CumInCAD is a Cumulative Index about publications in Computer Aided Architectural Design
supported by the sibling associations ACADIA, CAADRIA, eCAADe, SIGraDi, ASCAAD and CAAD futures

PDF papers
References
id sigradi2006_e172c
authors Donath, Dirk and González Böhme, Luis Felipe
year 2006
title A Constraint-Based Building Bulk Design Support
source SIGraDi 2006 - [Proceedings of the 10th Iberoamerican Congress of Digital Graphics] Santiago de Chile - Chile 21-23 November 2006, pp. 278-282
summary We introduce an architecture practice-oriented implementation strategy of constraint-based methods called BDS (Building Bulk Design Support) to supporting bulk analysis during the architectural programming phase. We examine the optmization problem of site coverage and building massing according to a set of standard planning and zoning regulations, and try a problem solving approach based on the paradigm of constraint satisfaction problems. The case study, which is focused on the paticipatory planning of very low-income dwellings within the Latin American context, serves as testbed for a prototypical application of the adopted methodology. The BDS constitutes a novel approach on computer-aided bulk analysis, regarding this particularly relevant context of application. In the case of participatively planned low-income housing projects, efficiency regarding time and cost of planning directly affects dwellers’ quality of life, whereas elementary programming tasks such as bulk analysis lack appropriate state-of-the-art technological support. Traditional architectural planning methods demand a large domain-specific knowledge base and skillful planners. A planning process, which is mainly driven by the formulation of planning-relevant constraints and sets of solution alternatives, suggests to avoid architects’ traditional procedure of: 1. Create an (yet not necessarily valid) instance of the eventual design solution by directly choosing specific values for its shape parameters. 2. Evaluate its validity by confronting the designed model to a set of applicable constraints, which have to be satisfied. Instead, the constraint-based design methodology poses a search procedure that operates in a space of pertinent constraint sets. A computer-aided interactive search procedure to find more valid design solution alternatives in less time and with less effort is particularly qualified to supply efficient support for participatory planning activities carried out between dwellers and planners. The set of solutions for a building-bulk design problem is constrained by both a large complex system of planning and zoning regulations and the geometry of the eventual design solution itself. Given a considerable amount of such regulations, a regular size geometric constraint satisfaction system proved to be capable of providing a highly efficient, interactive modeling and evaluation tool for the formulation in real time of valid solution alternatives for an ordinary building-bulk design problem. A BDS implementation will constitute one system module of a larger integrated system model called Esther. A BDS tool shall interact with other functional modules, like e.g. the FLS (Floor plan Layout Support), which also uses constraint-based design methods.
keywords constraint-based design; bulk analysis; participatory planning; low-income housing; design theory; design proces
series SIGRADI
email luis-felipe.gonzalez@archit.uni-weimar.de
full text file.pdf (403,317 bytes)
references Content-type: text/plain
Details Citation Select
100%; open Gross, M. (1986) Find in CUMINCAD Design as Exploring Constraints , Ph.D. dissertation, Massachusetts Institute of Technology, Cambridge

100%; open Kramer, G. (1994) Find in CUMINCAD A Geometric Constraint Engine , Constraint-Based Reasoning: Special Issues of Artificial Intelligence, Freuder, E. and Mackworth, A. (eds.), The MIT Press, pp. 327-360

last changed 2016/03/10 08:50
pick and add to favorite papersHOMELOGIN (you are user _anon_450521 from group guest) CUMINCAD Papers Powered by SciX Open Publishing Services 1.002