185
1 INTRODUCTION
Theaimofthispaper istoaddress the feasibility of
obtaining maritime awareness, i.e. continuous
knowledge of the whereabouts of ships, over sea
basin‐wide areas, by collecting and integrating data
fromvariousshipreportingsystems;andtoassessthe
qualityoftheresultingmaritimepicture.
Ship
reportingsystemsincludeLRIT(LongRange
Identification and Tracking; IMO 2008, Popa 2011),
AIS(AutomaticIdentificationSystem;ITU2010)and
VMS (Vessel MonitoringSystem, for fishing vessels;
FAO 1998), plus non‐automatic reporting through
radio call‐in. These have been set up for different
purposes and work in different ways,
but in all of
them, the ships send out short messages reporting
theiridentityandposition,andsometimesadditional
information. Thanks to the use of satellites as
communication platforms, these messages may now
be received from ships anywhere on the globe,
allowinganunprecedentedviewofshiptraffic.
For many applications,
there is a requirement to
have an up‐to‐date awareness of where, within a
certainarea ofinterest,alltheshipsare,whotheyare,
andwheretheyaregoing:i.e.toknowthe“Maritime
Situational Picture”. Such a requirement stems in
particular from authorities who are responsible for
maritime safety and security. A particular case is
counter‐piracy. The wide‐area Maritime Situational
Picture(MSP)isneededinpiracy‐affectedseasforthe
authorities to assess emerging risks, as merchant or
fishingshipsapproachlocationswherepiracyactivity
hasbeen reported, and to issuewarningsdirectlyto
ships at
risk. It is also needed to support the
deployment of inspection or interception assets.
Furthermore,fromthehistoriccollection of MSPs as
they evolve over time, ship traffic patterns can be
compiled, that allow the assessment of geographical
Basin-Wide Maritime Awareness From Multi-Source
Ship Reporting Data
H. Greidanus, M. Alvarez, T. Eriksen, P. Argentieri, T. Cokacar, A. Pesaresi, S. Falchetti,
D.Nappo,F.Mazzarella&A.Alessandrini
EuropeanCommission–JointResearchCentre,Ispra,Italy
ABSTRACT:Asystemwassetuptoingestautomaticshippositionreports(terrestrialandsatelliteAIS,LRIT)
andfusetheseintoaMaritimeSituationalPicture,trackingtheshipswithinanoceanbasin‐wideareaofinterest
inrealtime.Trialrunsweremade
overseveralmonths,collectingreportingdatafromanumberofdifferent
sources,overtheGulfofAdenandtheWesternIndianOcean. Alsosatelliteradarsurveillancewascarriedout
inordertosamplethepresenceofnon‐reporting ships.ThetrialshowedthatsatelliteAISisapowerfultoolfor
basin‐wide ship traffic monitoring; that multiple AIS satellites are needed for sufficient completeness and
updaterate;andthatcoastalAISandLRITstillprovideessentialcomplementstothesatelliteAISdata.The
radarsurveyshowedthatabouthalfoftheradar‐detectedshipsarenotseeninthereportingdata.
Theultimate
purposeofthisworkistosupportthecountriesaroundtheHornofAfricainthefightagainstpiracyandto
helpbuildtheircapacitytodelivermaritimesecurityandsafety.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 7
Number 2
June 2013
DOI:10.12716/1001.07.02.04
186
riskdistributionandthatcanfunctionasareference
enabling the recognition of abnormal behaviour that
mightindicateaproblem.
With these applications in mind, the PMAR
(Piracy, Maritime Awareness and Risks) study was
performedtoassesshowmaritimeauthoritiesaround
theHornofAfrica canacquire thelevelof
maritime
awareness needed to carry out counter‐piracy
responsibilities. Piracy off the Horn of Africa is a
regionalproblem(Duda&Wardin2012).Inorderto
facilitate a regional approach, and to permit data
sharingbetweendifferentcountries and government
sectors, the data and systems used should be
unclassified. For
use within Africa, the technologies
should match available infrastructural limits; and
they should be cost‐effective. As such, the PMAR
project was aimed at Regional Maritime Capacity
Building, and was part of an internationally
coordinated effort from the European Union to
combat piracy and increase maritime security off
Africa(seealso
Perkovicetal.2012).
Thispaperdiscusseshowacontinuous,real‐time
MSP may be maintained by maritime authorities in
Africa, derived from integrating the data from a
numberofshipreportingsystems.TheMSPservesin
the first place for counter‐piracy purposes, but also
formaritimesecurity, safety
andresource protection
purposes.Thepaperwilldiscusstheperformanceof
theMSPintermsof(a)thenumberofdifferentships
that are detected; (b) how often ship positions are
updated, determining how well the ships can be
tracked over time; (c) the marginal benefit of
additional data sources (e.g.
how many ships are
detected using one reporting data source, two
reporting data sources, etc.); (d) particular problems
with the data and their impacts; and (e) the
completenessoftheMSP,bysamplingnon‐reporting
shipswithsatelliteradar.
2 METHOD
First, an IT architecture was designed and
implemented(Fig.
7)which:(a)Continuouslyingests
incoming data streams from several ship reporting
systems. (b) Tracks each ship based on its MMSI
number (the main identifier in the AIS messages).
Sometimes(inLRIT),theshipisidentifiedbyitsIMO
number;inthatcase,ashipregisterisusedtoconvert
this
toanMMSInumber.(c)Predictsthe positionof
each ship to a certain reference time, based on the
ship’slastreportedpositionandspeedwhichmaybe
some hours old. If no speed is reported in the
message, then it is computed from the two most
recentpositions.The
referencetimeisthesameforall
ships,atorjustaheadofthecurrenttime,andinthis
way,areal‐timeMSPiscreated;thisstepisupdated
atfixedintervals,e.g.every15minutes.(d)Displays
the resulting MSP on a screen, whereby the ship
positions are clickable
to display information about
the ship and its past track. (e) Can sum all MSPs
computed over a certain time period to obtain ship
traffic density maps. The IT system furthermore: (f)
Ingestspositions of ships thathavebeendetected in
satelliteimages(“VDS”for Vessel Detection System;
Greidanus &
Kourti 2006). In contrast to the
continuous stream of positions from the ship
reporting systems, the VDS positions are only
availablewhenasatelliteimageistakenoverthearea,
andtheVDSshipsareunidentified.(g)Correlatesthe
VDSpositionswiththepositionsoftheknownships
(fromthereportingsystems),sothatadistinctioncan
be made between VDS ships that were already
known,andnon‐reportingVDSships.Finally,theIT
system can (h) Ingest piracy incident data (location,
time, incident description) and can plot these on a
map together with the MSP or historic ship density
maps
forriskassessmentpurposes.
With this system, two test campaigns were
executed;thefirsttohelpdesignthesystem,andthe
secondtotuneitandmeasuretheperformance.Data
were mainly collected from AIS, LRIT and satellite‐
borneSyntheticApertureRadar(SAR).AISisglobally
mandatedonSOLASvessels(mainly
shipsof300GT
andmore)anditsmessagesarebroadcastedonVHF
with high update rate (at intervals of seconds to
minutes). The AIS messages, which contain a lot of
information,canbereceivedbycoastalreceiversorby
dedicatedsatellitespassingoverhead.Fortheir usein
shiptracking
seee.g.B.J.Tetreault2005andCarthelet
al. 2007. LRIT is also globally mandated on SOLAS
vessels,butthemessagesaresentbysatcomdirectly
to the ship’s Flag State usually at 6‐hourly intervals
andcontainmuchlessdatathantheAISmessage.As
aSARsatellitepassesoverhead,
itcanmakeasnap‐
shotradarimageoftheseasurfaceofanextentofup
to several hundred kilometres on the side, enabling
the detection (but not identification) of the larger
ships (> 20 m). It is also possible to make more
detailedSARimagesthatcandetect
boatsassmallas
a meter in favourable conditions, but such images
onlyhaveaverylimitedextent(5‐10kmontheside)
andarethereforenotsuitableforsurveyingextended
areas.
Figure1.Trialarea(viewedonGoogleEarth).
The data were obtained from many providers,
both commercial and institutional. Table 1 gives an
overview of the data sources that were used in the
secondtestcampaign, on which theresults reported
here are based. (Some results of the first test
campaign were reported in Posada et al. 2011.)
References
to the AIS data sources are: Wychorski
(2010),Eriksenet al. (2010), Eiden (2010), Flessate &
187
Loretta (2010), Lorenzini (2010) and Martin & Allen
(2010). The AIS and LRIT data were continuously
collectedwithintheboxshowninFigure1duringthe
period 1 Aug 2011 – 31 Jan2012 (however not each
datasourcewasavailableduringthatentireperiod).
The SAR images were collected
during a limited
numberofdaysintheperiodOct–Dec2011,mostly
concentratedinthe Gulf of Aden and off Mombasa,
DarAsSalaamandtheSeychelles.Auxiliarydatathat
were used included publicly available ship registers
anddigitalmapdata.Furtherdatathatwerecollected
included VMS
reports, and optical images from the
ALOS‐PRISM, SPOT, IKONOS and DEIMOS
satellites,butthosewereonlyusedtoalimitedextent
andarenotdiscussedhere.
The tracking is designed to be run in real‐time.
BecausesatelliteAISdataonlybecomeavailablewith
adelayoftheorderof
hours(duetothetimeneeded
for downlinking and processing), prediction of the
ships’ current positions to the “now” is essential.
Furthermore, all ingested data are stored in a
database, and the tracking can be re‐run off‐line as
well.Whendoingso,itispossibletochoosea
sub‐set
of all available data sources. In this way, one can
explorethe impact of different combinationsof data
sources on the completeness and quality of the
resulting MSP. This is important to know, because
each data source has an associated cost, and one
wants to use the lowest‐cost
combination of data
sources that still provide the necessary level of
quality.
Table1. Main data sources used in the second trial. The
period15Nov–15Dec2011(31days)isusedasareference
periodfortheresultsinthispaper.
Datatype Provider C/I* #Platformsin
15Nov– 15Dec
Terrestrial
AIS
MSSIS I Setofcoastal
receiversplusa
fewreceiveson
mobile
p
latforms
Satellite
AIS
FFI I 2satellites:
NORAIS,AISSat‐1
LuxSpace/
Orbcomm
C 2satellites
exactEarth C 2satellites
LRIT EUFlagStates
/EMSA
I 17Flags
Satellite
SAR
TerraSAR‐X C 2satellites
Radarsat‐2 C 1satellite
CosmoSk
y
Med C 4satellites
Envisat‐ASAR I 1satellite
*:Commercial/Institutional
3 RESULTS
Resultswillbegivenherefortheone‐monthreference
period 15 Nov – 15 Dec 2011 (31 days), when the
systemwasperformingwellandmanydatastreams
wereonlineatthesametime.Duringthisperiod,on
average 49,000 messages (AIS, LRIT)
per day were
received(fromtheareaofFig.1).
Figure 8 shows an example of an MSP. Such a
pictureistypicallyrefreshedevery15minutes.
AscanbeseenfromTable1,therewere7different
AIS platforms available, the terrestrial MSSIS
(countedhereasoneplatform)and
6satellites(from3
providers). Figure 2 plots, as an example, all AIS
position reports received during one day, from one
singlesatellite(top),andfromall7platformstogether
(bottom).Itisimmediatelyapparentthatinorder to
domeaningfultrackingofshipsovertheopenocean,
morethanone
AISsatelliteisneeded.
InTable2,itcanbeseenthatwithMSSISonly,on
average110differentshipsareseenatanygiventime,
whereas with one satellite AIS platform, on average
607 are seen. This is because the coverage of the
satellite is much wider than that
of MSSIS which is
mostlycoastalandwithalimitednumber ofcoastal
stationsinthisarea.Whencombiningthedataofall
systems (MSSIS, LRIT and 6 AIS satellites, bottom
line),1011differentshipsareseenatanygiventime
onaverage.Thisisstillabitmorethanthe
931(one
line before last) from combining MSSIS, LRIT and
only4outofthe6AISplatforms.Sonotonlydoesthe
qualityof the tracks increase as more AIS pla tforms
are used, as illustrated in Figure 2, but also the
numberofshipsthatcanbetracked.
Figure2.AllAISpositionreportsreceivedduringoneday;
top: from one single satellite; bottom: from 7 different
platformstogether(6satellitesandMSSIS).
188
Although MMSI numbers are supposed to be
unique, sometimes the same MMSI is used by
different ships. This happens in particular with
invalid MMSI numbers such as 0, 1, 123456789, etc.
Thetrackingalgorithmhastoresolvesuchsituations.
(Some ship tracki ng and prediction issues were
discussed in Falchetti et
al. 2012.) Whereas the total
number of distinct MMSIs during the month was
found to be 5155 (Table 2), in fact a total of 5235
distinct tracks (actually distinct ships) could be
recognised. The right column “# Ships” in Table 2
indeedreferstotrackedships,nottoMMSInumbers.
Figure3depictsinadifferentwaythecumulative
value of adding more AIS data sources. It plots the
numberofdifferentMMSIsseenperdayasafunction
of using one AIS provider, 2, 3 and 4 (the four
providersfromTable1).(Notethatthefinalnumber
of just
under 1200 MMSIs in the top plot is higher
than the final 1011 from Table 2, because during a
wholedaymoreshipsareseenthanatoneinstant.)
Table2.ThenumberofdifferentMMSIsthatareseeninthe
areaduringtheentiremonth, andthenumberofdifferent
ships that are present at any one moment (on average);
given for MSSIS only (top line), for a single AIS satellite
(averaged over all 6 satellites, 2
nd
line), when combining
MSSIS, LRIT and four satellites from two providers
(averaged over the three possible combinations of two
providers, 3
rd
line), and when taking all systems together
(bottomline).
Systems #MMSIin
wholemonth
#Shipsatone
time(mean)
MSSIS 1851 110
SingleSat‐AIS(mean) 4363 607
FourSat‐AIS+LRIT+
MSSIS(mean)
5022 931
All 5155 1011
Inordertobeabletotracktheships,andpredict
theirpositionsatthetimeoftheMSP,theirpositions
must be updated frequently. Whereas coastal AIS is
receivedcontinuously(aslongastheshipisinrange),
satelliteAISisonlyrefreshedwhenasatellitepasses
over.The
orbitperiodofasatellitecanbeoftheorder
of 100 minutes, so a single satellite can provide
updatedpositionsonceagainafterthattime,butthen
theearthrotationcausestheareaofinteresttorevolve
outofthesatellite’sswath,sothatthenextupdateis
thenonly
aftermanyhours.Theuseofmorethanone
satelliteisneededtoobtainmorefrequentupdates.In
addition,theAIStransmittersfittedontheshipswere
not originally designed for satellite reception, and
only a certain fraction of all emitted messages are
actually received. This fraction depends on the
individualship,aswellasonthedensityoftheship
traffic.Theresultisthatforsomeships,manyupdates
maybereceived,whileforothersonlyfew.Figure4
showsthiseffect.Itplotsadistributionofthenumber
of ships as a function of how many times their
positionisupdated.Thehorizontalscale,thenumber
of messages received from the ship, is logarithmic.
Thepeakontheextremeleftaretheshipsfromwhich
only a single message is ever received (during the
one‐month period used here); there are 310 such
ships. Most ships are seen
with between 50 – 500
messages (during the month). From some ships
(extremeright),nearly10,000messagesare received.
Figure 6 shows message update intervals per ship,
againcompiledovertheone‐monthperiodandona
logarithmic horizontal time scale. The top graph is
using a single AIS satellite only.
The left peak with
updatetimesshorterthan10minutes,isfromupdates
received within one single satellite overpass. An
isolated peak is seen at around 100 minutes,
corresponding to the refreshafterone satellite orbit.
Thenfollow updatesafterlonger thansome9hours
whenthe earth has revolved
the area againinto the
satellite swath. The bottom graph is made from
combining all systems; the two gaps between 10
minutes and 9 hours are now filled, enabling much
bettertracking.
Figure3. How the number of different MMSIs found per
day, averaged over one month, increases as more AIS
providers(MSSIS,LuxSpace/Orbcomm,exactEarthandFFI)
areadded.Thetwocoloursrelatetotwoorders of adding
theproviders,redstartingwithMSSIS,greenstartingwith
oneofthesatelliteproviders.For
twodifferentperiodsof1
month.
Figure2(bottom)showsoneshippositiononland;
this is an error in the AIS message content. From
analysingthenumberofreportsonland(takinginto
accountabufferzonetocorrectforshipsinports),it
is estimated that of the order of 1 in 10,000 AIS
messages
maycontainanerror.Theerrorratevaries
noticeablywithplatform,indicatingthatsomeerrors
189
happen on receive. However, in some instances the
sameerroneousmessageisreceivedbytwodifferent
platforms, proving that errors can also occur on
transmit.ItisthoughtthatsomeoftheuniqueMMSIs
that are seen only once (from the 310 mentioned in
Fig.4)arealsotheresult
ofmessageerrors.
Figure4. This distribution shows how many ships occur
(vertical axis) for each number of messages per ship
received during one month (horizontal axis, logarithmic
scale).
Figure5. Histogram of the difference between the AIS
messageinternaltimestamp(seconds)andthesecondspart
of the externally affixed time stamp. All platforms are
combinedinthis graph,fromonemonthdata,eventhough
theyindividuallyshowquitedifferentbehaviours.
The AIS message itself does not specify the time
that it was broadcasted, but it does contain the
seconds part of that time. It is up to the receiver to
affix the full time stamp. A comparison was made
betweenthetimestamp(secondsonly)insidetheAIS
message, and the
seconds part of the externally
affixed time stamp. Significant differences were
found,varyingsystematicallywithAISplatformand,
perplatform,asafunctionoftime.Figure5displaysa
histogram of the differences between the two time
stamps, lumping together all platforms; ideally, this
shouldshowapeakaround0.
Alsointhetrackingit
was found that ship positions jumped between
messages received from different platforms. These
jumps indicate time errors of up to several minutes.
These time errors, due to inaccurate receiver clocks,
turn out to be rather problematic when estimating
speed,sometimesleadingtoseriouspredictionerrors.
Figure6.Histogramoftimeintervalsbetweenmessagesof
thesameship,forallshipstogetherduringonemonth.The
horizontalaxisisthe timeinterval(logarithmic scale). The
verticalaxisisthe numberoftime intervalsthat occurred.
Top:fromonesatelliteAISplatformonly.Bottom:fromall
platformstogether.
Concerning LRIT, there are 577 ships of which
LRITdataarereceived(againreferringtothe1 ‐month
period15Nov–15Dec2011),anumberthatcanbe
contrastedwiththetotalof5155ofTable2.Ofcourse
thenumberismuchlowerbecauseonlyEUFlagged
vessels
are considered. However, from these 577
ships,37wereseenexclusivelywithLRITandnoton
AIS.Onereasoncould be that theydidnot transmit
onAIS,whichwasatthattimetherecommendedbest
managementpracticewhensailingthroughhigh‐risk
areas. Alternatively, their AIS transmissions could
190
have been too weak to be picked up by the AIS
satellite receivers. In any case, this underlines the
valueofhavingLRITinadditiontoAIS.
Asforthenon‐reportingships,Figure9showsas
reddotsthenon‐reportingshipsthatwerefoundina
satelliteSAR
surveyfrom22Novto4Dec2011over
the Gulf of Aden. The linear concentration running
downtheGulfofAdenrepresentsthetransitcorridor
thatwassetupforbetterpiracyprotection.About55
% of the total number of ships detected in the SAR
imageswerenon‐
reporting.Thiscanbebecausethese
ships are not subject to AIS or LRIT carriage
requirements(mainlybecausetheywouldbesmaller
than300GT),orbecausetheyarenotreportingeven
thoughtheyshould,orbecausetheirAISsignalswere
notsuccessfullyreceived.
Figure7.Datafusionarchitectureasdescribedinthetext.
Figure8.DisplayofMaritimeSituationalPicture,for16Nov2011,07:01UTC.(Shipdetailsanonymised.)
191
Figure9.SatelliteSARimagestakenintheGulfofAdenintheperiod22Novto4Dec2011.Theblueboxesaretheoutlines
oftheimages; thedotsareshipdetections.Greendotsareshipsthatcouldbecorrelatedwithreportingships.Reddotsare
non‐reportingships.
ThesatellitesusedarementionedinTable1.
4 CONCLUSIONS
ByusingAISdata receivedfrom satellites,reporting
ships can be tracked and a Maritime Situational
Picture can be maintained across an entire ocean
basin. A single AIS satellite does not provide
sufficientcompletenessandupdaterate,sodatafrom
severalareneeded.Asmoresatellitesareadded,
the
number of different ships that are found keeps
increasing,butafter about5‐6 satellites, the increase
levelsoff.TimestampsonAISmessages,whichmust
beexternallyaffixedbythereceiver,mustbeaccurate
to seconds ideally, to prevent significant errors in
tracking and prediction using multi‐platform data;
this accuracy is not yet available. Although LRIT
reportslessfrequentlythanAIS,someshipsareonly
foundinLRIT,soitsadditionisvaluable.Additionof
coastalAISfurtherimprovesthecompletenessofthe
picture,aswellasthereal‐timetrackingperformance
incoastalareaswherereactiontimes
shouldbefaster
than far away from the coast. Observations with
satellite SAR show that about half of the SAR‐
detected ships do not report. Future work should
further analyse the SAR detections to establish to
what extent this is because they are too small to
report.
ACKNOWLEDGMENTS
FFI (Norwegian Defence
Research Establishment)
provided NORAIS and AISSat‐1 data. The EU
MemberStatesauthoritiesandEMSAprovidedLRIT
data. The Volpe Centre of the U.S. Department of
Transport provided MSSIS data. ESA provided
Envisat‐ASAR data. DLR provided ship detections
from TerraSAR‐X data. The Spanish FMC provided
VMSdata.Commercial
providersincludedLuxSpace,
exactEarth, KSAT, MDA, e‐Geos and InfoTerra. The
projectbenefittedmuchfromcooperationwithEUSC,
ItalianCoastGuard,SPAWAR,NAVAF,NURCand
NRL.
REFERENCES
Carthel,C.,Coraluppi,S.&Grasso,R.2007.FusionofAIS,
RADARandSARDataforMaritimeSurveillance.Proc.
SPIEEuropeRemoteSensingConference,Florence,Italy, Sep
2007.
Duda,D.&Wardin,K.2012.InfluenceofPirates’Activities
on Maritime Transport in the Gulf of Aden Region.
TransNav‐InternationalJournal
onMarineNavigationand
SafetyofSeaTransportationVol.6No.1:109‐115.
Eiden,G.2010.LuxSpaceSAT‐AISExperienceafter3Years
of Operation. European Symposium on Satellite AIS,
Bremen,6‐7Dec2010.
Eriksen,T.,Skauen,A.N.,Narheim,B.,Helleren,O.,Olsen,
Ø.&Olsen,R.B.2010.
TrackingShipTrafficwithSpace‐
Based AIS: Experience Gained in First Months of
Operations. Proc. Waterside Security Conference (WSS)
International, 3‐5 Nov. 2010: 1‐8. doi:
10.1109/WSSC.2010.5730241
Falchetti, S., Alvarez, M., Cokacar, T., Greidanus, H. &
Vespe,M.2012.ImprovingCooperativeVesselTracking
for a Maritime Integrated Surveillance Platform.
Proc.
NATO RTO SCI‐247 Symposium on “Port and Regional
MaritimeSecurity”,Lerici,Italy,21‐23May2012.
192
FAO (Food and Agriculture Organization) 1998. Fishing
Operations.1.Vessel MonitoringSystems.FAOTechnical
GuidelinesforResponsibleFisheriesNo.1Suppl.1.Rome,
FAO.ISSN1020‐5292.
Flessate, G. & Loretta, A. 2010. Orbcomm– Worldwide
AISDatafromSpace.TEXAS‐IVConference,Washington
DC,28‐30Sep2010.
Greidanus,
H.&Kourti,N.2006.FindingsoftheDECLIMS
project– Detection andClassification of Marine Traffic
from Space. Proc. SEASAR 2006: Advances in SAR
Oceanography from ENVISAT and ERS Missions, Frascati,
Italy,23‐26Jan2006;ESASP‐613.
IMO(InternationalMaritimeOrganisation)2008.Resolution
MSC.263(84), Revised Performance Standards
and
FunctionalRequirementsforLRIT.
ITU (International Telecommunications Union) 2010.
Recommendation ITU‐R M.1371‐4 (04/2010), Technical
characteristics for an automatic identification system
usingtime‐divisionmultipleaccessintheVHFmaritime
mobileband.
Lorenzini, D. 2010. SpaceQuest. CANEUS International
Workshop on Collective Security, Safety and Prosperity,
MarinadiCarrara,Italy,
20‐22Oct2010.
Martin.,D.& Allen, J.2010.ExactEarth – exactAISService
and Capabilities. TEXAS‐IV Conference, Washington DC,
28‐30Sep2010.
Perkovic,M.,Wardin,K.,Twrdy,E.,Harsch,R.,Gucma,M.
&Vidmar,P.2012.TechnologicalAdvancesandEfforts
to Reduce Piracy. TransNav‐International Journal
on
MarineNavigation andSafety of SeaTransportationVol.6
No.2:203‐206.
Popa, L. V. 2011. Ships Monitoring System. TransNav‐
InternationalJournalonMarineNavigationandSafetyofSea
TransportationVol.5No.4:549‐554.
Posada,M.,Greidanus,H.,Alvarez,M.,Vespe,M.,Cokacar,
T.& Falchetti,S.
2011.MaritimeAwarenessforCounter‐
PiracyintheGulfofAden.Proc.IEEE2011International
Geoscience & Remote Sensing Symposium (IGARSS 2011),
Vancouver,Canada,24‐29July2011:249‐252.
Tetreault, B. 2005. Use of the Automatic Identification
Systems (AIS) for Maritime Domain Awareness. Proc.
OCEAN 2005, Washington DC,
USA, Sep 2005 Vol. 2:
1590‐1594.
Wychorski, H. 2010. MSSIS – Maritime Safety & Security
Information System. TEXAS‐IV Conference, Washington
DC,28‐30Sep2010.
