59

1 INTRODUCTION

Thetransportationamountatsea hasbeenincreasing

allovertheworldanditisexpectedthatships’size

become bigger and marine traffic increases.

Therefore,itisessentialtoevaluatesafetyofmarine

trafficfor the purpose of improvementof efficiency

and safety of marine traffic. Traditionally, visua

l or

radarobservationhasbeenconductedtoinvestigate

actualtrafficflowofships.Thentheobserveddatais

used as base data for the evaluation of the

aforementionedmarinetraffic.Ontheotherhand,it

is becoming possible to observe the actual marine

trafficmore easily andfaster than before,according

to rapid spread of Automa

tic Identification System

(AIS).Goerlandt etal. (Goerlandt 2011)presented a

method for obtaining a realistic input data for the

marine traffic simulation through analysis of AIS

data.

Besides, description of collision avoidance

behaviours of ships are indispensable to simulate a

realistic marine traffic. However, the act

ual actions

for collision avoidance depend on circumstances

where ships are sailing, i.e. ship domains which

never let other ships enter are different between in

harboursandinthecoastalseaarea(e.g.Inoue1994,

Miyake 2015). Therefore, it is important to develop

and implement an algorithm of collision avoidance

corresponding to a ta

rget traffic or target area into

themarinetrafficsimulation.

The authors developed an automated marine

trafficsimulationsystemwithAISdata.Inthispaper,

we propose a series of systematic procedures for

marine traffic simulation including analysing for

collisionavoidancebehavioursusingAISdata.

Firstlytheprocedureforma

rinetrafficsimulation

withAISdataareintroduced.Andtheprocedurefor

analysing collision avoidance behaviours using AIS

data is introduced. Specifically, we proposed a

methodforidentifyingatimewhenagive‐wayship

evadeastand‐onship.Finally,theexaminationofthe

methodisshown.

Procedure for Marine Traffic Simulation with AIS

Data

R.Miyake&J.Fukuto

NationalMaritimeResearchInstitute,Tokyo,Japan

K.Hasegawa

OsakaUniversity,Osaka,Japan

ABSTRACT:Itisessentialtoevaluatesafetyofmarinetrafficfortheimprovementofefficiencyandsafetyof

marine traffic. Spreadof AISmakes observation of actual marine traffic more easily and faster than before.

Besides,descriptionofcollisionavoidancebehavioursofshipsareindispensabletosimulatearealist

icmarine

traffic.Itisimportanttodevelopandimplementanalgorithmofcollisionavoidancecorrespondingtoatarget

trafficortargetareaintothemarinetrafficsimulationbecauseactualactionsforcollisionavoidancedependon

circumstanceswhereshipsaresailing.Theauthorsdevelopedanautomatedmarinetrafficsimulationsystem

withAISdata.Andinthi

spaper,weproposedaseriesofsystematicproceduresformarinetrafficsimulation

includinganalysingforcollisionavoidancebehavioursusingAISdata.

http://www.transnav.eu

the International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 9

Number 1

March 2015

DOI:10.12716/1001.09.01.07

60

2 MARINETRAFFICSIMULATION

2.1 Generalstepsofmarinetrafficsimulationand

requireddata

General procedures of marine traffic simulation are

composed of the following steps: (e.g. Nakamura

2011)

Step1:observationofactualmarinetraffic;

Step2:descriptionoftheobservedmarinetraffic;

Step3:simulationoftheobserved

marinetraffic;and

Step4:evaluationofthesimulatedmarinetraffic.

Atthefirststep,dataofactualmarinetrafficina

targeted area is acquired through an investigation

based on visual or radar observation, or video

recording (e.g. The Japan Association of Marine

Safety1991).Atthesecondstep,

theacquireddatais

analysedtogetattributesofshipsandthetrafficsuch

asshiptypeandsize.Thentrafficvolumeorgeneral

trafficrouteincludingwaypointsareobtained.Atthe

third step, a series of traffic flow data, which

represents traffic conditions of target area, is

generated using these

 attributes. Then the target

trafficissimulatedbasedonthetrafficflowdata.In

this step, it is possible to include the future

prediction or new traffic systems such as traffic

separation schemes according to the purpose of

simulation. At the last step, the results of the

simulation are evaluated

with appropriate means

corresponding to the purposes. For instance,

frequency or position of one to one ship encounter

situationsisexaminedforthepurposeofassessment

oftheeffectofmodificationofashapeofchannel.

2.2 RequireddataobtainedfromAIS

Theauthorshaveproposedmodifiedproceduresfor

marine traffic

 simulation using AIS data. We

developed tools for obtaining some attributes from

AISdataandanautomateddatagenerationsystems

for these attributes. Table 1 shows analysable

attributes by traditional procedure with radar

observationandbytheproposedprocedurewithAIS

data, which are the required data for the marine

traffic simulation. The attributes denoted by “yes”

and “no” are available and not available,

respectively.Theattributes denoted by “hard”have

seldom been obtained from visual or radar

observationbecauseitrequires unrealisticamountof

worktoanalysethembyhand.Forthisreason,many

researchers have focused on whole marine

 traffic

ratherthanthemovementofindividualshipsintheir

analyses.

The procedure with AIS data makes analyses

easier,fasterandmore accurate thanthe traditional

procedure.This procedure alsomakes itpossible to

analysethedetailof the individualshipmovement,

in addition to the analyses of whole traffic

flow.

Here, it should be noted that AIS data does not

includetheinformationofsmallshipswhereAISare

notinstalled,suchasdomesticshipsoflessthan500

GTandinternationalshipsoflessthan300GT(IMO

2003).



Table1.Comparisonofanalysableattributeswhicharethe

requireddataforthemarinetrafficsimulation

_______________________________________________

ObjectRequiredAvailability

___________________

datawithRadar withAIS

_______________________________________________

ODdata(gate)  hard  yes

Individual ODdata(port)  hard  no

shipShip’sspeedhard  yes

movement Waypointshard  no

CA*behaviours  hard  yes

_______________________________________________

Shipsizeandtype yes  yes

Ship’sspeedyes  yes

Distribution Trajectoriesyes  yes

oftraffic  Trafficvolume  yes  yes

flowTrafficrouteyes  yes

ODtableyes  yes

_______________________________________________

*collisionavoidance

3 MODIFIEDPROCEDUREFORMARINE

TRAFFICSIMULATIONWITHAISDATA

Figure1showstheproposedmodifiedprocedureof

marine traffic simulation with AIS data. In this

figure, rectangles with grey shading indicate the

unique process  by using AIS data.  Rectangles

withoutshadingindicatethecommonprocess,which

arealso included in

the traditional procedures with

visual or radar observation. Rectangles drawn by

heavylineindicatetheautomatedprocesses.

Here, we briefly describe the outline of the

procedure with AIS data for marine traffic

simulation.

3.1 Step1‐GatesettingforODsurvey

Preparation of an origin and a destination survey

(OD

survey) in a targeted area  is conducted at

Process 1‐1. Gates, which represent a starting point

andanendpointoftrafficflowaswellasanorigin

and a destination of a ship, should be set at

appropriate places, considering the observed traffic

flowofthetargetedareain

ordertocoverallarterial

trafficroutes.

3.2 Step2‐ODsurveyandmodelling

At the next step, OD survey and modelling for the

trafficsimulationareconducted.AtProcess2‐1,OD

survey is conducted based on the set gates. At

Process2‐2,eachdataofeachship

whichpassedthe

gatebasedontheODsurveyisextractedasaseries

ofpositiondata,i.e.trajectory.Thenthetrajectories,

which enable to reproduce the target traffic for

marinetrafficsimulation,areclassifiedaccording to

the routes defined by origin and destination gates.

Irrelevant data for the purpose of

 analysis such as

anchoredshipsisfilteredoutinthisprocess.



61

AIS data

Decode

Extract target area

Gate setting for OD survey

Extract ships passed gates based on OD survey

Analysis of ship type and size

Category of ship grouped by ship type and size

Modelling of traffic route

Navigation data

Evaluation

Extract ships passing gate based on OD survey

Synchronized data of state of individual ship

Time series data of state quantity of encountered two ships

Analysis of collision avoidance behaviours

Development of algorithm for collision avoidance

Modelling of

occurrence frequency

of ships

number of ship

OD table

Ship speed

Traffic route

Step 1

Step 3

Step 4

Step 2

Process 2-2

Collision avoidance behaviours

Process 2-3

Process 2-4

Process 1-1

Process

2-5

Process 2-6

Process 5-2

Process 5-3

Process 5-4

Process 5-5

Analysis object

Whole traffic flow

OD survey

Process 2-1

OD survey

Marine traffic simulation

algorithm for collision

avoidance

Step 5

Process 5-1

Process 5-6



Figure1.ModifiedprocedureofmarinetrafficsimulationwithAISdata

Figure 2 shows the trajectories of all extracted

shipsthatpassedtheorigingateandthedestination

gate. Based on attribute data of the extracted ships

from AIS data, at Processes 2‐3 and 2‐4, ships are

grouped according to ships’ types and sizes. Then

categories of ships are created

 for the model. They

shouldbeclassifiedaccordingtotheappropriateship

size and type because the attributes of traffic route

dependonshipsizeandtype.

Through Processes 2‐5 and 2‐6, modelled data

such as traffic routes and occurrence frequency of

ships in the simulation is generated

using the

developed tools. Specifically, at Process 2‐5, the

numbers of ships in the respective categories are

counted,andthenODtableisgenerated.AtProcess

2‐6, average speed of ships or waypoints of traffic

flow are observed andthen a traffic routemodel is

generated for respective categories

 of ships’ types

andsizes.