Wednesday, November 20, 2002

Compilation on Space Allocation Problem: 1980 onwards


1974. Shaviv, Edna and Dov Gali
A Model for Space Allocation in Complex Buildings : A Computer Graphics Approach

Build International. England: Applied Science Publisher Ltd., 1974. vol. 7: pp. 493-517 : ill. includes bibliography

A model for the organization of activities in a complex building is presented. The model aimed at the minimization of an objective function based on circulation and subject to architectural constraints

1986, Layout Design Problems: Systematic Approaches

Computer-Aided Architectural Design Futures [CAAD Futures Conference Proceedings / ISBN 0-408-05300-3] Delft (The Netherlands), 18-19 September 1985, pp. 28-52

The complexity of the layout design problems known as the 'spatial allocation problems' gave rise to several approaches, which can be generally classified into two main streams. The first attempts to use the computer to generate solutions of the building layout, while in the second, computers are used only to evaluate manually generated solutions. In both classes the generation or evaluation of the layout are performed systematically. Computer algorithms for 'spatial allocation problems' first appeared more than twenty-five years ago (Koopmans, 1957). From 1957 to 1970 over thirty different programs were developed for generating the floor plan layout automatically, as is summarized in CAP-Computer Architecture Program, Vol. 2 (Stewart et al., 1970). It seems that any architect who entered the area of CAAD felt that it was his responsibility to find a solution to this prime architectural problem. Most of the programs were developed for batch processing, and were run on a mainframe without any sophisticated input/output devices. It is interesting to mention that, because of the lack of these sophisticated input/output devices, early researchers used the approach of automatic generation of optimal or quasioptimal layout solution under given constraints. Gradually, we find a recession and slowdown in the development of computer programs for generation of layout solutions. With the improvement of interactive input/output devices and user interfaces, the inclination today is to develop integrated systems in which the architectural solution is obtained manually by the architect and is introduced to the computer for the appraisal of the designer's layout solution (Maver, 1977). The manmachine integrative systems could work well, but it seems that in most of the integrated systems today, and in the commercial ones in particular, there is no route to any appraisal technique of the layout problem. Without any evaluation techniques in commercial integrated systems it seems that the geometrical database exists Just to create working drawings and sometimes also perspectives.


Gross , Mark D.
1996, Elements That Follow Your Rules: Constraint Based CAD Layout

Design Computation: Collaboration, Reasoning, Pedagogy [ACADIA Conference Proceedings / ISBN 1-880250-05-5] Tucson (Arizona / USA) October 31 - November 2, 1996, pp. 115-122

The paper reports on CKB (Construction Kit Builder) a prototype CAD program that designers can program with positioning and assembly rules for layout of building elements. The program's premise is that designing can be understood as a process of making and following rules for the selection, position, and dimension of built and space elements. CKB operates at two distinct levels of design: the technical system designer, who makes the rules, and the end designer, who lays out the material and space elements to make a design. CKB supports two kinds of rules with constraint based programming techniques: grid and zone based position rules, and assembly rules that position elements with respect to one another. The paper discusses the rationale for CKB and describes its implementation.

Damski, Jos� C. and Gero, John S.
1997, An Evolutionary Approach to Generating Constraint-Based Space Layout Topologies

CAAD Futures 1997 [Conference Proceedings / ISBN 0-7923-4726-9] M�nchen (Germany), 4-6 August 1997, pp. 855-864

This paper describes a system to produce space layout topologies for architectural plans using an evolutionary approach. The layout specification is defined as a set of topological and directional constraints, which are used as a fitness function in the evolutionary system. The halfplane representation is used to represent the genotypes in the evolutionary system, for both arrangements of halfplanes and the figures generated from those arrangements. As the halfplane representation proposed here does not distinguish between straight and non-straight boundaries, at the symbolic level the spaces and the layouts produced can also be bounded by straight or non-straight lines. The well known rectangular (polyomino) arrangements become a particular case only.

Li, S.-P., Frazer, J.H. and Tang M.-X.
2000, A Constraint Based Generative System for Floor Layouts

CAADRIA 2000 [Proceedings of the Fifth Conference on Computer Aided Architectural Design Research in Asia / ISBN 981-04-2491-4] Singapore 18-19 May 2000, pp. 441-450

This paper presents the current study of using a constraint based approach to solve floor layout problems. Nonlinear programming technique is used for the solution searching. This paper presents the authors' attempt to improve the nonlinear programming techniques for floor layout problems. Unlike most nonlinear programming systems, multiple optimized solutions can be provided with this system. The process of solving a layout problem, from constraint specification to solution searching, is described in detail. A case study is given in the last section before the conclusions to illustrate how the proposed model works.


Flemming, Ulrich and Sheng-Fen , Chien
1995, Schematic Layout Design in SEED Environment

Journal of Architectural Engineering -- December 1995 -- Volume 1, Issue 4, pp. 162-169

This paper describes SEED-Layout, a module of SEED that supports the generation of schematic layouts of the functional units specified in an architectural program. SEED-Layout provides capabilities that allowdesigners to generate and evaluate rapidly different layout alternatives and versions; to explore the trade-offs involved; and to engage generally in an iterative, highly explorative design process. The resulting"design space" is complex, and the paper describes current efforts to provide designers with intelligent "navigation" aids that encourage them to explore interesting portions of this space without "getting lost."The paper concludes with a brief description of the current implementation and directions for future work.

Flemming, Ulrich and Woodbury, Robert
1995, Software Environment to Support Early Phases in Building Design (SEED): Overview

Journal of Architectural Engineering -- December 1995 -- Volume 1, Issue 4, pp. 147-152

This paper describes the overall goals of SEED, the approach taken by its developers to achieve these goals, and the subprojects that comprise the entire project. SEED aims at providing computational support forthe early design phase in all aspects that can benefit from such support. It addresses specifically architectural programming, schematic layout design, and the generation of a fully three-dimensional configuration ofphysical building components like structure and enclosure. These tasks are handled by three individual modules, SEED-Pro, SEED-Layout, and SEED-Config. A standards processor is under development tosupport standards and code checking in any module, as is an object database to store and retrieve different design versions, alternatives, and past designs that can be reused and adapted in different contexts(case-based design). Usability issues, especially the interfaces to the modules, receive special attention. Subsequent papers elaborate on these efforts in greater detail. The present paper provides an overview of theentire project and introduces shared concepts presumed known in subsequent papers.

Aygen, Z. and Flemming, U.
1998, Classification of Precedents - A Hybrid Approach to Indexing and Retrieving Design Cases in SEED (a Software Environment for the Early Phases of Building Design)

CAADRIA �98 [Proceedings of The Third Conference on Computer Aided Architectural Design Research in Asia / ISBN 4-907662-009] Osaka (Japan) 22-24 April 1998, pp. 435-444

An efficient indexing of past solutions is crucial to case-based design (CBD) systems performing complex retrieval on large case-bases. This paper suggests a hybrid approach to the indexing and retrieval of design precedents. The suggested approach accounts for the issues of classification manifested in architectural discussions on type and CBD literature. The indexing scheme integrates description-logic based representation for classification and an object-based representation for precedents. The hybrid scheme constitutes a basis for the implementation of a generic case indexing and retrieval mechanism for SEED.


1986. Flemming, Ulrich, Robert F. Coyne and Timothy J. Glavin, et al
ROOS1 -- Version One of a Generative Expert System for the Design of Building Layouts

17 p. : ill. Pittsburgh: Engineering Design Research Center, Carnegie Mellon University, September, 1986

ROOS1 is a generative expert system for the design of building layouts. The system is intended to complement human designers' performance through (a) its ability to systematically search for alternative solutions with promising trade-offs; and (b) its ability to take a broad range of design concerns into account. Work on the system provides insights into the applicability of Artificial Intelligence techniques to space planning and building design in general. The system is based on a general generate-and-test paradigm. Its main components are a generator, a tester and a control strategy (which is to be expanded later into a genuine planner). The generator is restricted to the allocation of rectangles. The spatial relations above, below, to the left and the right are defined for pairs of objects in a layout and serve as basic design variables which define differences between solutions and govern the enumeration of alternatives. Within the class of layouts it is able to produce, the generator is completely general and able to generate all realizable sets of spatial relations for a given number of objects. In contrast, the tester is domain-specific and incorporates knowledge about the quality of layouts in a specific domain. The system can be applied to various domains by running it with the appropriate tester and, possibly, the appropriate control strategy. The control strategy itself mediates between planner and tester and, when expanded into a planner, is able to streamline the search for alternatives. The system will go through a sequence of versions with increasing complexity. Each version will have a conceptually clean and clear architecture, and it is the authors' intention to evaluate each architecture explicitly in terms of its promises and limitations with respect to various domains. The first of these versions is described in the present paper

1988. Flemming, Ulrich, Robert F. Coyne and Timothy J. Glavin, et al
A Generative Expert System for the Design of Building Layouts -- Version 2

Artificial Intelligence in Engineering: Design. editor. John J. Gero. Elsevier (Computational Mechanics Publications), 1988. PP. 445-464 : ill. includes bibliography

The paper describes an attempt to increase the intelligence of a CAD system by adding capabilities (1) to systematically enumerate alternative solutions to a design problem, and (2) to take, at the same time, a broad spectrum of criteria or concerns into account. These capabilities are intended to complement the designer's abilities and performance. In connection with such attempts, fundamental problems arise when the objects to be designed have shape and are located in space. These problems are identified, and an approach to solve them is outlined. This approach is currently being tested over a range of domains all of which deal with the design of layouts of rectangles subject to constraints and criteria. The search for alternatives takes place in a state space with properties that make it possible to systematically explore and evaluate the power of various search strategies or planning paradigms. The state space is established through a domain-independent generator, while the evaluation of points in that space is carried out by a domain-dependent tester built up through a process of knowledge acquisition familiar from work with expert systems

1989, Purnomo, H.
SPACE - Generative Expert System: An Expert System for Designing a Layout of Single-Family Houses Using the Expansiva Building System [MBdgSc Thesis]

Department of Architectural and Design Science, University of Sydney UNPUBLISHED. CADLINE has abstract only.

This dissertation describes an expert system for designing the layout of a single-family house using the Espansiva building system introduced by Jorn Utzon. The expert system uses two systems that are already available; the BUILD expert system shell as an automated reasoning system and the Eagle 3D modeler system for producing graphical output. Both programs run under the UNIX operating system on SUN microcomputers. The integration of BUILD, which is written in Prolog, with Eagle using one of the Eagle commands called 'weasel' is a major part of the implementation of the system.