On-si e demons a ions o C-ITS a chi ec u e and
au onomous ehicle ope a ion in he ci y o T ikala,
G eece
Ch is os Ballis1, Anna An onakopoulou2, E angelos Tsougiannis3,
Tasos Theodo idis4, Alexand os Liazos5, Odisseas Rap is6, Loukas Va i sas7,
Elena Pa a ouka8, Angelos Amdi is9
1 ICCS, [email p o ec ed], G eece, 2 ICCS, [email p o ec ed], G eece,
3 ICCS, [email p o ec ed], G eece, 4 ICCS, [email p o ec ed], G eece,
5 ICCS, [email p o ec ed], G eece, 6 e-T ikala, o[email p o ec ed], G eece,
7 e-T ikala, [email p o ec ed], G eece, 8 e-T ikala, [email p o ec ed], G eece
9 ICCS, [email p o ec ed], G eece
Abs ac . This pape del es in o he Coope a i e In elligen T anspo Sys ems
(C-ITS) a chi ec u e o he on-si e pilo conduc ed in he ci y o T ikala,
G eece, as pa o he EU- unded p ojec , IN2CCAM. The pilo es s and
demons a es inno a i e Coope a i e, Connec ed, and Au oma ed Mobili y
(CCAM) se ices h ough he deploymen o au onomous ehicles (AVs) and
hei in eg a ion wi h a C-ITS pla o m. Fea u ing a lee o au onomous elec ic
mini ans on a p ede ined ou e equipped wi h ad anced sma digi al in as uc-
u e, his ini ia i e aligns wi h he ci y's ision o ackle conges ion, p omo e he
use o sha ed anspo se ices, and enhance he equi y o he local anspo
sys em. Fu he mo e, a Mobili y as a Se ice (MaaS) mobile applica ion ena-
bles he public o plan mul imodal ips, in eg a ing on-demand sha ed passen-
ge anspo se ices wi h AVs, public anspo se ices, sha ed mic omobili y
solu ions, and ac i e mobili y, such as walking o cycling. The p oposed C-ITS
a chi ec u e includes: a) AV lee managemen and moni o ing, b) a G een
Ligh Op imal Speed Ad iso y (GLOSA) unc ionali y o achie e smoo he and
mo e uel-e icien jou neys, c) a a ic-based g een wa e sys em o con ibu e
o high a ic e iciency by adjus ing a ic signals acco ding o eal- ime con-
ges ion le els, and d) C-ITS messages be ween he in as uc u e and he AVs
o ale abou uncon olled c ossings o Vulne able Road Use s (VRUs). This
pape aims o ou line he a chi ec u al componen s o his pilo and discuss u-
u e wo k, which will include e alua ing he easibili y, e ec i eness, and soci-
e al impac o he p oposed inno a ions in he local ecosys em o T ikala.
Keywo ds: CCAM, C-ITS, AV, MaaS, GLOSA, VRU, pilo .
1 In oduc ion
Coope a i e, connec ed, and au oma ed mobili y (CCAM) p io i izes mo ing people
and goods along he oad ne wo ks in a sa e, quick, cos -e ec i e, com o able, and
2
en i onmen ally iendly manne . Coope a i e In elligen T anspo Sys ems (C-ITS)
echnologies, au oma ed ehicles (AVs), ad anced a ic managemen sys ems, and
Mobili y-as-a-Se ice (MaaS) pla o ms, among o he s, a e u ilised o his aim.
The scope o he p oposed wo k is o implemen a e e ence C-ITS a chi ec u al
schema wi h he scope o in eg a e an au onomous passenge anspo a ion se ice
in o he local anspo a ion ecosys em o he ci y. Wi hin he amewo k o he EU
IN2CCAM p ojec , he ci y o T ikala, G eece, se es as a hub o he on-si e demon-
s a ions o C-ITS solu ions discussed in his pape .
The T ikala pilo ea u es a lee o wo elec ic, e o i ed, and au oma ed
mini ans which ha e been g an ed pe mission o a el wi hou a d i e on a p ede-
ined ou e equipped wi h ad anced sma digi al in as uc u e. The 9.6 km long
ou e uns be ween he ci y cen e and he Science and Spo s Uni e si y and has been
selec ed as i se es he ailway s a ion, he ci y’s hema ic pa k, spo s acili ies, and
majo subu bs and illages ha a e unde se ed om he cu en public anspo line.
The au onomous se ice mainly a ge s uni e si y s uden s and esiden s o a eas on
he ou ski s o he ci y. I o e s on-demand ips which can be booked ia a mobile
applica ion ha ac s as a mul imodal jou ney planne o he ci y. Use s can use he
mobile applica ion o book ips be ween ixed loca ions along he ou e. In addi ion,
o ensu e seamless in eg a ion wi h he local public anspo sys em, he au onomous
se ice s ops ha e been selec ed o align wi h exis ing unde se ed bus s op loca ions.
Fig. 1. Sc eensho o he AV ou e in T ikala.
3
In summa y, he solu ions implemen ed in his pilo p ojec aim o imp o e he
sa e y and e iciency o he au onomous se ice, as well as o imp o e i s in e ac ion
wi h mixed a ic lows along sa e y-c i ical and conges ed sec ions o he ou e.
Namely, he ollowing ad ancemen s will be de eloped and deployed:
a) Ad anced a ic managemen se ices o enhance he managemen and moni-
o ing o he AV lee in mixed a ic low along he selec ed ou e.
b) A a ic-based g een wa e sys em which con ibu es o highe a ic e icien-
cy by adjus ing a ic signals acco ding o eal- ime a ic olume.
c) A G een Ligh Op imal Speed Ad iso y (GLOSA) unc ionali y o achie e
smoo he and mo e uel-e icien jou neys.
d) C-ITS messages be ween he in as uc u e and he AVs o ale abou uncon-
olled c ossings o Vulne able Road Use s (VRUs).
e) A Mobili y-as-a-Se ice (MaaS) mobile applica ion o enable he public o
plan mul imodal ips. This app in eg a es he on-demand sha ed passenge
anspo se ice wi h AVs, public anspo op ions, sha ed mic omobili y so-
lu ions, and ac i e modes like walking o cycling.
This pape is s uc u ed as ollows: Sec ion 2 ou lines he a chi ec u e o he C-ITS
solu ions included in his pilo , Sec ion 3 discusses he a p io i e alua ion o he inno-
a ions, Sec ion 4 p esen s he u u e wo k, and las ly, Sec ion 5 concludes his pape .
2 A chi ec u e
2.1 O e iew
The a chi ec u e suppo ing he on-si e demons a ions o he au onomous se ice in
T ikala, G eece, is designed o seamlessly in eg a e a ious componen s, all aimed a
enhancing u ban mobili y and sa e y. An ou line o he a chi ec u e is p esen ed be-
low.
4
Fig. 2. A chi ec u e o he C-ITS solu ions implemen ed in he T ikala pilo s udy.
The ou main componen s o his a chi ec u e a e:
a) The physical and digi al in as uc u e, which include Road-Side Uni s
(RSUs), ins alled a signalized in e sec ions, On-Boa d Uni s (OBUs),
moun ed on he oo s o he AVs, a ic ligh s equipped wi h a ic ligh
con olle s, a ic came as wi h objec de ec ion capabili ies, and a MaaS
applica ion.
b) The ITS backend, which includes a se o backend p ocesses along wi h a lo-
cal da abase o s o e he eco ded da a.
c) The T a ic Managemen Cen e (TMC), which inco po a es lee and ope a-
ional managemen capabili ies ela ed o he AV se ice, along wi h addi-
ional TMC capabili ies.
d) The AVs, wi h hei so wa e and ha dwa e subcomponen s.
2.2 T a ic-based g een wa e sys em
T a ic conges ion is among he mos challenging p oblems in u ban managemen . To
his aim, g een wa e s a egies aim o op imise a ic low by coo dina ing signal
imings along co ido s o educe conges ion and emissions. Such s a egies play an
essen ial ole in in elligen anspo a ion sys ems (Zheng e al., 2020). The ou come
o a g een wa e sys em would be less s essed mo o is s, mo e eliable anspo a ion,
punc ual bus se ices, dec eased noise pollu ion, and ewe pollu an s (SWARCO,
2024). On he con a y, g een wa e policies could lead o enhanced a ac i eness o
p i a ely owned ca s, and, hus, inc eased gas emissions, especially on some p ima y
a e ials (Blode and Jäge , 2021).
Rega ding s udies ha examine he impac s on a ic a e applying g een wa e
s a egies, one could e e o Tseng, Chang and Kuo (2017), who simula ed he u ili-
5
sa ion o ehicula ad hoc ne wo ks (VANETs) o apply i ual a ic ligh s a egies.
The esul s implied ha a ic low and a e age speed a e imp o ed by his ac ic,
while ene gy is also sa ed. In a di e en s udy, Shi e al. (2020) p oposed a high-
e iciency mul i-in e sec ion coo dina ion algo i hm based on Vehicle- o-E e y hing
(V2X) communica ions, simula ed i s applica ion, and showcased, likewise, ha a e -
age speed inc eased, whe eas olume- o-capaci y pe cen age and numbe o s ops
dec eased.
In his p ojec , his solu ion in ol es he use o s a iona y a i icial in elligence
(AI) a ic came as ins alled a one o he main in e sec ions along he ou e in
T ikala whe e he elec ic, e o i ed, and au oma ed mini ans a e pe mi ed o a el
au onomously du ing he T ikala demons a ions. The came as p o ide a con inuous
insigh in o he eal- ime a ic le els o he a ea and, consequen ly, allow o ad-
jus men s o he ele an a ic signal p og ams. P e equisi e o his au oma ed p o-
cess is he s o age and analysis o his o ic a ic da a, which include classi ied a ic
coun s and speeds. Fi een-minu e in e als a e used, and ypical a ic le els a e
de e mined o each. The p ocess in ol es analysing he cu en le el o a ic, and
by compa ing i agains he ypical si ua ion, i applies inc emen al changes o he
signal phases. In o he wo ds, he du a ion o he g een ime alloca ed o he main
ou e adjus s, wi hin a p e-selec ed ange, acco ding o he cu en a ic olume.
This is achie ed wi h he de elopmen o Applica ion P og amming In e aces (APIs)
which in e ac wi h T a ic Ligh Con ol and Managemen so wa e and apply he
ele an signalling p og ams.
2.3 G een Ligh Op imal Speed Ad iso y (GLOSA)
The GLOSA applica ion p o ides p ecise and imely in o ma ion abou a ic signal
imings and a ic ligh posi ions ia in as uc u e- o- ehicle (I2V) communica ion,
which allows d i e s o c uise wi h mo e cons an speeds and wi h less s ops a a ic
ligh s (Ka sa os e al., 2011). In his pilo , he GLOSA unc ionali y is being es ed
and demons a ed wi h one o he wo au oma ed mini ans. Signal Phase and Timing
Ex ended Messages (SPATEM) and MAP ( opology) Ex ended Messages (MAPEM)
a e b oadcas by an RSU loca ed on Pylis s ee and ecei ed by he OBU o he AV.
The o me includes eal- ime in o ma ion abou he a ic ligh signal phase and
imings, while he la e p o ides in o ma ion abou he opology and geome y o he
signalized junc ion. A web applica ion has been de eloped o display his in o ma ion
in a clean, in ui i e, and easy- o-comp ehend way o he sa e y d i e . The ou pu o
he p ocess is shown on a cus om Human-Machine In e ace (HMI) on a able in-
s alled inside he mini an. Fu he mo e, he Au oma ed D i ing Sys em (ADS) o he
ehicle u ilises he GLOSA- ecommended speeds o adjus he ehicle’s speed ac-
co dingly, i ha is possible gi en he a ic condi ions and he p esence o obs acles.
A o al o i e signalised in e sec ions ha e been equipped wi h RSUs which can
exchange messages ia ITS-G5, cellula ne wo k (e.g., 4G/5G), and Wi-Fi/E he ne ,
among o he s. Fo his GLOSA applica ion, he ollowing p ocesses a e implemen ed.
Fi s , a backend p ocess has been de eloped o access he SPATEM, eques he ehi-
cle’s loca ion and speed in o ma ion, and hen combine hem so ha hey a e used as
6
inpu in he GLOSA algo i hm. The GLOSA calcula ions ake place on he ly and he
sugges ed speeds a e displayed o he sa e y d i e s on a web applica ion. In his
demons a ion, he C-ITS messages a e ecei ed om he OBU when he mini an is
wi hin app oxima ely 250m o he signalised junc ion. The GLOSA algo i hm in-
ol es he main ollowing s eps.
a) S a s wi h checking whe he he e is a ele an a ic ligh in p oximi y.
b) I yes, he dis ance o he a ic ligh and he ime equi ed o each i (deno ed
as T l) a e calcula ed.
c) De e mines he colou o he a ic ligh a T l. Depending on he colou o he
a ic ligh , i ollows di e en p ocesses o calcula e a ecommended speed.
d) The p ocess loops back o check o ele an a ic ligh s in p oximi y, con inu-
ously adjus ing he speed as needed.
The low cha o he GLOSA algo i hm is shown in Fig. 3.
Fig. 3. Flow cha o he GLOSA algo i hm demons a ed in T ikala.
7
The ange o ecommended speeds has been se so ha hey a e wi hin he ange ha
he AV is allowed o c uise by law, in o he wo ds a maximum speed o 30kph. Las -
ly, e y low ecommended speeds a e excluded o a oid c ea ing sa e y haza ds due
o e y slow mo emen in a ic.
Fig. 4. Sc eensho o he web applica ion ha displays he GLOSA- ecommended speeds.
2.4 VRU de ec ion and wa ning o app oaching AVs
VRUs include a wide ange o oad use s, such as pedes ians, cyclis s and powe ed
wo wheele s (Schollie s, an Sambeek and Moe man, 2017). They ha e been iden i-
ied by policymake s as a g oup o a ic pa icipan s ha dese es speci ic a en ion
in e o s o educe he numbe o a ali ies and se ious inju ies (Eu opean Commis-
sion, Di ec o a e-Gene al o Mobili y and T anspo , 2020). VRUs a e mo e suscep-
ible o oad a ic acciden s and a ali ies han o he oad use s. Imp o ing hei sa e-
y s ands ou as a p ima y conce n o many municipali ies and go e nmen s, and
a ious measu e a e aken o mi iga e acciden isks and inc ease oad sa e y (Hejazi
and Boko , 2024). P e ious VRU-sa e y p ojec s aimed o imp o e sa e y by u ilizing
on-boa d senso s, such as came as o ada s, sys ems which a e pa icula ly use ul in
se e al use cases, howe e hey a e no e ec i e when he e is no line-o -sigh con-
nec i i y be ween ehicle and VRU (Schollie s, an Sambeek and Moe man, 2017).
So o e al. (2022) no e ha he e a e s ong incen i es o p o ide ehicles wi h
communica ion-based applica ions o inc ease he sa e y o hei occupan s and o he
oad use s. Fu he mo e, a VRU p o ec ion c oss-wo king g oup o he 5G Au omo-
i e Associa ion (5GAA) eleased a whi e pape in 2020 whe e hey ocused on he
ollowing use cases (5GAA Au omo i e Associa ion, 2020):
a) VRU high isk zones: i.e., s a ic o dynamic messages o d i e s o ehicles
when hey en e high isk a eas whe e he p esence o many VRUs is likely.
8
b) In e ac i e communica ions be ween VRU and ehicles: i.e., nego ia ions be-
ween he VRU’s de ice and a ehicle (e.g., a VRU asks he ehicle o pe -
mission o c oss he oad).
c) VRU sa e y messages and algo i hms: i.e., he mos common use case, whe e
isk assessmen is con inuously pe o med by he mos app op ia e uni (e.g.
in as uc u e o ehicle) and wa nings a e issued o a oid collisions.
The VRU de ec ion and wa ning sys em, implemen ed in T ikala, belongs o he
‘VRU sa e y messages and algo i hms’ ca ego y. I aims o imp o e he sa e y o bo h
VRUs and passenge s ha make use o he au oma ed se ice. S a iona y AI a ic
came as, moun ed on a ic poles, ha e been con igu ed o de ec VRUs ha c oss
he oad in an uncon olled way (i.e., wi hou using designa ed c osswalks o a ic
signals). When such an e en is de ec ed, a backend p ocess is ini ia ed o p omp he
ansmission o C-ITS messages om he RSU ins alled a he c ossing. The messages
a e ecei ed by he OBU o he AV, i he ehicle is nea by he in e sec ion, and,
subsequen ly, hey a e displayed on a able de ice (i.e., he same web applica ion ha
displays he GLOSA ou pu ), which is ins alled in on o he sa e y d i e o inc ease
hei awa eness.
Fig. 5. Sc eensho o he AI-based ideo analy ics de ec ing uncon olled c ossings on Pylis
s ee in T ikala, G eece.
9
Fig. 6. Sc eensho o he web applica ion ha displays VRU-wa nings o he sa e y d i e ( he
VRU-wa ning unc ionali y is in eg a ed in he web applica ion u ilised o GLOSA).
2.5 Mobili y-as-a-Se ice (MaaS) applica ion
A Mobili y-as-a-Se ice (MaaS) is a ela i ely new concep , o en c edi ed o Sampo
Hie anen o in oducing i o he Finnish T anspo a ion Minis y in 2006 (A ias-
Molina es and Ga cía-Paloma es, 2020). Among he se e al de ini ions gi en in he
p e ious yea s, Kama gianni and Ma yas (2017) de ined MaaS as ‘a use -cen ic,
mul imodal, sus ainable and in elligen mobili y managemen and dis ibu ion sys em,
in which a MaaS P o ide b ings oge he o e ings o mul iple mobili y se ice p o-
ide s (public and p i a e) and p o ides end-use s access o hem h ough a digi al
in e ace, allowing hem o seamlessly plan and pay o mobili y.’
As A ias-Molina es and Ga cía-Paloma es (2020) poin ou in hei e iew pape ,
anspo au ho i ies and ope a o s, ei he p i a e o public, should coope a e o
achie e op imal implemen a ion o MaaS, while a da a d i en app oach could p o e
bene icial o anspo planning and policy making. Polydo opoulou e al. (2020)
shaped he key elemen s o he MaaS business eco-sys em ia da a collec ion in Man-
ches e , Luxembou g, and Budapes , and highligh ed mobili y p o ide s, public
anspo , and egional au ho i ies as he key s akeholde s ha should build us
among one ano he . The gaps ega ding s anda diza ion and legal amewo k, as well
as he lack o open p og amming in e aces we e s a ed as ba ie s o ex ensi e MaaS
implemen a ion. Giesecke, Su akka and Hakonen (2016), whils ag eeing ha MaaS is
he nex pa adigm change in he anspo sec o , had al eady p o ided conce ns e-
ga ding i s concep ual issues, such as in e ope abili y wi h ITS, ac ual sus ainabili y,
and end-use accep ance. La e , he s udy o Bu le , Yigi canla and Paz (2021) e e s
o possible lack o coope a ion and us be ween public and p i a e coope a ions, as
well as o possible di icul y appealing o elde ly people and equen use s o p i a e-
ly-owned ehicles.
16
Schollie s, J., an Sambeek, M. and Moe man, K. (2017) ‘In eg a ion o ulne able oad use s
in coope a i e ITS sys ems’, Eu opean T anspo Resea ch Re iew, 9(2), p. 15. A ailable a :
h ps://doi.o g/10.1007/s12544-017-0230-3.
Shen, Y., Zhang, H. and Zhao, J. (2018) ‘In eg a ing sha ed au onomous ehicle in public
anspo a ion sys em: A supply-side simula ion o he i s -mile se ice in Singapo e’, T ans-
po a ion Resea ch Pa A: Policy and P ac ice, 113, pp. 125–136. A ailable a :
h ps://doi.o g/10.1016/j. a.2018.04.004.
Shi, Y. e al. (2020) ‘A Coo dina ion Algo i hm o Signalized Mul i-In e sec ion o Maximize
G een Wa e Band in V2X Ne wo k’, IEEE Access, 8, pp. 213706–213717. A ailable a :
h ps://doi.o g/10.1109/ACCESS.2020.3039263.
So o, I. e al. (2022) ‘A su ey on oad sa e y and a ic e iciency ehicula applica ions based
on C-V2X echnologies’, Vehicula Communica ions, 33, p. 100428. A ailable a :
h ps://doi.o g/10.1016/j. ehcom.2021.100428.
SWARCO (2024) G een Wa e in T a ic: Challenges. A ailable a :
h ps://www.swa co.com/mobili y- u u e/in elligen - anspo a ion-sys ems/g een-wa e- a ic
(Accessed: 25 Ap il 2024).
Tseng, C.-W., Chang, Y.-J. and Kuo, Y.-W. (2017) ‘G een wa e-based i ual a ic ligh
managemen scheme wi h VANETs “G een wa e-based i ual a ic ligh managemen
scheme wi h VANETs”’, In . J. Ad Hoc and Ubiqui ous Compu ing. A ailable a :
h p://ecomo e-p ojec .eu.
Venka esh, V. e al. (2003) ‘Use Accep ance o In o ma ion Technology: Towa d a Uni ied
View’, MIS Qua e ly, 27(3), pp. 425–478. A ailable a : h ps://doi.o g/10.2307/30036540.
Venka esh, V., Thong, J.Y.L. and Xu, X. (2012) ‘Consume Accep ance and Use o In o -
ma ion Technology: Ex ending he Uni ied Theo y o Accep ance and Use o Technology’,
MIS Qua e ly, 36(1), pp. 157–178. A ailable a : h ps://doi.o g/10.2307/41410412.
Zheng, Y. e al. (2020) ‘A No el App oach o Coo dina ing G een Wa e Sys em Wi h Adap a-
ion E olu iona y S a egy’, IEEE Access, 8, pp. 214115–214127. A ailable a :
h ps://doi.o g/10.1109/ACCESS.2020.3037129.
