Current paradigms in intelligent transportation systems

Published in IET In elligen T anspo Sys ems Recei ed on 11 h No embe 2009 Re ised on 12 h May 2010 doi: 10.1049/ie -i s.2009.0102 ISSN 1751-956X Cu en pa adigms in in elligen anspo a ion sys ems S.L. To al 1 M.R. Ma ı ´nez To es 2 F.J. Ba e o 1 M.R. A ahal 1 1 E.S. Ingenie os, Uni e si y o Se ille, A da. Camino de los Descub imien os s/n, Se ille 41092, Spain 2 E.U.E. Emp esa iales, Uni e si y o Se ille, A da. San F ancisco Ja ie s/n, Se ille 41018, Spain E-mail: [email p o ec ed] Abs ac : In elligen anspo a ion sys ems (ITS) cons i u e oday a mul idisciplina y field o s udy in ol ing a la ge numbe o di e en esea ch a eas. As a consequence, i is di ficul o achie e a s uc u ed iew o ITS, which is necessa y o uni y e o s and as guidance o u u e de elopmen s. This s udy aims o iden i y he main pa adigms in he field o ITS by seman ically analysing s udies ela ed o his gene al opic. An unde s anding abou which esea ch is conside ed aluable by he esea ch communi y o build upon may p o ide aluable insigh s in his field. As a esul o he s a is ical ea men o da a, up o 13 pa adigms a e ob ained. The scope o hese pa adigms and he ela ionships be ween hem ha e also been de ailed, p o iding a s uc u ed ision o ITS syn hesised in a map o m. 1 In oduc ion In elligen anspo a ion sys ems (ITS) ha e been in es iga ed o many yea s in Eu ope, No h Ame ica and Japan, wi h he aim o imp o ing he sa e y and e ficiency o oad anspo and en i onmen al conse a ion. To his end, new echnologies and compu e powe ha e been applied o eeway, a fic and ansi sys ems [1, 2]. ITS can be conside ed a global phenomenon, a ac ing wo ldwide in e es om anspo a ion p o essionals, he au omo i e indus y and poli ical decision make s [3]. ITS in ol es a la ge numbe o esea ch a eas sp ead o e many di e en echnological sec o s such as elec onics, con ol, communica ions, sensing, obo ics, signal p ocessing and in o ma ion sys ems [4, 5]. This mul idisciplina y na u e inc eases he p oblem’s complexi y because i equi es knowledge ans e and coope a ion among di e en esea ch a eas [6]. One o he main p oblems o being a complex and mul idisciplina y field is he di ficul y o dealing wi h ITS as a de elopmen and esea ch a ea [7, 8]. T adi ional opics o in e es a e changing and new ones a e eme ging due o he con inuous ad ance o eme gen echnologies and he economical, social and en i onmen al implica ions o ITS. The complexi y o he opic sugges s ha i will be o benefi o define a global, s uc u ed iew. This iew will be use ul o suppo close in eg a ion o ITS wi h con en ional anspo a ion ini ia i es as well as o p o ide guidance o u u e ITS deploymen s. Sha ing a common s uc u ed iew will also help he p omo ion o ITS s anda ds de elopmen , he iden ifica ion and confi ma ion o needs, p oblems, objec i es and issues, and he alignmen o esea che s, companies and use s o syne gy [3]. The idea o ob aining his s uc u ed iew o a pa icula esea ch and de elopmen a ea is no new, bu nowadays is aking on majo impo ance. The mos ecen and ambi ious a emp has been made by he Eu opean commission by launching he echnological pla o m called ARTEMIS [9]. The aim o ARTEMIS is o de elop and d i e a join Eu opean ision on embedded sys ems connec ing esea ch and de elopmen wi h inno a ion o align agmen ed R&D e o s along common s a egic agenda and looking o imp o emen s in Eu opean companies’ e ficiency and compe i i eness. ARTEMIS is ollowing a bo om-up scheme in which he mos p ominen Eu opean companies a e ge ing in ol ed in he defini ion o his unique Eu opean ision h ough he de elopmen o a s a egic esea ch agenda (SRA). In IET In ell. T ansp. Sys ., 2010, Vol. 4, Iss. 3, pp. 201–211 201 doi: 10.1049/ie -i s.2009.0102 &The Ins i u ion o Enginee ing and Technology 2010 www.ie dl.o g pa icula , ITS can be loca ed in one o he ou applica ions con ex s iden ified by ARTEMIS SRA de o ed o public in as uc u e [9]. Al hough some s uc u ed iew o ITS has been p oposed based on ma ke a eas [3] o on da abases and cumula i e expe ience [10], his s udy p oposes a quan i a i e and sys ema ic me hodology o iden i y he main pa adigms wi hin he b oade field o ITS. The s a ing poin is he in o ma ion p o ided by he Ins i u e o Scien ific In o ma ion’s (ISI) massi e da ase s. The abs ac s o he published pape s included in da ase s will ha e been analysed using ex ca ego isa ions ools o ob ain he final pa adigms. To mee hese objec i es, he pape has been s uc u ed as ollows. In Sec ion 2, ITS will be analysed as a field s udy, desc ibing p e ious a emp s o s uc u ing his esea ch opic. In Sec ion 3, he p oposed me hodology will be desc ibed in de ail, including a discussion abou some o he app oaches. The ob ained pa adigms in ITS will be p esen ed in Sec ion 4, h ough he applica ion o he p oposed me hodology. Finally, he main conclusions o he wo k a e included in Sec ion 5. 2 Analysis o ITS as a field s udy The e a e di e en ways o classi y and segmen he ITS field. Six majo ca ego ies we e e iewed in [6] om a echnological pe spec i e: †Ad anced a fic managemen sys ems (ATMS), used o imp o e a fic se ice quali y and o educe a fic delays. †Ad anced a elle s in o ma ion sys ems (ATIS), used o supply eal- ime a fic in o ma ion o a elle s. †Comme cial ehicles ope a ion (CVO), sys ems ha use di e en ITS echnologies o inc ease he sa e y and e ficiency o comme cial ehicles and flee s. †Ad anced public anspo a ions sys ems (APTS), which make use o elec onic echnologies o imp o e he ope a ion and e ficiency o high-occupa ion anspo s, such as buses and ains. †Ad anced ehicles con ol sys ems (AVCS), which join senso s, compu e s and con ol sys ems in d i ing assis ance solu ions. †Ad anced u al anspo s sys ems (ARTS), used o sol e p oblems a ising in u al zones (s eep g ades, blind co ne s, cu es, sca ce na iga ional signs, mix o use s, lack o al e na i e ou es). Al hough his classifica ion is shaped by eme ging echnologies and is use ul om he iewpoin o sys em designe s, some in e es ing opics ela ed o ITS a e excluded. Fo ins ance, anspo policy and planning, a fic modelling and o ecas ing, and he sociological and beha iou al influences o ITS a e no included in he p e ious classifica ion. The equi emen s and p e e ences o ITS use s may also be gi en inadequa e a en ion. Ano he way o looking a ITS is o conside ma ke a eas as a ep esen a ion o ITS use s and ope a o s wi h simila needs. Nine majo ma ke a eas, de ailed in Table 1,a e defined in [3]. Simila classifica ions can also be ound in [10], al hough hey conside up o 13 a eas by sub-di iding se e al o he nine ma ke a eas defined in Table 1. As a di e ence o he p e ious classifica ions, mainly ocused on one aspec o ITS, an in eg al analysis o he ITS field is p oposed in his pape . The s a ing poin will be he abs ac s and keywo ds o published pape s ela ed o ITS included in he ISI Web o Science da abase [11, 12]. ISI Web o Science includes jou nals o almos Table 1 Ma ke a eas in ITS Ma ke a ea Goal a ea 1 a fic managemen manage he en i e oad ne wo k on behal o he gene al public a ea 2 eme gency managemen espond o incidences and eme gencies (fi e, police, ambulance) a ea 3 anspo a ion planning ma ch anspo a ion supply wi h demand bo h now and in he u u e a ea 4 a elle in o ma ion supply in o ma ion o a elle and subsc ibe s a ea 5 comme cial ehicles p o ide a el in o ma ion and flee managemen se ices a ea 6 ansi managemen plan and ope a e ansi sys ems in bo h u ban and u al a eas a ea 7 in elligen ehicles enhance he capabili ies o oad ehicles h ough he use o elec onics, senso s, communica ions and con ol ac ua o echnologies a ea 8 inciden managemen conce ned wi h e ficiency and sa e y o he oadway ne wo k a ea 9 paymen sys ems encompass all he people ha ake money in e u n o p o iding a se ice ( oll oad ope a o s, ansi agencies, ca pa king ope a o s, e c.) 202 IET In ell. T ansp. Sys ., 2010, Vol. 4, Iss. 3, pp. 201–211 &The Ins i u ion o Enginee ing and Technology 2010 doi: 10.1049/ie -i s.2009.0102 www.ie dl.o g each scien ific field. Consequen ly, no only echnological and ma ke app oaches o ITS will be ga he ed, bu any scien ific discipline s ongly o weakly ela ed o ITS, such as planning and logis ic, psychology o social sciences. The selec ed pape s will be p ocessed using a s a is ical ex ca ego isa ion ool and, as a esul , majo pa adigms in he field o ITS will be iden ified. 3 Me hodology The s a ing poin o he p oposed me hodology consis s o pape ex ac ion om ISI da abases. Pa icula ly, pape s ela ed o he opic ‘in elligen anspo a ion sys ems’. Ins ead o analysing he ull ex , which would be an eno mous ask, a ep esen a i e piece o ex summa ising he whole pape has been selec ed, i.e. he abs ac s and keywo ds o index e ms. An abs ac is a condensed e sion o a pape ha highligh s he majo poin s co e ed and concisely desc ibes he con en and scope o he w i ing, while keywo ds e iew he w i ing’s con en s in abb e ia ed o m. Keywo ds a e included because hey emphasise he con en o he pape . Al hough keywo ds hemsel es can cons i u e an adequa e desc ip ion o he pape con en [13], hey can also es ic he numbe o di e en opics o be ob ained. No ice ha some imes keywo ds mus be chosen among a closed lis p o ided by he jou nal publishe . Consequen ly, he p oposed me hodology will conside bo h abs ac s and keywo ds wi h he aim o lea ing open he numbe o opics o be ob ained. I is implici ly assumed ha au ho s w i e good abs ac s ( ep esen a i e o he con en o he pape ) and ha keywo ds a e ca e ully chosen. In gene al, bibliome ic esea ch is de o ed o quan i a i e s udies o li e a u e. Se e al empi ical me hods can be ound in he li e a u e. Co-ci a ion me hods a e pe haps he mos employed [14], and hey ha e been equen ly used o analyse he in ellec ual s uc u e o many disciplines [15]. The basis o co-ci a ion me hods consis s o coun ing he numbe o imes ce ain ma ke s occu o co-occu , gi ing ise o in o ma ion on such au ho co-ci a ion [16], jou nal co-ci a ion, keywo d co-ci a ion, e c. [13]. The main d awback o adi ional bibliome ic echniques, such as au ho o jou nal co-ci a ion me hods, is ha hey a e no conce ned abou he con en o conside ed pape s bu on e e ences usually delayed be ween 2 and 5 yea s a e a pape is fi s d a ed. Al hough hey lead o in e es ing esul s, hey do no p o ide an immedia e pic u e o he ac ual con en o he esea ch opic deal wi h in he li e a u e. As a di e ence, seman ic analysis based on co-wo ds analysis (co-occu ences o wo ds in he publica ions on a gi en subjec ) has he po en ial o sol ing his kind o p oblem [13–17]. Seman ic analysis usually employs a ec o space model [18], in which documen s a e summa ised and ep esen ed by ec o s o wo ds ( e m ec o s). Howe e , a cen al p oblem in his kind o s a is ical analysis is he high dimensionali y o he ea u e space (one dimension o each unique wo d). The e o e, i is desi able o fi s p ojec he documen s in o a lowe -dimensional subspace in which he seman ic s uc u e o he documen space becomes clea [19]. In he low-dimensional seman ic space, he adi ional clus e ing algo i hms can hen be applied. To his end, spec al clus e ing [20, 21], clus e ing using la en seman ic indexing (LSI) [22] and clus e ing based on non- nega i e ma ix ac o isa ion [23, 24] a e he mos well- known echniques. Pa icula ly, LSI decomposes a e m documen ma ix using a echnique called singula alue decomposi ion o cons uc new ea u es as combina ions o he o iginal ea u es, significan ly educing he high- dimensionali y p oblem o he ea u e space [25]. Mo eo e , LSI conside s documen s ha ha e many wo ds in common o be ‘seman ically close’, and ones wi h ew wo ds in common o be ‘seman ically dis an ’. The LSI app oach makes h ee basic claims: ha seman ic in o ma ion can be de i ed om a wo d-documen co-occu ence ma ix; ha dimensionali y educ ion is an essen ial pa o his de i a ion; and ha wo ds and documen s can be ep esen ed as poin s in a Euclidean me ic space. A di e en app oach has been applied in his pape . This app oach is consis en wi h he fi s wo o hese claims, bu i di e s in he hi d, desc ibing a class o s a is ical models in which he seman ic p ope ies o wo ds and documen s a e exp essed in e ms o p obabilis ic opics [26]. The opic model is a s a is ical language model ha ela es wo ds and documen s h ough opics. I is based upon he idea ha documen s a e mix u es o opics, whe e a opic is a p obabili y dis ibu ion o e wo ds [26–28]. In his pape , he me hodology p oposed o ITS pa adigms iden ifica ion consis s o a ou -s ep p ocedu e based on he la en Di ichle alloca ion (LDA) me hod o [26] (see Appendix o mo e de ails) and illus a ed in Fig. 1. Figu e 1 Da aflow o he p oposed me hodology IET In ell. T ansp. Sys ., 2010, Vol. 4, Iss. 3, pp. 201–211 203 doi: 10.1049/ie -i s.2009.0102 &The Ins i u ion o Enginee ing and Technology 2010 www.ie dl.o g 1. In o ma ion ex ac ion: he ex ac ion p ocess in ol es access o ISI da abases o anno a e he abs ac and keywo ds o a pape dealing wi h he opic ‘in elligen anspo a ion sys ems’. 2. On ology: an on ology defines he basic e ms and ela ions comp ising he ocabula y o a opic a ea as well as he ules o combining e ms and ela ions be ween e ms [29].An on ological model o he domain is used as a acili a o h oughou all he p ocesses. This s ep p o ides a common ocabula y and specifies he seman ics o key ela ionships wi hin he domain. The selec ion is based on he ob ained equency-o -occu ence- a es o wo ds o e he o al co pus o ex ac ed ex s. 3. S uc u ing and p ocessing in o ma ion: in s a is ical na u al language p ocessing, one common way o modelling he con ibu ions o di e en opics o a documen is o ea each opic as a p obabili y dis ibu ion o e wo ds, iewing a documen as a p obabilis ic mix u e o hese opics [27]. 4. Ca ego isa ion: The algo i hm ou lined abo e can be used o find he opics ha accoun o he wo ds used in a se o documen s. In his s udy, e e y abs ac and i s associa ed keywo ds a e conside ed a documen . 4 Pa adigms in ITS A o al o 1147 pape s ela ed o he opic ‘in elligen anspo a ion sys ems’ has been ob ained om ISI da abases, co e ing subjec a eas like enginee ing, anspo a ion, compu e science, elecommunica ions, au oma ion and con ol sys ems, and ope a ions esea ch and managemen science. The majo i y o hem belong o he gene al ca ego y o Science & Technology, bu a small pe cen age is associa ed o Social Sciences and A s & Humani ies. Acco ding o he explained p ocedu e, abs ac s and keywo ds ha e been used as documen s in he e minology o seman ic analysis. Up o 64 132 wo ds ha e been ex ac ed om hese documen s, leading o a ocabula y (non- epea ed wo ds) o 5485 wo ds. Wo king wi h such an amoun o wo ds would be p ohibi i e in e ms o compu ing ime, so on ology is selec ed o acili a e he implemen a ion o he p ocedu e. The c i e ion o he on ology selec ion was based on he equency o e ms in he documen collec ion. Pa icula ly, wo ds ha occu ed in mo e han 15 documen s we e conside ed [30].Asa esul , an on ology o 267 wo ds we e ob ained. This is he mos manual-s ep o he p ocedu e, because he final esul mus be supe ised o emo e wo ds ha a e no di ec ly ela ed o ITS issues. Table 2 shows he key dimensions used in he opic model. T and ITER ep esen opic model un pa ame e s. The numbe o Gibbs sample i e a ions was chosen o be ITER ¼200. This is a la ge enough alue o gua an ee he con e gence o he algo i hm [21]. The numbe o opics was selec ed using he pe plexi y alue. Pe plexi y is a s anda d measu e o pe o mance o s a is ical models o na u al language [31, 32] defined by (1). pplex =exp −1 W W n=1 log P(wn|dn)  (1) The ole o pe plexi y has mos ly been discussed on an in ui i e le el as a e age unce ain y when p edic ing he nex wo d gi en i s his o y. Pe plexi y indica es he unce ain y in p edic ing a single wo d. A lowe pe plexi y sco e indica es be e gene alisa ion pe o mance. Pe plexi y a ies om 1 o W; lowe pe plexi y is be e , and he maximum pe plexi y o Wis eached when all wo ds in he ocabula y a e equally likely. In ou case s udy, he LDA algo i hm was un o a numbe o opics a ying be ween 1 and 30. The esul s a e illus a ed in Fig. 2. The minimum pe plexi y alue is eached o a numbe o opics equal o 13. Consequen ly, 13 was chosen as he numbe o selec ed opics. Table 3 shows he 13 opics ob ained, wi h he mos likely wo ds in each opic, and hei p obabili ies P(w| ). Table 2 Dimensions o he opic model Pa ame e Desc ip ion Value Dnumbe o documen s in co pus 449 N o al numbe o wo ds in co pus 34 132 La e age leng h o documen in wo ds (L¼N/D) 76 Vnumbe o wo ds in ocabula y 5485 Won ology 267 Tnumbe o opics – ITER numbe o i e a ions 200 Figu e 2 Pe plexi y as a unc ion o he numbe o opics 204 IET In ell. T ansp. Sys ., 2010, Vol. 4, Iss. 3, pp. 201–211 &The Ins i u ion o Enginee ing and Technology 2010 doi: 10.1049/ie -i s.2009.0102 www.ie dl.o g Table 3 Ob ained opics and p obabili ies om he LDA algo i hm TOPIC 1 TOPIC 2 TOPIC 3 TOPIC 4 me hod 0.165 sys em 0.101 ehicle 0.265 inciden 0.079 e ec 0.087 se ice 0.074 ehicles 0.118 pe o mance 0.067 benefi 0.073 sys ems 0.073 sys em 0.069 de ec ion 0.065 analysis 0.072 anspo 0.059 in o ma ion 0.069 echnique 0.061 me hods 0.055 decision 0.054 ansi 0.048 a fic 0.053 benefi s 0.049 in o ma ion 0.049 mobile 0.038 eeway 0.049 e alua ion 0.047 amewo k 0.039 ese ed 0.027 ne wo k 0.047 e ec i e 0.047 suppo 0.039 au oma ic 0.025 managemen 0.047 esul 0.037 en i onmen 0.037 ision 0.025 condi ions 0.047 me hodology 0.032 assess 0.036 in as uc u e 0.025 neu al 0.040 in elligen 0.031 in elligen 0.032 app oach 0.022 sys em 0.037 s a ion 0.028 se ices 0.031 igh s 0.021 eal- ime 0.033 po en ial 0.026 e alua e 0.030 unc ion 0.021 echniques 0.033 e ec s 0.024 ope a ions 0.023 changes 0.020 cha ac e is ics 0.031 accu acy 0.022 impac s 0.020 co ido 0.020 delay 0.031 al e na i e 0.022 echnologies 0.019 mo ion 0.018 de ec o 0.027 e ec i eness 0.021 analysis 0.018 mo ing 0.017 p esen ed 0.027 expe imen al 0.021 wea he 0.018 scena ios 0.017 s a egies 0.023 TOPIC 5 TOPIC 6 TOPIC 7 TOPIC 8 sys em 0.191 algo i hm 0.187 anspo 0.127 a fic 0.294 sys ems 0.134 algo i hms 0.073 anspo a ion 0.126 simula ion 0.118 design 0.078 p ocess 0.059 de elop 0.089 signal 0.072 communica ion 0.060 p esen 0.057 sys em 0.066 esul s 0.047 p esen 0.051 de elop 0.049 s a e 0.053 es ima e 0.043 echnology 0.046 app oach 0.045 de elopmen 0.048 esul 0.038 posi ion 0.034 applica ion 0.045 planning 0.041 pa ame e s 0.034 highway 0.032 de eloped 0.044 egion 0.039 es ima es 0.032 message 0.028 senso 0.043 a chi ec u e 0.038 in elligen 0.030 esea ch 0.024 pe o mance 0.041 in elligen 0.032 a e ial 0.028 in eg a ed 0.023 e ficien 0.036 deploymen 0.031 empo al 0.024 p esen ed 0.020 complex 0.036 de eloping 0.028 oadway 0.023 implemen a ion 0.020 applica ions 0.030 p og am 0.026 in e sec ion 0.022 po en ial 0.019 anspo a ion 0.026 issues 0.025 mic oscopic 0.020 communica ions 0.019 in elligen 0.022 managemen 0.023 spa ial 0.019 global 0.017 scheme 0.022 p ocess 0.022 empi ical 0.017 Con inued IET In ell. T ansp. Sys ., 2010, Vol. 4, Iss. 3, pp. 201–211 205 doi: 10.1049/ie -i s.2009.0102 &The Ins i u ion o Enginee ing and Technology 2010 www.ie dl.o g Table 3 Con inued TOPIC 5 TOPIC 6 TOPIC 7 TOPIC 8 con ex 0.017 p ocessing 0.021 public 0.019 emissions 0.017 wi eless 0.016 senso s 0.019 egional 0.019 e alua ed 0.014 TOPIC 9 TOPIC 10 TOPIC 11 TOPIC 12 anspo a ion 0.144 con ol 0.138 ne wo k 0.164 model 0.356 anspo 0.142 d i e 0.092 p oblem 0.116 models 0.121 sys em 0.087 sys em 0.087 dynamic 0.089 a fic 0.067 applica ion 0.070 ehicle 0.067 ne wo ks 0.069 de elop 0.056 sys ems 0.068 sys ems 0.065 p oblems 0.042 anspo a ion 0.054 loca ion 0.045 esul 0.053 p esen 0.040 app oach 0.039 in elligen 0.044 d i ing 0.052 compu a ion 0.035 modelling 0.036 echnologies 0.043 d i e s 0.041 solu ion 0.033 de eloped 0.035 use 0.042 esul s 0.039 ou ing 0.033 anspo 0.030 applica ions 0.040 sa e y 0.033 guidance 0.030 u ban 0.029 ad anced 0.025 condi ions 0.022 assignmen 0.030 in elligen 0.028 esea ch 0.024 significan 0.021 s uc u e 0.028 o ecas ing 0.025 expe ience 0.022 beha iou 0.020 op imal 0.027 esea ch 0.024 unc ion 0.020 acking 0.019 p esen ed 0.024 demand 0.017 equi emen s 0.019 came a 0.016 eal- ime 0.022 sys ems 0.016 quali y 0.017 au onomous 0.016 op imisa ion 0.022 eg ession 0.015 comme cial 0.016 dynamics 0.015 sho es 0.020 p esen 0.012 solu ion 0.015 human 0.015 compu a ional 0.020 easibili y 0.012 TOPIC 13 a el 0.205 – – – – – – p edic 0.088 – – – – – – in o ma ion 0.071 – – – – – – imes 0.068 – – – – – – p edic ion 0.045 – – – – – – a is 0.045 – – – – – – esul 0.042 – – – – – – a elle 0.040 – – – – – – choice 0.035 – – – – – – esul s 0.031 – – – – – – es ima e 0.025 – – – – – – a el- ime 0.021 – – – – – – u u e 0.020 – – – – – – s a is ical 0.019 – – – – – – Con inued 206 IET In ell. T ansp. Sys ., 2010, Vol. 4, Iss. 3, pp. 201–211 &The Ins i u ion o Enginee ing and Technology 2010 doi: 10.1049/ie -i s.2009.0102 www.ie dl.o g Each opic can be de i ed om i s co esponding bag o wo ds, leading o he ollowing opic ca ego isa ion lis : †Topic 1. E alua ion o e ec i eness and benefi s o ITS: This opic deals wi h he analysis o implica ions and po en ial use o ITS including new challenges and oppo uni ies, and he an icipa ion o use beha iou in he p ocess o designing new ITS echnologies. †Topic 2. S udy o sys ems and ools suppo ing decision making and anspo a ion planning: I includes he de elopmen o models o a el demand decision and anspo a ion planning modelling ools, echnology in eg a ion and da abases suppo ing a chi ed da a use se ices. †Topic 3. Au oma ic ehicle de ec ion sys ems: Vehicle de ec ion appea s o be one o he mos p omising a eas in a fic su eillance and con ol. This concep en ails he de ec ion o ehicles and ex ac ion o a fic pa ame e s in eal- ime om images gene a ed by ideo came as o e looking a a fic scene. †Topic 4. Inciden managemen : Inciden managemen includes eme gency esponse deploymen and e ou ing o bypass he a ec ed a ea, impac o eeway lane closu es, and inciden de ec ion algo i hms. †Topic 5. Mobile and wi eless communica ion in ITS: This opic co e s in e ehicle communica ion ne wo ks, in- ehicle communica ions ), oad- o- ehicle communica ions, wi eless p o ocols, GPRS and hi d- gene a ion sys ems. †Topic 6. P ocessing algo i hms o ITS applica ions: This opic is de o ed o ad anced p ocessing algo i hms o sol ing ITS p oblems, such as pa h p oblems in dynamic ne wo ks, o igin–des ina ion es ima ion and p edic ion, and image and ideo p ocessing algo i hms. †Topic 7. Ad anced a fic managemen sys ems (ATMS): They a e ocused on he de elopmen o ITS o imp o e sa e y and quali y o se ice as well as he e ficiency o exis ing oadway u ilisa ion. †Topic 8. T a fic simula ion: A a fic simula ion sys em consis s o a a fic-flow simula ion code, which is able o simula e a fic on a eeway ne wo k. I usually conside s a mic oscopic ep esen a ion (whe e each indi idual ehicle is ep esen ed) o a mac oscopic model cap u ing a fic dynamics. The pu pose o hese sys ems consis s o pe o ming a fic-flow simula ion o applica ions like a fic condi ions p edic ion in eal- ime, a fic con ol and d i e s’ guidance, link a el ime calcula ion and signal con ol s a egy. This opic also co e s issues like modelling d i e beha iou unde he influence o ex e nal ac o s. †Topic 9. Comme cial Vehicles Ope a ion (CVO): This opic is ocused on he impac o ITS on comme cial ehicles and flee s o imp o ing anspo a ion sa e y and e ficiency. †Topic 10. Ad anced ehicles con ol sys ems (AVCS): AVCS a e based on sys ems ha p o ide inc eased sa e y and/o con ol o he d i e ei he by means o imp o ing he in o ma ion abou he d i ing en i onmen o by ac i ely aiding he d i e in he d i ing ask. They include on- boa d au onomous in elligen c uise con ol sys ems, ABS and ac ion con ol sys ems, ac i e suspension sys ems, ehicle s abili y sys ems, in- ehicle collision wa ning sys ems, e c. †Topic 11. Dynamic ou e selec ion algo i hms: Dynamic ou e selec ion p oblems a e sea ch p oblems o finding an op imal ou e om a s a ing o a des ina ion poin on a oad map wi hin a ime limi . Since he ime o a e se a link will depend upon a fic olume encoun e ed on ha link, link imes a e dynamic. †Topic 12. Models o a fic demand o ecas ing: T a fic conges ion is a majo ope a ional p oblem on ITS. Reducing conges ion e ec s equi es de eloping models ha can accu a ely p edic a fic demand. †Topic 13. Ad anced a elle s in o ma ion sys ems (ATIS): The unc ion o ATIS is o assis a elle s wi h planning, pe cep ion, analysis and decision making o imp o e he con enience and e ficiency o a el. The ob ained esul s depic an exhaus i e d aw o ITS esea ch a eas. Some ma ke -specific opics a e supp essed compa ed o p e ious classifica ion me hods desc ibed in Sec ion 2, like APTS o ARTS in [6] o eme gency managemen and paymen sys ems in [3–10], bu new esea ch opics a e meanwhile iden ified. In ac , up o six new esea ch a eas ha e now been de ec ed om he ob ained 13 opics. Fo ins ance, specific opics like e alua ion o e ec i eness and benefi s o ITS, mobile and Table 3 Con inued TOPIC 13 p opaga ion 0.019 – – – – – – conges ion 0.018 – – – – – – es ima ed 0.017 – – – – – – igh s 0.017 – – – – – – IET In ell. T ansp. Sys ., 2010, Vol. 4, Iss. 3, pp. 201–211 207 doi: 10.1049/ie -i s.2009.0102 &The Ins i u ion o Enginee ing and Technology 2010 www.ie dl.o g wi eless communica ions in ITS o models o a fic demand o ecas ing ha e no been p e iously es ablished. They ha e eme ged as he esul o he inco po a ion o new echnologies, and also due o he necessi y o modelling and assessing hei social and economical impac . The p oposed ca ego isa ion should help esea che s and p ac i ione s o cla i y hei posi ion o u u e wo k because he ob ained 13 opics ep esen he mos ecen issues and eme gen echnologies in he ITS wo ld. The opic co ela ion ma ix (Table 4) shows ha opics a e poo ly co ela ed wi h each o he , which means ha opics a e well defined and hei scope is clea ly delimi ed. Ne e heless, i is impossible o achie e pe ec ly delimi ed opics wi h independen scopes. The e is always some deg ee o o e lap. O e lapping can be used o ob ain se e al majo pa adigms a ending o he opic simila i y. Fo his pu pose, a mul i a ia e s a is ical echnique like mul idimensional scaling was used [33]. This analysis consis s o p ojec ing he wo ks on a wo-dimensional map, using he da a om he co ela ion ma ix as inpu da a. Fig. 3 shows he ob ained map. The ob ained RSQ coe ficien (0.93430) and K uskal’s s ess (0.15374) sugges ha goodness o fi is e y accep able ( he RSQ coe ficien is he squa ed co ela ion index R 2 ha measu es he model fi o he da a, and i s minimum accep able sco e is 0.6 Figu e 3 Mul idimensional scaling Table 4 Topic co ela ion ma ix T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12 T13 T1 1.00 0.13 0.10 0.07 20.05 20.03 0.06 0.14 0.07 20.01 20.10 0.03 0.06 T2 0.13 1.00 0.01 0.00 0.12 20.07 0.16 0.07 0.16 20.07 0.03 0.10 0.04 T3 0.10 0.01 1.00 0.03 0.13 0.01 20.06 0.10 0.07 0.13 20.02 20.02 0.08 T4 0.07 0.00 0.03 1.00 20.05 0.14 20.05 0.19 20.02 0.05 20.02 0.21 0.21 T5 20.05 0.12 0.13 20.05 1.00 0.05 0.21 20.02 0.28 0.05 20.03 20.09 20.08 T6 20.03 20.07 0.01 0.14 0.05 1.00 20.04 0.04 0.01 0.03 0.34 20.02 0.00 T7 0.06 0.16 20.06 20.05 0.21 20.04 1.00 20.06 0.12 20.10 20.06 20.02 20.07 T8 0.14 0.07 0.10 0.19 20.02 0.04 20.06 1.00 20.01 0.08 0.03 0.25 0.15 T9 0.07 0.16 0.07 20.02 0.28 0.01 0.12 20.01 1.00 20.01 0.04 20.03 0.02 T10 20.01 20.07 0.13 0.05 0.05 0.03 20.10 0.08 20.01 1.00 20.04 0.00 20.10 T11 20.10 0.03 20.02 20.02 20.03 0.34 20.06 0.03 0.04 20.04 1.00 0.04 0.13 T12 0.03 0.10 20.02 0.21 20.09 20.02 20.02 0.25 20.03 0.00 0.04 1.00 0.30 T13 0.06 0.04 0.08 0.21 20.08 0.00 20.07 0.15 0.02 20.10 0.13 0.30 1.00 208 IET In ell. T ansp. Sys ., 2010, Vol. 4, Iss. 3, pp. 201–211 &The Ins i u ion o Enginee ing and Technology 2010 doi: 10.1049/ie -i s.2009.0102 www.ie dl.o g [34]) and ha he map exhibi s a good app oxima ion o eali y. The p oximi ies o opics on he map show hei simila i y and sugges a meaning o he axes. The ho izon al axis is ela ed o he applica ion a ea. Topics loca ed o he le pa o he map (T4, T8, T12 and T13) a e ocused on he imp o emen o a fic in u ban a ea. Topics loca ed on he cen e o he map conside ITS ools om wo di e en pe spec i es, he implemen a ion de ails ep esen ed by opics T3, T6, T10 and T11 a he op o he map and hei benefi s and po en ial use ep esen ed by T1 and T2 a he bo om o he map. Finally, opics loca ed o he igh pa o he map a e ocused on ehicle and a fic managemen (T5, T6 and T7). The e ical axis is ela ed o he le el o ha dwa e implemen a ion, as i can be clea ly deduced om opics in he uppe hal o he map, conce ned wi h he implemen a ion de ails o ITS ools (T3, T6, T10 and T11) o he mobile and wi eless communica ion possibili ies suppo ing a fic managemen (T5). Compa ing hese esul s wi h p e ious classifica ions de ailed in Sec ion 2, no ice ha ‘Benefi s and po en ial use o ITS ools’ we e no conside ed om a echnological pe spec i e and he ‘implemen a ion de ails o ITS ools’ we e no conside ed om a ma ke a ea pe spec i e. Consequen ly, he map o Fig. 3 summa ises a mo e comple e ision o ITS. 5 Conclusion The main con ibu ion o his pape is a global iew o ITS, which could be used by u u e esea che s as a s a e o he a o me hods, echniques and applica ion a eas. The analysis is in ended o o e new pe spec i es in o wha is iewed as impo an o build upon, p o iding aluable insigh s in o bo h wha esea ch is impo an and whe e he field o ITS is heading. The p oposed me hodology o p oducing his global iew is based on he seman ic analysis o keywo ds and abs ac s o pape s indexed by he ISI. As a di e ence o au ho o jou nal co-ci a ion me hods, seman ic analysis is ocused on he con en o pape s, so esul s a e no biased by ci es o i ele an li e a u e o ecu en ci ed pape s. A o al o 1147 pape s ha e been analysed co e ing a la ge a ie y o opics ela ed o ITS, including he mos ecen issues and eme gen echnologies. As a esul o he analysis, 13 pa adigms we e ob ained. Using a mul idimensional scaling, hey ha e been ep esen ed on a bidimensional map, illus a ing he mos ela ed pa adigms as well as he b idges among hem. Fu he mo e, his s udy also defines a s a ing poin o o he analyses aimed a a be e unde s anding o he ITS field. This con inuous analysis is conside ed necessa y as ITS is an e olu iona y field influenced by changes in echnology, wi h changing se ices and suppo o end use s. 6 Acknowledgmen s The au ho s g a e ully acknowledge suppo p o ided by he Spanish Minis y o Educa ion and Science (p ojec wi h e e ence DPI2007-60128) and he Conseje ı ´ade Inno acio ´n, Ciencia y Emp esa (Resea ch P ojec wi h e e ence P07-TIC-02621). 7 Re e ences [1] ANDRISANO O.,VERDONE R.,NAKAGAWA M.: ‘In elligen anspo a ion sys ems: he ole o hi d gene a ion mobile adio ne wo ks’, IEEE Commun. Mag., 2000, 38, (9), pp. 144–151 [2] TORAL S.,VARGAS M.,BARRERO F.: ‘Embedded mul imedia p ocesso s o oad- a fic pa ame e es ima ion’, Compu e , 2009, 42, (12), pp. 61–68 [3] MCQUEEN B.,MCQUEEN J.: ‘In elligen anspo a ion sys ems a chi ec u es’ (A ech House, 1999) [4] NATVIG M.K.,WESTERHEIM H.: ‘Na ional mul imodal a el in o ma ion – a s a egy based on s akeholde in ol emen and in elligen anspo a ion sys em a chi ec u e’, IET In ell. T ansp. Sys ., 2007, 1,(2), pp. 102–109 [5] TORAL S.,VARGAS M.,BARRERO F.,ORTEGA M.G.:‘Imp o ed sigma-del a backg ound es ima ion o ehicle de ec ion’, Elec on. Le ., 2009, 45, (1), pp. 32–34 [6] FIGUEIREDO L.,JESUS I.,MACHADO J.A.T.,FERREIRA J.R.,MARTINS DE CARVALHO J.L.: ‘Towa ds he de elopmen o in elligen anspo a ion sys ems’. P oc. 2001 IEEE In elligen T anspo a ion Sys ems, Oakland, CA, USA, 2001, pp. 1206–1211 [7] WANG F.-Y.,MIRCHANDANI P.B.,ZHENG N.:‘Ad ancesand ends in esea ch and de elopmen o in elligen anspo a ion sys ems: an in oduc ion o he special issue’, IEEE T ans. In ell. T ansp. Sys ., 2004, 5,(4), pp. 222–223 [8] ZHENG N.,KAWASHIMA H.: ‘Ad ances and ends in esea ch and de elopmen o ehicula elec onics and sa e y: special sec ion’, IEEE T ans. In ell. T ansp. Sys ., 2007, 8, (1), pp. 106–107 [9] A emis S a egic Resea ch Agenda Wo king G oup: ‘S a egic esea ch agenda’ (Ideo L d, 2006, 1s edn. edn.) [10] US Depa men o T anspo a ion, Fede al Highway Adminis a ion: ‘In elligen anspo a ion sys ems benefi s, cos s and lessons lea ned’ (Mi e ek Sys ems, 2005) IET In ell. T ansp. Sys ., 2010, Vol. 4, Iss. 3, pp. 201–211 209 doi: 10.1049/ie -i s.2009.0102 &The Ins i u ion o Enginee ing and Technology 2010 www.ie dl.o g