to

the path-line.

Thus, the presented algorithm, due to its

formalized form, can be integrated into the USV

autopilot based on the microcontroller.

4 SONAR

The most useful part of USV development is

integration with measurement devices. Current

development is targeted to bathymetry. Selection of

sonar depends on many reasons as precision,

resolution, one or multi-beam ray. But in case of USV

the most important is connection with MCU - sonar

has to be equipped with suitable interface, in most

cases this is RS-232 with NMEA standard messages

[3]. We choose Garmin Echomap 42 [4], this sonar has

5 Volts NMEA interface, but it`s signal is inverted.

Because of Atmega2560 does not have option to

invert UART signal, we have to add inverter before

signal goes to MCU using NPN transistor (Figure 9).

Figure 9. NMEA inverter

Placing sonar in internal compartment protects it

from water sprays (Figure 10).

Figure 10. Placing sonar in USV

5 POST-PROCESSING SOFTWARE

DEVELOPMENT

5.1 Statement of the bathymetric data visualization

problem

It is proposed to visualize bathymetric data using a

heat map. To do this, select a color palette and fill

with the color each area with the same depth. The

initial data for building a heat map are:

− geographical coordinates of points (latitude and

longitude) and depth values measured by the

sonar at these points;

− geographical coordinates of the boundary points

of the test site, to create the boundaries of the

displayed heat map;

− color palette (range and number of color shades).

To solve this problem, there are special services.

For example, the Heatmaptool [5] online service

allows you to use various color palettes, control the

radius, zoom, and opacity of heat spots and update

the map in real time. However, this service is paid

and designed to visually display statistical data in the

selected region.

5.2 Heat Map algorithm

The problem of building a heat map is to fill on the

map not only the individual points at which depth

measurements were made, but to fill the entire area

with the color. This problem is solved using the

Inverse distance weighting (IDW) [6] method, which

consists in determining the weight of each point by

interpolating neighboring points with a known

weight at a given point.

The heat map algorithm consists from the

following steps [7]:

1 Converting the geographical coordinates of the

points to the x and y coordinates of the Cartesian

coordinates.

2 Formation of the border of the polygon for

cropping the resulting image of the heat map.

3 Implementation of the IDW method for

constructing a heat map image.

4 Converting points in a Cartesian coordinates to

geographical coordinates and combining a heat

map with a geographical map of the area.

857

The algorithm is implemented using developed

software in the C # programming language.

1 Primary processing of coordinates and

bathymetric data. The heat map is built according

to three parameters: geographic coordinates

(longitude and latitude) and depth. USV capture

this data, which is saved in a file with the .csv

extension. To process bathymetric information, it

is necessary to read the data from the file line by

line, split them into lexemes and write the

necessary parameters to the buffer. The read data

are in symbolic format and for further processing

and building a heat map, they are converted into

real numeric format.

The dimensions of the heat map field correspond

to the linear dimensions determined by the

maximum and minimum values of latitude and

longitude.

To process depth values, you need to convert them

from text format to numeric format and convert

them to a specified range.

2 Creation of polygon border for cropping the heat

map image. Polygon boundary is an array of

points located on the boundary of the exploration

water area. The border of the polygon for

cropping the image field is created at the stage of

forming the dimensions of the image field and

converting geographic coordinates to X0Y

coordinates. This is due to the fact that the

polygon is also specified by geographic

coordinates, which will need to be converted to

X0Y coordinates at the same scale.

In Figure 11 shows an image obtained using the

Google Earth web application, with the help of

which the geographic coordinates of this polygon

were obtained.

Figure 11. Exploration area of Inkerman career (Sevastopol)

in Google Earth application

Considering the coordinates of the polygon

corrected the linear dimensions of the plotting

area of the heat map image and converted the

geographic coordinates of the polygon into X0Y

coordinates. The image cropping contour is built

using these coordinates.

3 Building a heat map using the IDW method. A

heat map of the depth of an area of interest is a

raster image that is obtained by calculating

interpolated depth values using known depth

values at fixed points. Thus, interpolation is

performed in those places where we are not

interested in depth, for example, on land. In this

case, the resulting image must be cropped along

the contour of the exploration water area. Image

cropping is performed using the formed contour

along the boundary points of the polygon. The

color palette is chosen arbitrarily, usually warmer

tones correspond to smaller depths, colder ones

correspond to larger ones.

An results of building a depth map for an artificial

reservoir in Inkerman (Sevastopol) is shown in

Figure 12 and the coastal waters of Holland Bay

(Sevastopol) is shown in Figure 13.

Figure 12. Heat map of Inkerman (Sevastopol)

Figure 13. Heat map of Holland Bay (Sevastopol)

6 CONCLUSION

Developed USV shown in the paper has simple

design which is can be reproduced with minimal cost.

The main limitations of our USV is working only in

calm waters without strong wind and waves.

Small draft of USV less than 3 inches allows to

mapping depth of lakes or bays very close to

shoreline, this is almost impossible with big boats

with crew on board.

ACKNOWLEDGEMENTS

The reported study was funded of the internal grant

by Sevastopol State University, project number

516/06-31

858

REFERENCES

[1] Arduino Mega official site,

https://store.arduino.cc/arduino-mega-2560-rev3.

[2] Nikishin V. V., Durmanov M. A. “Firmware and

Software Structure Of Unmanned Naval Vehicle”, Int.

Conf. “Modern problems of radio electronics and

telecommunications”, Sevastopol State University,

Sevastopol, Russia, 2019.

[3] National Marine Electronics Association,

https://www.nmea.org.

[4] Garmin Echomap sonar, https://buy.garmin.com/en-

US/US/p/592785.

[5] HeatmapTool.com, https://www.heatmaptool.com.

[6] GisGeography. Inverse Distance Weighting (IDW)

Interpolation, https://gisgeography.com/inverse-

distance-weighting-idw-interpolation.

[7] Nikishin V. V., Durmanov M. A., Sizyh D. A. “Heatmap

for Water Depth Visualization”, Int. Conf. “Modern

problems of radio electronics and telecommunications”,

Sevastopol State University, Sevastopol, Russia, 2019.