117
1 INTRODUCTION
Thedynamicdevelopmentoflocalcommunitiesand
tourism in coastal regions and inland waterways
creates the need for new forms and functions of
transportation. Solutions for transferring road
transportto rail, wateror air aresought. Oneof the
manywaysofreducingroadandurbancongestionis
the possibilit
y of transferring wheeled vehicles to
water. The use of this means of transport is all the
more conducive when the shape of the area around
the water reservoir allows reducing the distance
between the transport points. Examples of such
activities are the initiative of many local authorities
seeking to launch water tra
ms between the most
frequented places. Waterway functions can be filled
inadhoc,usingexistingconventionalmannedunits,
or by undertaking long‐term planning, taking into
account functionality and reducing negative
environmental impact while maintaining maximum
safety standards and low operating costs, by
deployingfullyautonomousunits.
Figure1. Ecological conceptual models Vindskip and E/V
Orcelle[1].
Italsoimportanttopayattentiontothedesigned
units, so they emit the minimal pollution into the
environment. For this reason, engineered solutions
Design of an Autonomous Transport System for Coastal
Areas
A.Łebkowski
GdyniaMaritimeUniversity,Gdynia,Poland
ABSTRACT:Thearticlepresentsaprojectofanautonomoustransportsystemthatcanbedeployedincoastal
waters,baysorbetweenislands.Presentedsolutionsanddevelopmenttrendsinthetransportofautonomous
andunmannedunits(ghostships)arepresented.Thestructureofthecontrolsystemofaut
onomousunitsis
discussedtogetherwith thepresentation ofappliedsolutions inthefield ofartificial intelligence. The paper
presentstheconceptofatransportsystemconsistingofautonomouselectricpoweredvesselsdesignedtocarry
passengers, bikes, mopeds, motorcycles or passenger cars. The transport task is to be implemented in an
optimalway,tha
tis,mosteconomicallyandatthesametimeassafeaspossible.Forthisreason,thestructure
oftheelectricpropulsionsystemthatcanbefoundonsuchunitsisshown.Theresultsofsimulationstudiesof
autonomoussystemoperationusingsimulatorofmarinenavigationalenvironmentarepresented.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 12
Number 1
March 2018
DOI:10.12716/1001.12.01.13
118
includeenginesfedwithgaseousfuels,applicationof
sails, changing propulsion from gasoline/diesel to
electric power derived from photovoltaic panels or
hydrogenfuelcells[1‐6].
Figure2.WindChallengerProject[2,3].
Figure3.LNGFueledContainerShipbyKawasakiH.I.[4].
TherearealsoshipsthatuseFlettnerʹsaerodynamic
rotor with Magnus effect to support the classical
propulsion[5,6].
Figure4.VikingGracewithone Norse‐powerRotorSail[5].
In order to reduce losses associated with motion,
structures are designed, whose aerodynamic and
hydrodynamic resistance to the hull of the ship in
relation to its parts above and below the water, as
wellastheshapeofthepropulsors,arethesmallest.
Figure5.E‐Ship1withfourRotorSail[6].
With the possibility of developing services with
autonomous units, one can extend the range of
services offered by transporting passengers, goods
andvehiclesnotonlytothelocationsdefinedbythe
timetable,buttoeachofthetransshipmentpointsin
thearea.Inaddition,autonomousunitscanbeusedto
collect
pollutants and flotsam in coastal and inland
waters,aswellasaroundportsandmarinas.
Atthemoment,twobasic conceptsregardingthe
use of unmanned ships can be found. The first is
based on the ability to remotely navigate through
remotecommunicationsandcamera systems,andthe
secondemploys
artificialintelligence‐basedapproach
toautonomousvesselmanagementwithouttheneed
for remote operator supervision. Examples of this
type of structural design can be found in the Rolls‐
Royce project proposals for carriers using seagoing
shiphullscapableoftravelinglongdistancesbetween
ports[7].
Figure6.DroneCargoShipbyRollce‐Royce[7].
Another proposal from the consortium of
Massachusetts Institute of Technology (MIT), Delft
University of Technology (TUD) and Wageningen
University and Research (WUR) [8], is the urban
water transport concept for cities with access to a
large waterway network, such as: Venice,
Amsterdam,Rotterdam,Bruges,Stockholm,Bangkok,
Suzhou,ElGounaorSt.
Petersburg.
119
Figure7.AMSROBOATQ&A[8].
In turn, Kongsberg intends to launch a fully
automatic ferry for passenger carriage on a fixed
routearoundTrondheim[9].
Figure8. Conceptual model of a passenger‐cargo ferry by
Kongsberg[9].
In this paper, the author presents the concept of
coastal water transport organization using small
autonomous ecological passenger and passenger
transportvehicles.
2 STRUCTUREOFAUTONOMOUSTRANSPORT
SYSTEMFORCOASTALAREAS[1‐7]
In the last decades, many artificial intelligence
methodsandalgorithmsusedtosupportnavigational
decisionsinvariousDSS
systemshavebeenfoundin
world literature [10 ‐25]. Part of the presented
methodsandalgorithmsaftersomemodificationsare
suitable for use in the management of autonomous
maritimeunits.
For steering autonomous units in coastal and
inland waters, the author proposes to use an agent
system [23‐25]. For
the first time, the agent system
wasproposedtodirectvesseltrafficin2007[23].The
agentsystemconsistsofanITagentplatformwhere
agentsaredeployedtoperformspecifictasks.Agent
platforms are installed on computers that control
autonomousunitsandshorestations.
A
N
A
T
A
SN
Figure9.Structureoftheplatformoftheagentsystem.
Thesingleagentplatformconsistsofthreeagents:
AT‐trajectoryagent,ASN‐navigationalagent,and
AN‐negotiation agent. The task of the trajectory
agentATistodeterminetherouteoftheautonomous
unit, based on the information provided by the
navigationagentfromtheshorestation.Atthesame
time,
thetrajectoryagentisresponsibleforcorrecting
agivenroutechangeintheeventofacollisionhazard
(otherunits movinginagivenareaandnavigational
limits). The trajectory agent is also responsible for
precise steering of the autonomous unit, including
precisemooring,aswellasdeparturefromthe
wharf.
TheroleoftheASNistocollectinformationaboutthe
currentnavigationalsituationaroundtheshipandits
in‐depth analysis of collision hazards with dynamic
objects as well as static objects, which are mostly
technical facilities of the autonomous community
harbors. Geographic information of the navigational
environmentin
whichtheautonomousunitmovesis
mainly provided by the ECDIS in combination with
the LIDAR sensor system or optionally video
cameras.ThetaskoftheANnegotiatoristonegotiate
with other agents located on the platforms of the
autonomouscoastalandinlandtransportsystem.For
thispurpose,the
negotiatingagentmayinteractwith
agents located on the agentsʹ platforms of other
autonomous entities, but also receive and transmit
informationfromagentslocatedattheshorestations.
Thanks to such a system structure, it is possible to
transfer information between autonomous units
operatingina givenwatercourseto avoidcollisions,
but also to disseminate information related to
dynamicallyemergingthreatstotrafficsafety.
Theconfigurationand tasksof theagent platform
locatedontheshorestationareslightlydifferentthan
the agent platform located on autonomous units.
Similar to the structure of an autonomous agent
platform, the agent platform of
the shore station
consists of three agents: AC‐a communications
agent, an ASN‐a navigation agent and an AN –
negotiatoragent.
ThetaskoftheCoastCommunicationsAgentisto
handle customer orders. AC analyzes orders
submitted by stand‐alone clients and users and
transferscoordinatestoANʹsCoastal
Agentplatform
negotiator, which is responsible for transferring the
geographiccoordinatesofthepointoftakingdatatoa
specific standalone unit with the type of execution
order‐passenger or passenger transport. The AN
120
negotiator will also exchange information related to
the technical condition of autonomous units (fuel
quantity,batterycharge)inadditiontothenavigation
information provided by the tank. The negotiating
agent has a higher decision factor from negotiation
agents located on autonomous units. Owing to this
property, shore station negotiation agencies
can
providecommandsfortheorganizationoftraffictoa
giventrajectoryagentonaplatformthatislocatedon
autonomousagentplatforms[24].
Inordertouseanautonomouswatercraftunit,the
user/clientmustlogontotheuserdatabaseviathe
web site. Then the user
must send an SMS message
fromtheregistered telephone number indicating the
locationhe /she wouldlike touse, the startdate of
the transport, the mode of transport (passenger
transportorvehicletransport)andthedestination.In
return, the user will receive information about the
confirmation of selected booking
parameters,
optionally, the necessities to make changes to the
reservation(e.g.,informationfortheuseroninability
to land the vehicle at the specified location) or the
nearest possible date of the transport order. Upon
arrivalattheplaceofboarding,theuserconfirmshis
arrival by entering the code
on the touch panel
mountedonboardthecraft.Theuser,dependingon
the transport order, will be informed through the
interactive communication interface about the steps
he/sheshouldtaketosafelycommencethejourney.
This is particularly important when choosing the
transportation optionof a passenger vehicle and
the
relatedloadingandsecuringofthevehicle.
Autonomous vehicle control is performed
automatically according to the userʹs destination or
fixed route. When ordering a transport by the
autonomous craft, it is possible to select specific
locationsinagivenarea,ormakeselectionatouristic
orscenicroute.
The task of the agent system is to direct the
autonomous units to the user pickup points in a
timelyfashionsothatthewaitingtimeisasshortas
possible.
The structure of the agent system controlling the
movement of autonomous, ecological units is
presentedinFigure10.
Coastal waterborne
transport using small
autonomousecologicalunitscanbecarriedoutusing
digitalVHFradio,GPRSandAIS.
The task of the agent system is to automatically
andautonomously conductthe transport on a given
watercourse. It was necessary to develop a way of
communicatingbetweenagents operatingwithin the
agent
platformaswellasthewayofexchangingdata
between agent platforms occurring throughout the
system.Agentactionsarecharacterizedbytheability
tomake autonomousdecisions without theneed for
operator intervention. Agents have the ability to
exchangeinformationwiththeoperator,buttheymay
also attempt to negotiate their
maneuvers with
navigatorsofotherunits operatinginthatparticular
area. Agents also have the ability to communicate
with other agents working within a single platform
andtheentireagentsystemsothatdynamicchanges
in the navigational environment are properly
interpretedandoptimaldecisionsaremadeonthem.
Figure10. Structure of the agent system controlling the
movementofautonomous,ecologicalunits.
3 SEAAUTONOMOUSSHUTTLE
For a coastal waterway system using small
autonomous ecological passenger, bicycle, mopeds,
motorcycles or passenger vehicles, passengers must
use adequately equipped vessels like the Sea
Autonomous Shuttle (SAS). The structure of the
autonomous equipment of the ecological unit is
showninFigure11.
SAS operation is based
on the information
provided by the devices and systems installed on
board the unit as well as transmitted by radio from
shore stations. Operation of the autonomous unit is
mainly based on data received from the GPS /
GLONASS / GALILEO, AIS, radar, ARPA,
anemometer, log, echo sounder and electronic
mapping system. Additionally, in particular during
mooring, signals from the LIDAR sensor system or
optionalvideocamerasystemareused.Afterdefining
a destination for SAS, the control system starts the
designated route. The route is determined based on
thegeographicalcoordinatesofthecurrentpositionof
theSASunitand
thedestination.Usingthetechnique
of evolutionary algorithms [24], the route for the
autonomousvesselisdetermined.Thecorrectrouteis
takencareofbytheATtrajectory agent.
E-map
LOG
Echosounder
AIS
RADARGPS
GPRS
VHF
L
I
D
A
R
Figure11. Structure of the autonomous vessel
equipment.
121
Whenacollisionriskforthevesselisdetected,the
navigation situation is analyzed and appropriate
action is taken, depending on its level of security.
These actions include negotiating with encountered
unitstodeterminetheoptimalrouteforthetransition.
Whennegotiatingisimpossible, anoptimalroutefor
aparticular
navigationsituationisdeterminedor an
anti‐collision maneuver is determined. Then the
process of automatic control after the route of the
transitioniscarriedout.Thisprocessiscarriedoutby
the trajectory agent, which also includes the precise
controloftheautonomousunitwhendepartingfrom
the berth,
mooring and maneuvering in narrow
passages.Theseactionsarepossiblewithinformation
fromthepositioningsystem,theradarsystem,andin
particular the LIDAR sensor system mounted in the
corneroftheautonomousvesseloroptionallyofthe
video camera system. The data received is
appropriatelytransformedsothat thealgorithm
that
determinestheshipʹsrouteandtheprecisecontrolof
following that route can be processed safely. 3D
objectsidentifiedinthemaritimenavigationareabya
radar, AIS, LIDAR sensor system or optional video
camera system are brought to the 2D plane. As a
result,allobjectsprojected
ona2Dplanearecollision
hazards for a ship on which a standalone agent
platform is installed. Depending on the motion
parameters, identified objects can be dynamic or
static. For such specific navigational limits, artificial
intelligence methods can be used to determine the
optimalrouteoftransitions.Evolutionaryalgorithms
are
used in the described system of directing the
movementofautonomouscraftsincoastalandinland
waterstodetermineoptimaltransitionroutes.
Theautonomous,ecologicalSAScraftisequipped
witha5kWelectricdrivepoweredbyabatterypack.
In addition, the autonomousunit for precise control
uses two thrusters:
bow and stern thruster, with
powerof1kWeach.Theenergystoredinthebattery
pack allows the unit to be operated for about 10
hours.In addition, the battery pack can be powered
byanauxiliarygenerator,whichallowsittooperate
when the energy in the main battery pack
is
consumed.Ontheroofoftheunit,andonthefrontof
the deck, there are photovoltaic panel units which
supplyelectricity directlyto aninverter that powers
an electric motor or to the battery pack. The unit is
also equipped with connectors to automatically
connecttothebattery
chargerinstalledonshore.
Figure12. Appearance of an exemplary autonomous unit
withpositioningsensors.
4 VERIFICATIONOFTHEAUTONOMOUS
TRANSPORTSYSTEMFORCOASTALAREAS
A network simulator of the navigational navigation
environmentwasusedtoverifytheoperation ofthe
coastal waterway system using small autonomous
ecological units. This simulator has the ability to
execute navigational scenarios along with the
parametersofthewatercoursein
whichtheyoperate.
Thesimulatorconsistsofacentralunit‐theserveron
which the navigation environment is modeled and
localstationsemulatingsingleautonomousunits.The
navigational scenarios concerned the verification of
thesysteminconflictsituationsbetweenautonomous
entities. The mathematical model of the developed
modelofthe
autonomousunitincludesthedynamic
propertiesofthehull,themainpropulsionconsisting
ofasinglefixedpitchpropeller,afinrudderandtwo
transversethrusters:fore andaft. Themodel of ship
dynamics also takes into account the influence of
hydrometeorological disturbances in the form of
wind, wave and sea
currents, as well as changes in
vessel dynamics caused by the shallow‐water effect
[25].
Verification of the motion control system of
autonomous units in coastal and inland waters was
basedon a seriesof computer simulationsusing the
simulatorofthenavigationalnavigationenvironment.
Below is an example of
a navigational situation
illustratingthemovementofautonomousunitsinthe
watersoftheGdańskBay(Figure13).
Figure13.Navigationsituation depictingthemovementof
autonomousunitsinthewatersoftheGdańskBay.
Agent platforms were launched on all units
involved in the imaged situation. Out of 6 units, 4
were equipped with an agent system. The motion
parameters and weather conditions parameters are
presentedintheFigure14.
The routes of Unit 1 and 2 are designed to
transport passengers from boarding place
to
destination place and at the same time to take
advantageofthetouristmodesothattheindividual
routecanbeplannedtoincludeareaswithinteresting
places or views of natural landmarks. Based on the
simulations,itcan beconcludedthat thesystem has
correctly implemented the control
process between
thestartingpointandthetargetpoint.Intheeventof
a collision risk, in this case exceeding the Closest
Point of Approach distance (CPA), the system
correctly corrected the current route. By using an
agent system to control autonomous units, it is
possible to set smaller values for
the smallest CPA
122
distance parameter, given that the individual agent
platforms have precise information about the route
parameters of other units that cooperate within the
system.
Figure14.Motionparametersofselectedautonomousunits
anddescriptionofmeteorologicalconditions.
Defining smaller CPA values has a direct impact
onreducingthesizeofareasaroundtheaffectedunits
in a given area, the violation of which results in an
increased risk to shipping safety. Defined areas are
taken into account by a specialized evolutionary
algorithm [24] as one of the artificial
intelligence
methods used by the AT agent trajectory agent to
determine the optimal route for the current
navigational conditions. This gives us the ability to
designate the shortest and the most secure passage
route. Unfortunately, this comfort is not guaranteed
by an agent system when a vessel is moving in
a
given area, and is not possible to establish
communication with it and thus negotiation by AN
negotiatorisimpossible.Inthiscase,theagentsystem
willperformtheroute determinationprocessforthe
standaloneunitbasedontheinformationprovidedby
the radar, LIDAR and AIS systems, taking into
account
the International Maritime Organizationʹs
COLREGsregulations.
In the next navigational situation, 5 autonomous
unitsequippedwithanagentsystemwereinvolved.
The motion parameters and weather conditions
parametersarepresentedinFigure15.Thankstothe
cooperation of agents installed on the agent
platforms, the optimal operation of the
autonomous
units from the starting point to the destination was
achieved. This gives one the potential to reduce the
operating costs by saving time and reducing energy
consumption.
123
Figure15.Motionparametersofselectedautonomousunits
anddescriptionofmeteorologicalconditions.
5 CONCLUSIONS
Simulationstudieshaveshownthattheuseofan
agent system to manage an autonomous coastal
water transport system using small ecological
unitshaspositiveeffects.
The use of the system in practice enhances the
safety of the users of the area, including rescue
and exploration operations
where autonomous
entitiescanparticipatebyusingthevideocamera
system.
Research has shown that even when a single
autonomous unit loses contact with a shore
station, it is possible for it to operate
autonomously in the navigation environment.
Using an agent system reduces the risk of
collisions with
other objects, raising the level of
securityinagivenarea.
The use of autonomous units opens up new
opportunitiesfortourismandbusinessservices.
Themainobstacletotheuseofautonomousunits
are legal regulations and issues of legal and
financialliabilityintheeventofan
accident.
The use of fully automated autonomous units
increases the level of navigational safety and the
efficiency of coastal and inland transport. The
algorithms used to determine the route of travel
fully respect the rules of the International
Maritime Organization COLREGs, which cannot
besaidofmanynavigators.
The use of autonomous units introduces new
possibilities for ship design in the future, where
crew social rooms can be repurposed to handle
morecargoinstead.
REFERENCES
[1]ShadboltP.,ʹVindskipʹcargoshipusesitshullasagiant
sail.CNN11.2015,www.edition.cnn.com.
[2]Ouchi,K., Zhu,T., Hirata,J., Tanaka,Y., Kawagoe,Y.,
Takashina,J., Suzuki,K.: Development of Energy‐saving
WindshieldforLargeContainerShip,Conferenceproc.,
the Japan Society of Naval Architects and Ocean
Engineers,21,p.159‐162,2015.11(inJapanese).
[3]
Hane,F., Aoki,I., Ouchi,K.: Development of Auto‐pilot
forSailingShip,Conferenceproc.,theJapanSocietyof
Naval Architects and Ocean Engineers, 21, p.171‐173,
2015.11(inJapanese).
[4]www.global.kawasaki.com,(12.2016).
[5]NorsepowerRotorSailSolution.www.norsepower.com,
(01.2017).
[6]DailyNews.E‐Ship1winswindpropulsioninnovation
award.No.4,
2016.
[7]Autonomous ships The next step. Ship Intelligence
Marine. Rollce‐Royce 2016. www.rollce‐
royce.com/marine.
[8]Amsterdam to Get World’s First Fleet of Autonomous
Boats.www.ams‐institute.org/roboat‐qa/,(09.2016).
[9]Automated Ships, Kongsberg to build fully‐automated
vessel for offshore operations. www.safety4sea.com,
(11.2016).
[10]Escario J.B., Jimenez J.F., Giron‐Sierra J.M.,
Optimisationofautonomousshipmanoeuvresapplying
AntColonyOptimisationmetaheuristic.ExpertSystems
With Applications, Vol. 39, Issue 11, 2012, p.10120‐
10139.
[11]Escario J.B., Jimenez J.F., Giron‐Sierra J.M.,
OptimizationofAutonomousShipManeuversapplying
Swarm Intelligence. IEEE International Conference on
Systems Man and Cybernetics Conference Proceedings
2010.
[12]Ottesen
A.E.,SituationAwarenessinRemoteOperation
ofAutonomousShips.ShoreControlCenterGuidelines
Norway2014.
[13]Gierusz W.,Łebkowski A., The researching ship
ʺGdyniaʺ.PolishMaritimeResearch,Vol.19,2012,p.11‐
18.
[14]Kula K., Autopilot Using the Nonlinear Inverse Ship
Model. Activities In Navigation: Marine Navigation
AndSafety
OfSeaTransportation,2015,p.101‐107.
[15]Kula K., On‐line autotuning of PID controller for
desired closed‐loop response. 20th International
ConferenceonMethodsandModelsinAutomationand
Robotics(MMAR)2015,p.707‐711.
[16]KulaK.,Model‐basedcontrollerforshiptrack‐keeping
using Neural Network. IEEE
2ND International
Conference on Cybernetics (CYBCONF 2015), 2015, p.
178‐183.
[17]LevanderO.,Autonomousships onthehighseas.IEEE
Journals&Magazines,Vol.54,Issue2,2017, p.26‐31.
[18]Johansen T.A., Perez T., Cristofaro A., Ship Collision
Avoidance and COLREGS Compliance Using
Simulation‐Based Control Behavior
Selection With
Predictive Hazard Assessment. IEEE Transactions on
Intelligent Transportation Systems. Vol. 17, Issue 12,
2016,p.3407‐3422.
[19]Perera L.P., Oliveira P., Soares C.G., System
Identification of Vessel Steering With Unstructured
124
UncertaintiesbyPersistent Excitation Maneuvers. IEEE
JournalofOceanicEngineering,Vol.41,Isuue3,2016,p.
515‐528.
[20]Lisowski J., Computational intelligence methods of a
safe ship control. Procedia Computer Science, Vol. 35,
2014,p.634‐643.
[21]Lisowski J., Optimization‐supported decision‐making
in the marine game environment. Solid
State
Phenomena,Vol.210,2014,p.215‐222.
[22]Lisowski J., Game control methods in avoidance of
ships collisions. Polish Maritime Research,Vol. 19,
2012,p.3‐10.
[23]ŁebkowskiA.,DziedzickiK.,AgentSysteminDirecting
theMovementoftheShip.I
st
InternationalTech‐Science
Conference NATCON 2007, Scientific Journal of Polish
NavalAcademy,No.170,2007,p.1‐12.
[24]Łebkowski A., Evolutionary Methods in the
Management of Vessel Traffic. Information,
CommunicationandEnvironment2015,p.259‐266.
[25]ŁebkowskiA.,Controlofshipmovementbytheagent
system.PolishJournal
ofEnvironmentalStudiesVol.17,
No.3C,2008.
[26]Łebkowski A., Negotiations between the agent
platforms. Scientific Publications Gdynia Maritime
University,Gdynia2013.
[27]Mohamed‐Seghir M., Safe shipʹs control in a fuzzy
environment using a genetic algorithm. Solid State
PhenomenaVol.180,2012,p.70‐75.
[28]Mohamed‐Seghir M.,
The branch‐and‐bound method,
genetic algorithm, and dynamic programming to
determine a safe ship trajectory in fuzzy environment.
ProcediaComputerScience,Vol.35,2014,p.348‐357.
[29]Mohamed‐Seghir M., Computational Intelligence
MethodforShipTrajectoryPlanning.21STInternational
Conference On Methods And Models In Automation
AndRobotics
(MMAR),2016,p.636‐640.
[30]RybczakM.,LinearMatrixInequalitiesinmultivariable
shipʹssteering.PolishMaritime Research,Vol.19,2012,
p.37‐44.
[31]GieruszW.,TomeraM.,Logicthrustallocationapplied
to multivariable control of the training ship. Control
EngineeringPractice,Vol.14,Issue5,2006,p.511
‐524.
[32]Gierusz W., Waszkiel A., Determination of suction
forcesandmomentonparallelmanoeuvringvesselsfor
a future control system. Mechatronic Systems,
Mechanics And Materials,Book Series: Solid State
Phenomena,Vol.180,2012,p.281‐287.
[33]TomeraM.,Amultivariablelowspeedcontrollerfora
ship autopilot with
experimental results. 20th
International Conference on Methods and Models in
AutomationandRobotics(MMAR),2015,p.17‐22.
[34]MillerA.,InteractionForcesBetweenTwoShipsDuring
Underway Replenishment. Journal of Navigation, Vol.
69,Issue6,2016,p.1197‐1214.
[35]Weintrit A., Neumann T., Safety of Marine Transport
Introduction.2015,
p.9‐10.
[36]WeintritA.,E‐NavigationRevolution‐MaritimeCloud
Concept. Communications in Computer and
InformationScience,Vol.471,2014,p.80‐90.
