The computer-aided spatial analysis of building models is an important feature for efficient data management and productive downstream processes in Building Information Modelling. Users of building models have to deal with complex data sets. At the same time, it is also important to have a deep understanding of the functionality and design of a building. Topological and directional queries provide a means of gaining insights into the spatial structure of even a large-scale project. To handle the massive amount of data involved in these projects, the previously introduced Spatial Query Language for Building Information Models is extended by a new, more efficient approach for its topological and directional predicates. Instead of the Octree representation of components employed in the past, the new methods operate directly on the boundary representation, i.e. triangulated meshes. In addition, R-Trees are used for a two-stage spatial indexing of the scene. This makes it possible to achieve a substantial improvement in performance for topological and directional predicates of the Spatial Query Language. 1. Efficient Spatial Analysis in large-scale BIM Projects The application of building information modeling (BIM) and its paradigms is taken for granted in the construction industry nowadays. In large-scale projects in particular, BIM methods lead to better management of the huge amount of data and the processes involved. It is possible to detect and rectify errors in the design and project scheduling at an early stage before the construction work commences. However, undetected mistakes in the planning of a building can lead to difficulties in the downstream process. High-quality results and efficient workflow in the construction phase can only be achieved if the data basis is accurate and functional. Formal spatial analysis can be employed to check the geometric validity of building information models. To this end, (Borrmann & Rank, 2009a, 2009b) devised a Spatial Query Language for Building Information Models. Algorithms based on Octree representations were proposed for implementing the topological and directional operators provided by this language. Due to the nature of the Octree representation and the associated algorithms, however, unsatisfactory processing times had to be taken into account when applied to full-scale building information models. The aim of this paper is to introduce a new implementation approach designed to overcome these restrictions. Instead of the previously used strategies based on Octrees, the algorithm presented here operates directly on the boundary representation of the operands, thus considerably accelerating the execution of topological and directional predicates. This makes it possible to process complex spatial queries within reasonable runtimes from the viewpoint of an end user. It is practically impossible to analyze large-scale models entirely manually. The spatial validation that is the subject of this paper is of crucial importance, especially in complex building models with a huge number of components and fine geometrical representations. In keeping with this requirement, the algorithms of the topological and directional predicates make use of a two-stage spatial indexing. At both levels, R-Trees (Guttman, 1984) are used.
[1]
Ernst Rank,et al.
Specification and implementation of directional operators in a 3D spatial query language for building information models
,
2009,
Adv. Eng. Informatics.
[2]
Steven A. Gaal,et al.
Point Set Topology
,
1964
.
[3]
Hanan Samet,et al.
The Design and Analysis of Spatial Data Structures
,
1989
.
[4]
Dinesh Manocha,et al.
Collision queries using oriented bounding boxes
,
2000
.
[5]
Tomas Akenine-Möller,et al.
A Fast Triangle-Triangle Intersection Test
,
1997,
J. Graphics, GPU, & Game Tools.
[6]
Stefanie Schraufstetter,et al.
AABB-Bäume als Grundlage effizienter Algorithmen für Operatoren einer räumlichen Anfragesprache
,
2007
.
[7]
Ernst Rank,et al.
Topological analysis of 3D building models using a spatial query language
,
2009,
Adv. Eng. Informatics.
[8]
Tomas Möller,et al.
A fast triangle-triangle intersection test
,
1997
.
[9]
Marios Hadjieleftheriou,et al.
R-Trees - A Dynamic Index Structure for Spatial Searching
,
2008,
ACM SIGSPATIAL International Workshop on Advances in Geographic Information Systems.
[10]
Tetsuo Tomiyama,et al.
Advanced Engineering Informatics
,
2007,
Adv. Eng. Informatics.