Planning a digi ally in eg a ed u u e:
1
GIS and BIM o oman a chi ec u e,
2
he case o Pompeii A chaeological
3
Pa k
4
Ang isani Gio anni1, Bo iello Gio anni2, Bosco Angela*2,
5
Ce a Vale ia1 Fo e F ancesca2, F egonese Luigi3,
6
Rosignoli Olga3, Scandu a Simona1
7
8
1 Uni e si à degli S udi di Napoli Fede ico II – Naples, I aly
9
2 Uni e si à di Napoli L’O ien ale – Naples, I aly
10
3 Poli ecnico di Milano – Milan, I aly
11
12
13
*Angela Bosco
14
Co espondence: abosco@unio .i
15
16
ABSTRACT
17
The ongoing wa e o digi iza ion ac oss scien i ic domains is eshaping he ields o cul u al
18
he i age and a chaeology. Ne e heless, as demons a ed by p e ious esea ch, he
19
excep ional na u e o he i age his o ic asse s o en con lic s wi h so wa e a chi ec u es, such
20
as da a schemas and communica ion p o ocols, ha a e no concei ed wi h hose speci ic
21
equi emen s in mind. The a chaeological emains p ese ed wi hin he Pompeii A chaeological
22
Pa k cons i u e an ex ao dina y case s udy, whe e a chi ec u al ea u es ha e been p ese ed
23
almos in hei en i e y. Gi en his con igu a ion, any digi al managemen ool designed o
24
suppo he si e’s conse a ion mus in eg a e he unc ionali ies o bo h Building In o ma ion
25
Modelling (BIM) and Geog aphic In o ma ion Sys ems (GIS), ensu ing hei seamless
26
in e ope abili y.
27
This s udy o e s a quan i a i e e alua ion o his in e ope abili y challenge- equen ly men ioned
28
in a icles bu no demons a ed ye -by iden i ying he p ecise in o ma ion loss ha occu s du ing
29
he digi iza ion p ocess and he ollowing alignmen wi h es ablished da a s anda ds.
30
Speci ically, a seman ic mapping be ween he en i ies in ol ed in he p ocess is p esen ed: he
31
open- o ma s anda ds IFC ( o BIM) and Ci yGML ( o GIS) a e compa ed wi h a pu pose-buil
32
axonomy de eloped o Vesu ian Roman a chi ec u e.
33
The esul ing co espondence ma ix indica es he concep ual ela ionships ac oss schemas,
34
assessing hei ins umen al ele ance o a chaeologis s, conse a o s, and si e manage s.
35
To u he en ich he seman ic ep esen a ion o a chaeological da a, he esea ch explo ed he
36
possibili ies gi en by on ologies and hesau i. This app oach allowed o he de ini ion o
37
domain-speci ic concep s and ela ionships ha a e no adequa ely cap u ed by exis ing BIM
38
and GIS s anda ds, he eby aligning he da a mo e closely wi h a chaeological p ac ices.
39
All his cons i u es a ounda ional s ep owa d he de elopmen o an in eg a ed digi al sys em
40
ha b idges he BIM, GIS and on ological domains, while emaining esponsi e o he
41
me hodological p inciples and ope a ional needs a icula ed by a chaeological expe s.
42
Keywo ds: BIG_SMAART, BIM, GIS, seman ics, Pompeian a chi ec u e
43
In oduc ion
44
This con ibu ion is pa o he b oade con ex o he digi al amewo k o he digi isa ion o
45
a chaeological asse s. Nowadays, he e a e many s udies ha highligh he excep ional na u e
46
o his o ical he i age and how i clashes wi h so wa e, schemas, o da a p o ocols ha ha e no
47
been speci ically cons uc ed wi h a chaeology's needs in mind (Bosco e al., 2021; Bosco e
48
al., 2019; Ga agnani e al., 2020).
49
Du ing he las decades -as well as in he pas - a chaeology has looked i elessly o c ea i e
50
and sa is ying solu ions o ep esen bo h he complex s a us quo and he de elopmen o
51
a chaeological si es. I s complica ed long ela ionship wi h digi al ep esen a ion ools is pa o
52
his on-going esea ch, one ha s a ed se e al yea s ago, mainly wi h GIS (Geog aphic
53
In o ma ion Sys ems) ep esen a ion. Then, wi h he popula i y o BIM inc easing in he sec o ,
54
he demand o i s in eg a ions o GIS d as ically g ew, and he i s expe imen s in his ega d
55
s a ed almos simul aneously (Delpozzo & Balle i, 2023)
56
Wi hin his con ex , he aim o his in es iga ion is o de ine a heo e ical s uc u e necessa y
57
o he o ganisa ion o he he e ogeneous da a ha con ibu e o he s udy o a chaeological
58
e idence, iden i ying in he ele an a ailable schemas he poin s o connec ion be ween
59
ep esen a ion sys ems such as BIM, GIS and he a chaeological domain.
60
The BIG_SMAART p ojec , unded by he I alian Minis y o Uni e si ies and Resea ch,
61
in ol es he co-ope a ion o h ee uni e si ies: he Uni e si y o Naples L'O ien ale, he
62
Uni e si y Fede ico II o Naples and he Poli ecnico di Milano.
63
64
Figu e 1 - P ojec wo k low wi h di ision o he a ious phases
65
The aim o his p ojec is o p o ide a quan i a i e assessmen o he in e ope abili y be ween
66
hese sys ems. Al hough he opic is widely known in he li e a u e, i is s ill a e o ind esea ch
67
demons a ing wha in o ma ion is los du ing digi isa ion p ocesses and he compliance
68
be ween he mos widely used da a schemas in BIM and GIS.
69
Conside ing hese assump ions, o de ine he speci ic wo k low (Fig. 1), i was decided o
70
de ine h ee main esea ch ques ions:
71
(a) do cu en GIS and BIM da a schemas o exis ing on ologies dedica ed o a chaeology
72
mee he equi emen s o he ep esen a ion o ancien a chi ec u e and especially Vesu ian
73
a chi ec u e?
74
(b) does a da a schema suppo hese equi emen s be e han i s equi alen ?
75
(c) how can BIM and GIS me hods be in eg a ed in an a chaeological con ex ?
76
The esea ch aims o add ess hese ques ions h ough he analysis o a conc e e case s udy
77
in which da a o ma s ha e been applied and e alua ed.
78
To his pu pose, he p ojec 's me hodology is based on he analysis o how he seman ic
79
da a p o ided by hese sys ems o e lap wi h he a chaeological needs. Since he
80
a chaeological wo ld is so wide- anging, he expe imen a ion ocuses on he Vesu ian a ea,
81
which is unique because i was c ys allised in 79 AD. Thanks o his condi ion, he con ex
82
e eals ex ensi e in o ma ion abou li e and echnology in he Roman wo ld o he Republican
83
and ea ly Impe ial pe iods.
84
Mo e speci ically, we p opose a seman ic mapping be ween he en i ies ha a e pa o he
85
p ocess, s a ing om an analysis o he mos widely used open o ma s anda ds o BIM and
86
GIS, namely IFC and Ci yGML. We hen p oceeded o compa e he seman ic possibili ies
87
o e ed by hese o ma s wi h he seman ic ee c ea ed speci ically o Roman a chi ec u e in
88
he Vesu ius a ea.
89
Digi al Da a Managemen in he Pompeian con ex
90
The a chaeological analysis o e idence and cons uc ion echniques was ex ended o allow
91
he c ea ion o a seman ic ee (a schema ic hie a chical model) use ul o b eaking down he
92
single elemen s o he complex a chaeological eali y o Pompeii. The choice o he si e is linked
93
o a ious easons, which can be summa ised in he ollowing poin s:
94
• Ex ensi e, well-p ese ed a chaeological e idence, on which o es as many
95
cases as possible;
96
• E idence o which he e is a good le el o p e ious documen a ion, e en when
97
i conce ns exca a ions ca ied ou in pas cen u ies;
98
• A complex eali y o manage, cha ac e ised by mode n in as uc u e ha adap s
99
o ancien e idence;
100
• The olume o e idence equi es an in o ma ic sys em capable o p o iding
101
e ec i e managemen suppo .
102
O e ime, he Pa k has implemen ed, h ough he G ea Pompeii P ojec , esea ch ac ions
103
ocused on he collec ion and sys ema isa ion o da a wi h a p ecise iew o p e en i e
104
main enance. In pa icula , he so-called “Knowledge Plan”
1
has led o he c ea ion o SiPompei,
105
a geo e e enced and ela ional pla o m ha desc ibes and ca alogues he en i e ancien ci y,
106
joined by Open Pompeii (h ps://open.pompeiisi es.o g/index.h ml), a digi al a chi e wi h ee
107
access aimed a acili a ing he dissemina ion o da a ha is in eg a ed wi h SIAV (in o ma ion
108
sys em o he Vesu ian a ea) and Tolomeo o a chi al ma e ials and his o ical documen a ion.
109
1
The da a om he GPP_Knowledge Plan has been kindly p o ided by he A chaeological Pa k o
Pompeii.
The Knowledge Plan has also p oduced new 3D su ey documen a ion wi h lase -scanning
110
and pho og amme ic me hodologies co e ing he en i e su ace o he Pa k as o 2015. This
111
h ee-dimensional documen a ion, howe e , is un ela ed o he sys ems men ioned abo e.
112
I he e o e ollows ha he conse a ion and managemen equi emen s o such a complex
113
si e would necessi a e he combined use o BIM and GIS unc ionali ies.
114
Based on hese assump ions, he ini ial phase o he esea ch in ol ed an in-dep h e iew
115
o he s a e o he a ega ding he in eg a ion o BIM and GIS applied o a chaeology. The
116
analysis o he li e a u e made i possible o iden i y he da a o ma s mos equen ly a es ed
117
in p e ious s udies, he so wa e mos adop ed in his ield o in es iga ion, as well as o de ine
118
he main p oblems me by o he esea ch g oups wi h he aim o se ing up an e icien mapping
119
be ween he en i ies belonging o hese knowledge s uc u es we a e aiming o communica e.
120
The choice o he case s udy
121
To e i y he heo e ical model de ined a he ou se o he p ojec , an applica ion case was
122
iden i ied and selec ed.
123
The choice o he House - Bake y VIII, 6, 1-9/11 (Fig. 2), is linked o se e al ac o s: he
124
loca ion o he complex be ween wo public a eas, he Ci il o um and he T iangula o um; a
125
complex s a ig aphic sequence wi h a ious modi ica ions and es o a ions, including mode n
126
ones; he ype o building s uc u e ha was ans o med om a p i a e house in o a p oduc i e
127
ac i i y wi h he addi ion o peculia ins alla ions such as he o en and mills ones.
128
The a chaeological complex was al eady disco e ed a he end o he 19 h cen u y, bu i is
129
s ill almos uns udied
2
. The e o e, a census and ca aloguing o he a chi ec u al and unc ional
130
elemen s o he complex, as well as he building echniques and ma e ials used o hei
131
cons uc ion, was ca ied ou (Bo iello e al., 2025).
132
Al hough no widely known o he gene al public, he Domus — in addi ion o occupying a
133
s a egic posi ion wi hin he opog aphical amewo k o he ancien ci y — is pe ec ly sui ed o
134
he expe imen al needs o he p ojec due o i s s uc u al complexi y and i s mul i-laye ed
135
ch onological and unc ional s a i ica ion.
136
2
The i s exca a ions ook place be ween 1881 and 1882.
137
Figu e 2 - Pho o o he o en ins alled in place o he cubiculum eas in he
138
a ium o he House - Bake y VIII, 6, 1-9/11 (Cou esy o he A chaeological
139
Pa k o Pompeii)
140
The seman ic b eakdown o Vesu ian cons uc ions
141
To make he managemen o he e idence as sui able as possible o i s use wi h BIM and
142
GIS pla o ms, a seman ic b eakdown o p i a e cons uc ion elemen s in Pompeii was ca ied
143
ou (Fig. 3). The seman ic scheme was based on he HBIM model, whe e en i ies a e desc ibed
144
as: Class, Ca ego y, Type and Ins ance, a emp ing o adap i wi hin he limi s o a chaeological
145
equi emen s. The main componen s o he Roman building we e di ided in o h ee disciplines
146
(s uc u al, a chi ec u al and plan enginee ing), based on he unc ion o he objec in he
147
Roman building complex. Each discipline was hen di ided in o classes o se s, i.e. elemen s
148
ha sha e he same o mal and unc ional cha ac e is ics (wall, coin, pilla ), which can be di ided
149
in o subclasses o subse s, based on he a ia ion o ce ain cha ac e is ics (e.g. ec angula
150
pilla , squa e pilla , o al pilla ). This is ollowed by ca ego ies, which allow he class o be b oken
151
down in o cons i uen elemen s (e.g. column, which is subdi ided in o capi al, sha and base).
152
Likewise, he subca ego y allows he o de o building echnique used o be de ined (e.g. opus
153
ince um sha ). Finally, he ype indica es he ma e ial used (e.g. limes one); he ins ance, on
154
he o he hand, cons i u es he indi idual objec c ea ed (Bo iello e al., 2025).
155
156
Figu e 3 - Seman ic s uc u e o Pompeian a chi ec u e (de eloped by G. Bo iello based on
157
he da a in a ached n.1)
158
In addi ion o being assigned an ID du ing modelling, each ins ance was accompanied by a
159
se o da a ha allows o he managemen o a chaeological in o ma ion in a ious ways. The
160
modelling phase and subsequen implemen a ion o BIM objec s we e s uc u ed o combine,
161
on he one hand, a b oad pa e n o a chaeological equi emen s and, on he o he hand, o
162
a oid o e loading he BIM sys em. All his equi ed ope a ional choices o be made (Ca pen ie o
163
- D’Au ia 2023): he addi ion o nume ous amilies a ailable locally would make i possible o
164
compensa e o he lack o a chaeological elemen s wi hin he BIM pla o m; howe e , his could
165
lead o he possible loss o essen ial da a du ing expo .
166
Among o he necessa y choices, i was use ul o e e o he da a managemen sys em o
167
he A chaeological Pa k o Pompeii o he nomencla u e o he walls. This sys em in ol es
168
di iding a single wall in o wo di e en sides, so in he in o ma ion se o each ins ance, i was
169
necessa y o add a ield ha would allow he wo sepa a e sides o be joined in o a single wall
170
en i y.
171
Ano he choice in ol ed he in oduc ion o a code o iden i y elemen s o he han walls, such
172
as loo s, ceilings and cladding.
173
Las ly, necessa y choices conce ned he econs uc ion o he quoins, iden i ied as sec ions
174
o mason y, and he doo ways, which we e modelled as emp y spaces. These examples a e
175
jus a ew o he di icul ies ha a e e iden in he modelling phase. Fo his eason, i was clea
176
ha he e was a need o compa e he seman ics o Roman a chi ec u e wi h IFC o ma s and
177
he main on ologies o he a chaeological domain.
178
The seman ic compa ison be ween da a schemas
179
IFC
3
is bo h he schema and exchange o ma mo e widely used in he BIM domain. The
180
hie a chical s uc u e ha a anges classes in descending o de and connec s hem h ough
181
speci ic ela ionships is ac ually e y complex and a he same ime desc ip i e: a he poin o
182
de elopmen i eached nowadays (wi h he e sion 4x3 o he s anda d) i can be a gued ha
183
3
BuildingSMART Alliance 2024, Indus y Founda ion Classes Release 4.3 ADD2 (IFC 4.3
ADD2) (h ps://s anda ds.buildingsma .o g/).
IFC, o he a ailable and app o ed-on da a schema is he mo e ad anced in he seman ic
184
desc ip ion o buil asse s; i was pu o he compa ison wi h he c ea ed “pompeian” classes
185
and a mapping (connec ion) was pe o med o he mo e seman ically co esponden . The
186
co espondence was egis e ed bo h in a sp eadshee and g aphically.
187
As an example, in Fig. 4, we posi ion he class i cWall, i s declina ion in “Ope a Vi a a” and
188
he ma e ials ha make i up and de e mine he ype. The a ious wall “ins ances” (en i ies in
189
he p ojec ) will inhe i hese cha ac e is ics.
190
In a compa ison wo k o comple e mapping o en i ies, a missing pe cen age ela ionship
191
was es ima ed o se e al elemen s such as quoin, a ch o h eshold, o which he e a e no IFC
192
classes.
193
Going in o he speci ics o he class “Wall”, he declina ion in he ypology s anda d was also
194
analysed. As can be seen g aphically om he mapping connec ions in Fig. 5, he “A chaeo-
195
Wall” class (on he le ) co esponds o he i cWall class o he 4x3 s anda d (on he igh ), he
196
la es o icial e sion. The ypologies, on he o he hand, low in o an i cWallType unique class
197
whe e he a ious speci ica ions due o building wo ks and ma e ials can subsequen ly be
198
c ea ed inside he p ojec . Op ionally, he a ious ins ances can ake on an addi ional
199
Enume a i e Type which, howe e , does no help o be e speci y he seman ics a he le el
200
equi ed by he a chaeological needs.
201
The same mapping p ocess was ca ied ou wi h Ci yGML , he de- ac o seman ic s anda d
202
in he geospa ial sciences. Designed o ep esen a ions a he u ban scale, which includes
203
mo phological da a o buil and na u al en i onmen s, simila ly o IFC has a dually s uc u ed
204
na u e (wi h geome y and in o ma ion combined) bu i is ope a es in he Geog aphy Ma kup
205
Language (GML).
206
Due o he numbe o classes (named “ ea u es”) a ailable, Ci yGML is subdi ided in
207
in e connec ed modules, ha connec h ough he same hie a chy. This s anda d is la e and
208
less desc ip i e han IFC, and cus omisa ion mechanisms ha e o be employed ha de ia e
209
om he p ede ined classes o each good. The de ini ion o ypes is less comple e han he IFC
210
because i mus be ep oduced wi h a cus omised mechanism called “CodeLis ”. Using he
211
example illus a ed be o e, speci ically o he class “WALL”, his lows in o he “Building
212
Cons uc i e Elemen ”, a e y gene ic concep , while he ypes o low in o he “CodeLis ”, which
213
is no globally app o ed (Rosignoli e al., 2025).
214
215
Figu e 4 - he comple e a chaeological classes “ ee” on he le and a zoom-in on he en i y
216
wall in “ope a ince a” (de eloped by O. Rosignoli in Mi o, based on he da a in a ached n.2)
217
218
Figu e 5 - in he image, an ex ac om he isual mapping pe o med be ween
219
he classes o he ex-no o hie a chical “ ee” cons uc ed o he a chaeological
220
en i ies and he pa ial hie a chical schema o IFC 4X3. The ex ac ocuses on
221
wall en i ies. I is clea om he numbe o a ows ha o en con e ge in he same
222
di ec ion owa ds IFC classes ha he e is a loss o seman ic g anula i y.The
223
comple e a chaeological classes “ ee” on he le and a zoom-in on he en i y
224
wall in “ope a ince a” (de eloped by O. Rosignoli in Mi o, based on he da a in
225
a ached n.2)
226
The possibili ies p o ided by on ologies: an o e iew
227
The need o align his seman ic decomposi ion wi h he mos popula s anda d o ma s
228
equi ed he in eg a ion o da a wi h addi ional sys ems capable o suppo ing his ype o
229
desc ip ion. This led o a e lec ion on concep ual ep esen a ion ools and seman ic
230
in e ope abili y close o he a chaeological domain, di ec ing an analysis o wha is al eady
231
a ailable in he cu en landscape and wha al e na i es hey p esen ed o he desc ip ion o
232
he en i ies o he a chaeological seman ic ee. E idence o his can be ound in on ologies, i.e.
233
concep ual models ha p o ide a o mal s uc u e o desc ibe in o ma ion in a sys ema ic and
234
in e ope able way. The sys em o on ologies has ound wide applica ion in cul u al he i age as
235
i p o ides a common language o he desc ip ion o he asse , allows a s anda diza ion o
236
e ms and concep s o ensu e a uni o m desc ip ion, a ou s in e ope abili y be ween sys ems
237
and suppo s mo e ad anced and p ecise esea ch, since i enables da a o be que ied based
238
on concep s and hei ela ionships, no only on ex ual e ms. In addi ion, hese sys ems allow
239
o p o iding adequa e suppo in he iden i ica ion o classes and p ope ies o be included in
240
he seman ic ee o he a chaeological building, whe e he concep ual s uc u e o e ed by he
241
BIM and GIS sys ems does no su icien ly mee he needs o a chaeological documen a ion.
242
In pa icula , he esea ch in ol ed he analysis o in e na ional and na ional on ologies and
243
hesau i:
244
