CI PA 2003 XIX 11 ' International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
requirements, processes and techniques called for are similar to 
those found in the fields of surveying, construction research, 
archaeology, the history of art and architecture, as well as in the 
care and preservation of monuments. A common starting point 
is a graphical survey. Our aim, therefore, is the conception of an 
integrated building information system, combined with a 
digitally-supported survey. 
2.1 Capture and Structure 
The starting point for the conception of an ordering system was 
to achieve the maximum generalisation whilst not ignoring the 
complexity and specificity inherent in architecture. 
Principal emphasis is laid on the systematisation of the built 
structure of a building, and the establishment of relevant 
planning and use-related information. Taken into account are 
the planning methods of architects working in the field of 
building restoration and reconstruction. 
A typical problem when surveying existing buildings is that a 
large amount of information is recorded without an overview of 
the overall situation. The tendency is to concentrate on details, 
whereby more straightforward structural connections within the 
building remain overlooked. These problems can be countered 
through the use of an ordering system that is used from the 
outset. 
The basic principle is a room-by-room process. Buildings are 
perceived as a series of different rooms, each room being a 
functional unit in itself. The appearance of a room is described 
by its surfaces. Built elements can be described as material 
'rooms', defined by their surfaces. 
There are two primary ordering principles: 
(A) Room structure - the spatial subdivision of the buildings 
complexes can be arranged both as entire buildings or 
individual rooms, and 
(B) Element structure - the hierarchy of built elements in the 
building that which defines space and from which the geometry 
of the building is measured. 
The project takes the approach that there is no generally 
applicable building model. The aim is instead to provide the 
user with appropriate tools to adapt a flexible building model to 
the specific situation at hand. The coupling of dynamic data 
structures with static basic structures allows flexible models to 
be adapted using fixed algorithms. 
informal data relational data 
Figure 1. Information container 
formal data 
width: 1 2 
height 4.0 
depth: 1.2 
information 
geometric data 
A series of attributes and their range of possibilities were 
identified, based upon practical experience in the field of 
architectural surveying. 
The following are examples of object properties and 
characteristics relevant in working with existing buildings: 
method of construction and material 
building damage and extent of damage 
constructional qualities 
colour 
historic information and value etc. 
Nevertheless, the architectural survey tends to concentrate on 
the geometric qualities. We have developed four different levels 
of abstraction corresponding to the phases in architectural 
practice (fig. 2): 
taxonomy or icon oriented (level 1) 
sketch orientated (level2) 
2D-plan orientated (level3) 
3D-model orientated (level4) 
These correspond to the individual phases in the surveying 
process. 
brt kutfch*c2 
Level 1 : taxonomy 
• î..:. cj“j3L xx. ja. 
f vi/L _• y«f' j 
Level 3: 2D-plan oriented 
Level 2: sketch 
Level 4: 3D-model oriented 
Figure 2. Different levels of detail for describing building 
substance (in this case building geometry) 
2.2 Complex digital building models 
The computer-supported building survey is not simply a 
geometric description of a building. It should also provide a 
multitude of features and characteristics relevant both to the 
buildings future use and to its later CAD processing. 
In addition to surveying geometric data, formal (specific 
building data), informal (text, images, sketches etc.) and 
relational data (probable structural relationships) can be 
captured. The captured data is stored within an information 
container that can be used within both the spatial structure as 
well as the building element structure (fig.l). Geometric data is 
“only one attribute” of such an information container. 
3. PROTOTYPICAL IMPLEMENTATION 
Based upon the analysis and the identification of deficiencies, a 
hypothetical system was developed with the aim of providing a 
set of tools for the computer-assisted surveying of buildings. 
This developed prototype-software 'freak 1 illustrates the path 
from sketch to measured 3D-model (fig. 3). The user initially 
creates a three-dimensional sketch of the object to be surveyed 
which provides a rough topological and geometrical direction. 
Through the addition of measurements, for instance through 
tacheometry, photogrammetry or manual measurements, the 
sketch takes on more and more of the actual situation up to a 
given degree of accuracy. This process is assisted through the 
use of geometrical constraints, for instance parallel surfaces etc.