289
1 INTRODUCTION
Approximately,90%oftheworldtradeiscapturedby
the maritime industry, comprising more than 60,000
merchant ships and over a million seafarers [1]. In
recent years, the maritime industry has strived to
improvereliabilityandshipstructuretominimisethe
riskofcasualtiesandimproveperformance.However,
thecasualtiesarestillhigh.Oneofthekeyreasonsis
the heavy dependence of the system on human
performance, which leads to human error and thus
losses[2].Anyactorlackofactionthatresultsinor
contributes to a casualty or near‐casualty is
considered a human
error by the Maritime
TransportationResearchBoard[3].
An analysis of 39 collisions reveals that most
maritimeaccidentsarecausedbydecisionerrors[4].
Afewcontributingfactorsatthepreconditionallevel
included an operatorʹslack of situational awareness,
attention deficit, and inefficient inter‐ship
communication. There was often inadequate
operational planning at the leadership level [5].
Human error and/or violations are estimated to be
responsiblefor75%to96%ofmaritimeaccidents[1].
Severalstudies are being conducted thatexamine
near‐miss errors that could occur in the daily
operationsofmachineryspaces.Thesetasksareoften
performed by
marine engineers based on their
experienceratherthanfollowingrules.TheAustralian
SafetyBureau reports that mostoil spills are caused
by errors during the critical task of transferring oil
between tanks [6]. Numerous studies have been
conducted on techniques for preventing maritime
accidents. Research has evolved over the last half
‐
century from focusing on naval architecture to
examining human error. It is likely to continue
Optimization of Daily Operations in the Marine Industr
y
Using Ant Colony Optimization (ACO)-An Artificial
Intelligence (AI) Approach
A.Sardar
1
,M.Anantharaman
1
,V.Garaniya
1
&F.Khan
2
1
UniversityofTasmania,Launceston,Tasmania,Australia
2
MemorialUniversity,St.John’s,NewfoundlandandLabrador,Canada
ABSTRACT:Themaritimeindustryplaysacrucialroleintheglobaleconomy,withroughly90%ofworldtrade
beingconductedthroughtheuseofmerchantshipsandmorethanamillionseafarers.Despiterecenteffortsto
improvereliabilityandshipstructure,the
heavydependenceonhumanperformancehasledtoahighnumber
ofcasualtiesintheindustry.Decisionerrorsaretheprimarycauseofmaritimeaccidents,withfactorssuchas
lackofsituationalawarenessandattentiondeficitcontributingtotheseerrors.Toaddressthisissue,thestudy
proposes an Ant Colony Optimization
(ACO) based algorithm to design and validate a verified set of
instructionsforperformingeachdailyoperationaltaskinastandardisedmanner.ThisAI‐basedapproachcan
optimisethepathforcomplextasks,provideclearandsequentialinstructions,improveefficiency,andreduce
thelikelihoodofhumanerrorbyminimisingpersonal
preferenceandfalseassumptions.Theproposedsolution
can be transformed into a globally accessible, standardised instructions manual, which can significantly
contributetominimisinghumanerrorduringdailyoperationaltasksonships.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 17
Number 2
June 2023
DOI:10.12716/1001.17.02.04
290
exploring the socioeconomic factors involved in
maritimeaccidents[7].
Recently, the maritime industry has taken
initiatives to globalize standards to ensure safety on
the vessels. The International Convention for the
SafetyofLifeatSea(SOLAS)isconsideredtobethe
mostwidelyacceptedtreatyforthesafetyofmerchant
ships. In 1994, the International Maritime
Organization (IMO) added International Safety
Management Code (ISM) to SOLAS, to improve
management,operations,andpreventivemea surefor
pollution on ships. Regulations for maritime safety
are developed using a Formal Safety Assessment
(FSA). Human variables, technical variables, and
organizational variables are all considered by
FSA.
However,itdoeshaveitslimitations.Relyingonthe
expertʹsquantitativeriskassessmentputstheFSAin
the realm of human error [8]. Human‐technology
interaction has been optimised through the
development of novel techniques for enhanced
automation monitoring and appropriate risk
assessment[9].
With the advent of technology
and its
implementation to improve engineering systems,
artificial intelligence (AI) has been usedto optimise,
improve, and implement the best strategies to cope
with problems on vessels. However, the focus has
beenrestrictedtoemergencysituationsonly,whereas
humanerrorequallycontributestodaily operational
tasks. Our research aims at
providing a potential
solutiontoreducetheriskofhumanerrorduringthe
dailyoperationaltasksonvessels.Weproposetheuse
of an Ant Colony Optimization (ACO) based
algorithm to design and validate a verified set of
instructionsforperformingeachoperationaltaskthat
is integrated, globally accessible, and standardised.
Following an AI‐based approach would help us
optimisethe path for complextasks, provide a clear
setofsequentialinstructions,improveefficiency,and
reducethe likelihood ofhumanerror by minimising
personalpreferenceandfalseassumptions.Moreover,
these are meant to be created once and for all,
applicable to
all operational tasks, andapplicable to
alltypesofships.
Originallydevelopedin1992byMarcoDorigoto
solve the Travelling Salesman Problem (TSP), the
ACOhasproventobebeneficial[10].Itsaimwasto
findtheshortestHamiltoniancycletrackbetweentwo
citiesandback.AntColonyOptimization(ACO)
isa
metaheuristic algorithm inspired by the natural
behaviourofantstofigureoutthebestpathfromone
point to another. It is an artificial intelligence (AI)
based technique, which simulates ant behaviour
based on their trail of pheromones. It explores
differentpathsusingpheromonetrailsandthenfinds
the best solution using this process. In the maritime
industry, ACO can be implemented to optimise
engine operations by searching for the best
parameters and procedures to carry out a task. For
instance, optimization of fuel consumption [11],
managingengineemissions[12],enginemaintenance
andrepair[13],navigation,andweatherrouting
[14],
are some of the areas where it can be implemented.
PathplanningproblemscanbesolvedusingtheACO
algorithminthemaritimeindustry,toavoidcollisions
[15].Maritimeappliedthesamealgorithmtooptimise
group project management, including efficient and
effectiveschedulingoffunds,manpower,equipment,
and materials
[16]. The algorithm can consider
various factors simultaneously. Despite the fact that
theACOalgorithmisaheuristicmethodanddoesnʹt
guarantee global optimality, it is able to provide a
reliable result within a reasonable time frame. A
variety of problems related to combinatorial
optimizationcanbehandledusingthe
algorithm[17].
2 METHODOLOGY
The oil transfer pump is a crucial component found
onallships.Itsmainpurposeistotransferoil,which
is not only necessary for reaching a vesselʹs
destination but also poses a significant risk if
mishandled[18].Therefore,itisimperativetoensure
that
marine oil transfer pumps are reliable and
efficient. In most cases, ships utilise heavy fuel oil
(HFO)topowertheirengines,whichisfirststoredin
bunkertankslocatedinthedoublebottom,afterbeing
received from the port or bunker barge. The oil
transferpumpthentransferstheHFOto
thesettling
tank in the engine room, which is its primary
function.
The marine industry heavilyreliesonoil transfer
pumps for various purposes, including returning
excessoilfromtheoverflowtanktothestoragetank
andtransferringdieseloilfrom one tanktoanother.
The importance of having reliable
oil transfer
operations cannot be overstated since even the
slightesterrorcanresultinhazardoussituations,such
as oil spills. Therefore, it is crucial to prioritise the
maintenanceandproperfunctioningofthesepumps
inthemarineindustry.
Oiltransferisacrucialtaskonboardships,but it
also poses significant
hazards. To ensure safety,
automatic alarms, and switches should be in place.
Each tank, including the settling and service tanks,
should have low and high alarms. The low alarm
signals the need for more oil, while the high alarm
and switch alert the crew when the tank is close to
overflowing.Insuchasituation,theoiltransferpump
shouldautomaticallyswitchoff.Ifnot,analarmwill
sound,andengineersmustmanuallyshutthepump
off as excess oil flows into the overflow tank.
Althoughautomaticswitchesarenecessary,itisalso
vital to have manual controls and an emergency
switch located outside of the engine spaces. This
allows for quick response in case of a dangerous
situation,suchasanoilspillorfire,andallowscrew
memberstostoptheoilflowfromasafelocation.
Several research studies have investigated near‐
miss errors that may occur during
daily operational
tasksinmachineryspaces,whichcanpotentiallylead
toaccidents.Inmanycases,marineengineersrelyon
their experience rather than following a set of
established rules when performing these tasks [19].
The Australian Transport Safety Bureau (ATSB)
reportsthatthemajorityofoilspillsresultfromerrors
during
the critical task of transferring oil from one
tanktoanother.Thesefindingshighlighttheneedfor
effectiveguidelinesandtrainingtopreventaccidents
and ensure safe practices in the maritime industry
[20].
291
Our study suggests a solution to reduce the
occurrence of human errors in ship operations. We
recommendthecreationofastandardizedinstruction
manual,whichwouldbegloballyaccessibleandfully
integrated.Thismanualwouldprovideaverifiedset
ofinstructionsforeachoperationaltaskandfollowing
it would reduce
the likelihood of human error. To
developandverifytheseinstructions,weproposethe
use of an Ant Colony Optimization (ACO) based
algorithm.
3 ANTCOLONYOPTIMIZATION–THE
CONCEPT
Thebasicideaofantcolonyoptimization(ACO)isto
simulatethe behaviour of antssearching for foodin
order
tofind an optimalsolutiontoan optimization
problem. The algorithm works by constructing
solutionsiteratively,witheachiterationconsistingof
a set of ants building asolutionby movingthrough
the problemspace. The conceptual representation of
ACOforpathoptimizationisshowninFigure1.
Todothis,ACO
usesagraphrepresentationofthe
problem, where the nodes represent the decision
variables, and the edges represent the possible
solutionsorconnectionsbetweenthosevariables.The
ants move through this graph, leaving pheromone
trailsontheedgestheytraverse.Thesetrailsrepresent
thecumulativeexperienceofthecolonyand
influence
the probability that other ants will follow the same
pathinthefuture.
Figure1. Concept diagram of Ant Colony Optimization
(ACO)Algorithm.WhereNandFdenoteNestandFood,a
istheongoingdirectionandbisreturningdirection.Part1:
shows the early process where ants start finding a path
between nest and foodand lay pheromone. Part 2: shows
the
intermediate process where ants went through all
possible paths. Part 3: shows the most adopted path with
thehighestpheromonelevel.
Here are some of the key equations used in the
ACOalgorithm:
Pheromonetrailupdateequation:
1
ij ij ij
p
(1)
where τ
ij is the amount of pheromone on the edge
connecting nodes i and j, ρ is the pheromone
evaporationrate,andΔτ
ijistheamountofpheromone
deposited on the edge by the ants that found a
solutionusingthatedge.
Probabilisticsolutionconstructionequation:
ij ij
ij
kl kl
tt
pt
tt
(2)
where p
ij(t) is the probability of an antchoosing the
edge connecting nodes i and j at time t, τ
ij(t) is the
amountof pheromoneon the edgeat time t,η
ij(t) is
theheuristicvalueoftheedgeattimet,αandβare
parameters that control the influence of the
pheromoneandheuristicinformation,andthesumis
taken over all possible edges k and l at the antʹs
currentlocation.
Localpheromoneupdateequation:
0
1
ij ij
(3)
where τ
0 is a constant value representing the initial
amount of pheromone on the edges, and α is a
parameter that controls the influence of local
pheromoneupdates.Thisequationisusedtoupdate
thepheromonetrailontheedgeimmediatelyafteran
anthasusedittoconstructasolution.
Global
pheromoneupdateequation:
1
ij ij best
(4)
whereΔτ
bestistheamountofpheromonedepositedon
theedgebythebestantinthecurrentiteration.This
equationisusedtoupdatethepheromonetrailonall
edgesaftereachiteration.
Objectivefunction:
f(x)=objectivefunctionvalueforsolutionx (5)
The objective function is used to evaluate
the
qualityofthesolutionsfoundbytheants.
Thegoalof the optimization problem is to finda
solution that maximises or minimises this function.
Overall, the ACO algorithm is a powerful
optimizationtechnique thathasbeenusedtosolvea
wide range of complex problems in fields such as
transportation,logistics,andtelecommunications.The
key to its successis the combination of probabilistic
solution construction, pheromone trail updates, and
heuristicinformationtoguidethesearchprocess.
4 CASESTUDY
Title: Optimization of Diesel Oil (DO) Transfer
ProcessinMaritimeUsingAntColonyOptimization
Introduction: The purification of diesel oil
is a
criticalprocessinthemaritimeindustry,asitensures
thesafeandefficientoperationofmarineengines.The
process involves removingimpuritiessuchaswater,
sediment, and other contaminants from the fuel.
Figure 2 presents the components involved in the
292
initiation of the DO transferring process from a
storagetanktoaservicetank.Thereareseveralmulti‐
step complex daily operational tasks involved to
complete the DO purification process. The person
interactingwiththeseoperationscertainlyknowshow
to carry an individual task but the sequence and
distance are
not optimised and are either based on
human preference or subjected to random choice.
Optimization of such complex processes using AI
could reduce the overall time of carrying out the
processandwouldmakeitmoreefficient.Inthiscase
study,wewilluseantcolonyoptimizationtooptimise
the
purificationprocessforamarinevessel,withthe
goal of maximising the efficiency of theprocess and
minimisingthetimerequiredforpurification.
Figure2. Diesel oil purification piping diagram of M.T
FERU.
Problem Statement: The diesel oil purification
processona marine vessel involves passingthefuel
through a series of filters and separators to remove
impurities. The process requires the regulation of
several components. The goal of the optimization
processistofindtheoptimisedpathtocompletethe
processtoensure
efficiency.
PreliminaryAnalysis:Toconductanevaluation,a
cohortof15engineerspossessingaminimumofthree
yearsofrelevantexperiencewasselectivelyrecruited.
Theseengineerswereassignedthedesignatedtaskof
transferringDOfrom the storage tankto the service
tank in a simulator, involving steps as depicted in
Figure 3. Prior to the evaluation, the engineers had
received comprehensive training in operating the
simulator. Each engineer was instructed to perform
the task utilising their own heuristic approach,
thereby implementing a unique method deemed
optimal by the individual. It was observed that the
sequenceofstepsandtotaldistanceto
accomplishthis
taskby each engineer,varied greatly, validating our
hypothesisthatdailyoperationaltasksrelyheavilyon
humanpreferenceandarepronetoerror.Therefore,
weneedagloballystandardisedmethodtooptimise
thesecomplexoperations,therebyreducingtheerror
riskandincreasingtheprocessefficiency.
Solution Approach: We
will use ant colony
optimization to optimise the diesel oil purification
process. ACO is a powerful optimization technique
that is well‐suited to problems with multiple steps
and complex interactions between them. The
algorithmworksbysimulatingthebehaviourofants
searching for food, with the goal of finding the
shortest
pathtothefoodsource.Inourcase,wewill
useACOtosearchfortheoptimalpathway ofdaily
operational tasks to accomplish the purification
process. One such task is transferring DO from the
storage tank to the servicetank. In order to transfer
the DO, the engineer officer
on watch must transfer
fuel from the bottom tank to the Settling tank; this
transfer requires 10 tasks to be completed
successfully.ThetasksarementionedinFigure3.
Figure3.Stepsinvolvedinthetransferofdieseloilfromthe
storagetanktotheservicetank.
SolutionMethodology:ThestepsinvolvedinACO
optimization are represented in the algorithm
flowchartinFigure4.ThefirststepinapplyingACO
to the diesel oil purification process is to define the
problem parameters and objective function. The
objective function for this Ant Colony Optimization
algorithmisthefitnessfunction,
whichcalculatesthe
total distance of the tour for a given ant. In
mathematicalnotation,thiscanberepresentedas:
Fitness=∑(i=1ton‐1)dist(tour(i),tour(i+1)) (6)
where dist. is the distance function between two
nodesandtouristhesequenceofnodesvisitedbyan
ant. The
goal of the algorithm is to minimize this
fitnessfunctionbyfindingtheshortesttourpossible.
Figure4.PseudocodeforMATLABimplementationofACO
optimizationalgorithm.
293
Inthiscase,theproblemparametersaredistance,
and the objective function is the efficiency of the
purificationprocess.Next,wewillinitializetheACO
algorithm by defining the number of ants, the
pheromone trail, and the heuristic information. The
pheromonetrailrepresentsthecumulativeexperience
oftheants,while
theheuristicinformationrepresents
the domain‐specific knowledge of the problem. We
will use the local and global pheromone update
equations to update the pheromone trail as theants
search for the optimal solution. Once the ACO
algorithmisinitialised, wewillrunitforaspecified
numberofiterations,
witheachiterationrepresenting
acompletesearchcyclebytheants.Ateachiteration,
theantswillconstructasolution by selectingvalues
for each task, based on the pheromone trail and
heuristicinformation.Thequalityofthesolutionwill
beevaluatedusingtheobjectivefunction,andthebest
solutionfound
byanyantintheiterationwillbeused
to update the pheromone trail using the global
pheromoneupdateequation.TheACOalgorithmwill
continue to search for the optimal solution until a
stoppingcriterionismet,suchasamaximumnumber
ofiterationsoraconvergenceofthesolution
quality.
Figure5. ResultsofACOoptimization tofindoutthebest
possible path to complete the tasks involved in diesel oil
purification,where thexandy axispresentsthedistances
betweeneachcomponent.
Results: The application of ACO to the diesel oil
purification process on a marine vessel resulted in
significant improvements in the efficiency of the
process by optimising the best‐suited path for the
complexdailyoperationofthedieseloil purification
process as shown in Figure 5. By optimising the
distance
and prioritising the sequence of steps, the
daily operational tasks related topurification can be
performed more effectively and in less time. It is
importanttonotethatotheroptimizationtechniques,
suchasMonteCarlosimulations,canbeusedtofind
optimalsolutions.However,ACOwaschosenforthis
study
due to its effectiveness in solving complex
problemswithlargesearchspacesandfindingglobal
optima more efficiently. This is especially important
in the context of optimizing daily operations in the
marineindustry,wherethesearchspacecan be vast
andtheoptimalsolutionmaynotbeeasilyapparent.
Therefore, while
other optimization techniques may
havepotentialtobeusedinthiscontext,theselection
of ACO in this study was based on its ability to
efficiently explore the search space and find the
optimal solution. The use of ACO allowed us to
explorethesolutionspacemorethoroughlyandfind
theoptimalsolution,whichwouldhavebeendifficult
orimpossibleusingotheroptimizationtechniques.In
conclusion,theuseofantcolonyoptimizationforthe
optimization of the diesel oil purification process in
the maritime is a powerful tool that can lead to
significant improvements in the efficiency of the
process. We
can extend the use of AI to optimize
multi‐parameter processes. Furthermore, the
technique can be applied to other optimization
problems in the maritime industry, as well as other
industrieswithsimilarproblems.
5 DISCUSSIONANDCONCLUSION
Themaritimeindustryisconsideredadrivingfactor
in the global economy. The
high dependence on
human performance in the maritime industry often
leads to a high number of casualties, and decision
errors are the primary cause of maritime accidents.
Previous studies have addressed this issue by
implementing various approaches such as decision
support systems [21, 22], fuzzy logic [23, 24], and
genetic algorithms
[25, 26]. However, this study
proposes an Ant Colony Optimization (ACO) based
algorithmtooptimizethepathforcomplextasksand
provide clear and sequential instructions for daily
operationaltasks.Thisalgorithmhasbeenpreviously
applied in various fields, includinglogistics[27,28],
transportation[29,30],andcommunicationnetworks
[29,
31],andhasshownpromisingresults.Thereare
severaladvantagestousingACOfortheoptimization
ofcomplexproblemssuchasthedieseloilpurification
processinthemaritime.First,ACOisabletohandle
problems with multiple parameters and complex
interactions between them. Second, ACO is able to
explore
thesolutionspacemorethoroughlyandfind
theglobaloptimumsolution,ratherthangettingstuck
in local optima. Finally, ACO is able to adapt to
changesintheproblemenvironment.TheuseofACO
for the optimization of the diesel oil purification
process in the maritime is a powerful tool that
can
lead to significant improvementsin the efficiency of
theprocess.
In our research, we have applied the ACO
algorithm to daily operational tasks in the maritime
industry,suchastransferringfuelfromstoragetanks
toservicetanks.Theproposedsolutionprovidesaset
ofverifiedinstructionsforperformingeach
taskina
standardised manner, thus minimising personal
preference and false assumptions. The proposed
solution has numerous advantages, such as
improving efficiency, reducing the likelihood of
human error, and providing a globally accessible
standardized instructions manual. This solution can
haveasignificantimpactonthemaritimeindustryby
reducingthenumber
ofcasualtiesandimprovingthe
overall safety of operations. In conclusion, our
research article proposes an innovative approach to
optimise daily operational tasks in the maritime
industryusingAI‐basedtechniques.Thestudybuilds
on previous research and provides a comprehensive
294
discussionoftheproposedsolutionʹsadvantagesand
potential impact on the industry. Overall, this
research has the potential to contribute tothe safety
andefficiencyofthemaritimeindustry.Infuture,this
technique can be applied to other optimization
problems in the maritime industry, as well as other
industries
with similar problems, such as
transportation, logistics, and telecommunications.
Moreover, multi‐factor complex studies can be
conducted using this model to evaluate large‐scale
implementation and validate its use for process
optimization.
Table1.ACOOptimisedthesequenceofstepsinvolvedintheprocessofdieseloiltransferfromthestoragetanktothe
servicetank.
___________________________________________________________________________________________________
Process No. StepsNormal Transfer Valve Location
in‐service condition Position (X,YCoordinates)
condition
___________________________________________________________________________________________________
Transfer 1 ConfirmDOServiceTankHigh‐Levelalarmis‐‐‐(100,85)
ofdieseloperationalbyactivatingthefloatswitch
oilfrom 2 DOStorageTankOutletValveShut Open (V‐PL121) (15,5)
storage 3 DOstorage&servicetankcommonsuctionShut Shut (V‐PL123) (22,7)
to 4
DOtransferpumpinletShut Shut (V‐PL124) (37,7)
service 5 PurifierfeedpumpsuctionvalveShut Open (V‐PL211) (55,5)
tank 6 PurifierinletV/VShut Open (V‐PL212) (62,5)
7 OpenPurifierdischargetotheDOServiceTank Shut Open (V‐PL213) (73,4)
8 OpenPurifierSludgeValveShut Open (V‐PL291) (87,5)
9 OpentheOperatingWaterTankoutletvalveShut Open (V‐PG204) (87,22)
10 StarttheDOPurifiermotorandallowittorunup‐‐‐(110,95)
tospeed.StarttheDOpurifierFeedpump,select
‘START’onthePurifierControlPanel.Checkfor
anyleakageincludingbowlleakage
___________________________________________________________________________________________________
REFERENCES
[1]Allianz Global Corporate and Speciality. Safety &
Shipping Review 2019 | AGCS
https://www.agcs.allianz.com/news‐and‐
insights/news/safety‐shipping‐review‐2019.html
(accessedJan19,2023).
[2]deMaya,B.N.;Farag,Y.;Bantan,H.;Kurt,R.;Turan,O.;
Uflaz, E.; Basappa, R. D.; Sotiralis, P.; Ventikos, N. P.
Human Factors’ Contribution into Maritime
Accidents
byApplyingtheSHIELDHFTaxonomy. SNAME14th
International Marine Design Conference, IMDC 2022,
2022.https://doi.org/10.5957/IMDC‐2022‐336.
[3]Sánchez‐Beaskoetxea, J.; Basterretxea‐Iribar, I.; Sotés, I.;
Machado, M. de las M. M. Human Error in Marine
Accidents: Is the Crew Normally to Blame? Maritime
Transport Research, 2021, 2,
100016.
https://doi.org/10.1016/J.MARTRA.2021.100016.
[4]Chauvin,C.;Lardjane,S.;Morel,G.;Clostermann,J.P.;
Langard, B. Human and Organisational Factors in
MaritimeAccidents:AnalysisofCollisionsatSeaUsing
the HFACS. Accid Anal Prev, 2013, 59, 26–37.
https://doi.org/10.1016/J.AAP.2013.05.006.
[5]Hulme,A.;Stanton,N.A.;Walker,G.H.;Waterson,P.;
Salmon,P.M.
AccidentAnalysisinPractice:AReview
of Human Factors Analysis and Classification System
(HFACS) Applications inthePeerReviewedAcademic
Literature. https://doi.org/10.1177/1071181319631086,
2019, 63 (1), 1849–1853.
https://doi.org/10.1177/1071181319631086.
[6]Li,P.;Cai,Q.;Lin,W.;Chen,B.;Zhang,B.OffshoreOil
SpillResponsePracticesandEmergingChallenges.Mar
Pollut Bull, 2016, 110
(1), 6–27.
https://doi.org/10.1016/J.MARPOLBUL.2016.06.020.
[7]Luo, M.; Shin, S. Half‐Century Research Developments
in Maritime Accidents: Future Directions. Accid Anal
Prev, 2019, 123, 448–460.
https://doi.org/10.1016/J.AAP.2016.04.010.
[8]Schröder‐Hinrichs, J.‐U.; Praetorius, G.; Graziano, A.;
Kataria, A.; Baldauf, M. Introducing the Concept of
ResilienceintoMaritimeSafety.2016,176–182.
[9]Pazouki,
K.;Forbes,N.; Norman,R.A.;Woodward,M.
D. Investigation on the Impact of Human‐Automation
InteractioninMaritimeOperations.OceanEngineering,
2018, 153, 297–304.
https://doi.org/10.1016/J.OCEANENG.2018.01.103.
[10]Dorigo, M.; Gambardella, L. M. Ant Colonies for the
Travelling Salesman Problem. Biosystems, 1997, 43 (2),
73–81.https://doi.org/10.1016/S0303‐2647(97)01708‐5.
[11]Wang,J.;
Dong,L.ShipEnergy‐SavingRoutePlanning
Based on Dynamic Fuel Consumption Model.
https://doi.org/10.1117/12.2645621, 2022, 12302, 873–877.
https://doi.org/10.1117/12.2645621.
[12]Xiang, Y.; Yang, X. An ECMS for Multi‐Objective
EnergyManagementStrategy ofParallelDieselElectric
Hybrid Ship Based on Ant Colony Optimization
Algorithm.Energies2021,Vol.14,Page810,2021,14
(4),
810.https://doi.org/10.3390/EN14040810.
[13]Chen,D.Z.;Wei,C.;Jia,G.L.;Hu,Z.H.Shortest‐Path
Optimization of Ship Diesel Engine Disassembly and
Assembly Based on AND/OR Network. Complexity,
2020,2020.https://doi.org/10.1155/2020/2919615.
[14]Ma,W.;Lu,T.;Ma,D.;Wang,D.;Qu,F.ShipRouteand
Speed Multi‐Objective
Optimization Considering
Weather Conditions and Emission Control Area
Regulations.
https://doi.org/10.1080/03088839.2020.1825853, 2020, 48
(8), 1053–1068.
https://doi.org/10.1080/03088839.2020.1825853.
[15]Lazarowska,A.Ship’sTrajectoryPlanningforCollision
Avoidance at Sea Based on Ant Colony Optimization.
The Journal of Navigation, 2015, 68 (2), 291–307.
https://doi.org/10.1017/S0373463314000708.
[16]Shen, X. Research on Optimization Model of Marine
IndustryStrategic
AdjustmentunderComplexMaritime
Conditions Based on Ant Colony Algorithm. Polish
Maritime Research, 2018, S 2, 164–169.
https://doi.org/10.2478/POMR‐2018‐0088.
[17]Yang, J.; Zhuang, Y. An Improved Ant Colony
Optimization Algorithm for Solving a Complex
Combinatorial Optimization Problem. Appl Soft
Comput, 2010, 10 (2), 653–660.
https://doi.org/10.1016/J.ASOC.2009.08.040.
[18]Dong,L.;Xiao,
Q.;Jia,Y.;Fang,T.ReviewofResearch
on Intelligent Diagnosis of Oil Transfer Pump
Malfunction. Petroleum, 2022.
https://doi.org/10.1016/J.PETLM.2022.01.002.
[19]Emad,G.;Roth,W.M.ContradictionsinthePracticesof
Training for and Assessment of Competency: A Case
Study from the Maritime Domain. Education and
Training, 2008, 50 (3), 260–272.
https://doi.org/10.1108/00400910810874026/FULL/PDF.
295
[20]Irving,P.;Holloway,M.;Hook,S.;Ross,A.;Stalvies,C.
PreparingforOilSpillMonitoring.OilSpillMonitoring
Handbook,2016.
[21]Ożoga, B.; Montewka, J. Towards a Decision Support
System for Maritime Navigation on Heavily Trafficked
Basins. Ocean Engineering, 2018, 159, 88–97.
https://doi.org/10.1016/J.OCEANENG.2018.03.073.
[22]Mansouri, S. A.; Lee,
H.; Aluko, O. Multi‐Objective
Decision Support to Enhance Environmental
Sustainability in Maritime Shipping: A Review and
FutureDirections.TranspResELogistTranspRev,2015,
78,3–18.https://doi.org/10.1016/J.TRE.2015.01.012.
[23]Malyszko, M. Fuzzy Logic in Selection of Maritime
SearchandRescueUnits.AppliedSciences2022,Vol.12,
Page 21, 2021, 12
(1), 21.
https://doi.org/10.3390/APP12010021.
[24]Poornikoo,M.;Øvergård,K.I.LevelsofAutomationin
Maritime AutonomousSurfaceShips(MASS): A Fuzzy
Logic Approach. Maritime Economics and Logistics,
2022,24(2),278–301.https://doi.org/10.1057/S41278‐022‐
00215‐Z/FIGURES/9.
[25]Chen, C. H.; Khoo, L. P.; Chong, Y. T.; Yin, X. F.
Knowledge Discovery Using
Genetic Algorithm for
MaritimeSituationalAwareness.ExpertSystAppl,2014,
41 (6), 2742–2753.
https://doi.org/10.1016/J.ESWA.2013.09.042.
[26]Li, L.; Gu, Q.; Liu, L. Research on Path Planning
AlgorithmforMulti‐UavMaritimeTargetsSearchBased
on Genetic Algorithm. Proceedings of 2020 IEEE
International Conference on Information Technology,
Big Data and Artificial Intelligence,
ICIBA 2020, 2020,
840–843.
https://doi.org/10.1109/ICIBA50161.2020.9277470.
[27]Huang, K.; Hsieh, C.‐Y.; Chou, Y.‐C. An Ant‐Based
Algorithm for the Cross Docking Scheduling Problem
for Distribution Centers. International Forum on
Shipping, Ports and Airports (IFSPA) 2013: Trade,
Supply Chain Activities and Transport: Contemporary
LogisticsandMaritimeIssues,2013,522–535.
[28]Lisowski,
J. Optimization Methods in Maritime
TransportandLogistics.PolishMaritimeResearch,2018,
25(4),30–38.https://doi.org/10.2478/POMR‐2018‐0129.
[29]Azadeh,M.A.;Shoja,B.M.;Kazemian,P.;Hojati,Z.T.
A Hybrid Ant Colony‐Computer Simulation Approach
forOptimumPlanningandControlofMaritimeTraffic.
International Journal of Industrial and
Systems
Engineering, 2013, 15 (1), 69–89.
https://doi.org/10.1504/IJISE.2013.055512.
[30]Fu, B.; Song, X.; Guo, Z.; Zhang, P. An Optimization
ModelforContainerTransportationNetworkwithACO
Approach. 2007 IEEE Congress on Evolutionary
Computation, CEC 2007, 2007, 4768–4775.
https://doi.org/10.1109/CEC.2007.4425098.
[31]Lazarowska,A.SafeShipControlMethodwiththeUse
of Ant Colony Optimization.
Solid State Phenomena,
2014, 210, 234–244.
https://doi.org/10.4028/WWW.SCIENTIFIC.NET/SSP.210
.234.
