Matchmaker : an instrument for matching demand for andsupply of buildingAdams, W.; Daru, R.

Published in:Open House International

Published: 01/01/1996

Document VersionPublisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the author's version of the article upon submission and before peer-review. There can be important differencesbetween the submitted version and the official published version of record. People interested in the research are advised to contact theauthor for the final version of the publication, or visit the DOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ?

Take down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Download date: 21. Aug. 2018

•

•

Reierences :

Akin, Ö. (1982) Introduction, in Akin, Ö. and E. Weinel (eds.), Representation and Architecture. Information Dynamics Silver Spring, Maryland, pp. 1-26. '

Bly, S. (1988) A Use of Drawing Surfaces in Different Collabo­rative Settings, Proceedings of the Second Conference on Com­puter-Supported Cooperative Work, Portland, Oregon, Septem­ber 26-28, 1988, pp. 244-249.

Carrara, G. Y. Kalay and G. Novembri (1990) A Computational Frame"':'ork for Supporting Creative Architectural Design, Evaluahng and _Predzcting Design Performance : Proceedings of the 3rd Internahonal SUNY Buffalo CAD Symposium, Buffalo, New York, March 23-25, 1990, pp. 17-33.

Coons, S._ (1963) ~n Outline of the Requirements fora Comput­er-Alded Des1gn System, Proceedings of the Spring Joint Con­ference, Detroit, Michigan, May 21-23, 1963, pp. 299-304.

Coyne, R. and A. Snodgrass (1991) Is designing mysterious? Challengmg the dual knowledge thesis, Design Studies 12, no. 3 (July 1991), pp. 124-131.

Ehn, P. (1988) Work-Oriented Design of Computer Artifacts. Arbetslivscentrum, Stockholm.

Greenberg, S. (1991) Computer-supported cooperative work and groupware : an introduetion to the special issues, Internation­al Joumal of Man-Machine Studies 34, no. 2 (February 1991), pp. 133-141.

Habra_ken, N. J. and M. Cross (1988) Concept design games, De­szgn Studzes 9, no. 3 (July 1988), pp. 150-158.

Kuhn, T. (1970) The Structure of Scientific Revolutions. University of Ch1cago Press, Chicago.

Lawson, B . . (1990) How Designers Think : The Design Process Demystified. Butterworth Architecture, Boston.

An Outline of the Requirements 30

Markus, T. (1969) The Role of Building Performance Measurement and Appraisal in Design Method, in Broadbent, G. and A. Ward (eds.), Design Methods in Architecture. Wittenborn, New York, pp. 109-117.

Minneman,. S. and S. Bly (1991) Managing à Trois : A Study of a Mulh-User Drawing Tool in Distributed Design Work, Hu­man Factors in Computing Systems : Reaching Through Technol­ogy, Proceedings of the CHI '91 Conference, New Orleans, Lou­isiana, April 27 to May 2, 1991, pp. 217-224.

Olson, M. ~nd S. Bly (1991) The Portland experience : a report on a d1stnbuted research group, International Joumal of Man­Machine Studies 34, no. 2 (February 1991), pp. 211-228.

Schön, D. (1983) The Reflective Practitioner. Basic Books, New York.

Schön: D. (1985) The_ Design Studio : An Exploration of its Tradi­tzons and Potentzals. RIBA Publications, London.

Schön, D. (1988) Designing : rules, types and worlds, Design Stud­ies 9, no. 3 (July 1988), pp. 181-190.

Simon, H. (1981) The Sciences of the Artificia/. MIT Press, Cam­bridge.

Tang, J. and L. Leiff~r (1988) A Framework for Understanding the Workspace Activ1ty of Design Teams, Proceedings of the Sec­ond Conference on Computer-Supported Cooperative Work, Port­land, Oregon, September 26-28, 1988, pp. 244-249.

Winograd: T. and F. Flores (1987) Understanding Computers and Cognztzon : A New Foundation for Design Addison-Wesley Reading, Massachusetts. '

Author's address :

University of Michigan Doctoral Program in Architecture Ann Arbor, MI 48109-2069, USA.

Open House International Vol. 21 No. 1, 1996

Matchmaker - An instrument for matching demand for and supply of buildings

Wim Adams and Roei Daru

Abstract:

To match supply and demand of buiidings various approaches are possibie. In this paper the well known, on SABA aigorithm based, methods are assessed and compared to a graphic interactive approach. In a tooi, deveioped by us, using this approach strongiy developed power of human visuai information processing and pattern formation, is used in collaboration with the computer power of Jast and clear visualisations support­ing for or giving feedback and consequences on decisions taken. To eiucidate the working and idiosyncrasies of this tooi a case concerning architecturai heritage is discussed.

1. Vacancy of monumental buildings:

In Europè"'(in the EEC and in other countries, in particular middle European countries) a great number of manurnental buildings, protected or not, will become vacant in the coming years. To avert the worst consequences of their obsolescence (be it technological, organisational, politica!, economie or cultural), it is necessary to shorten the time of vacancy as much as possible, and at best to eli.mi­nate vacancy altogether.

On the other side, there are a number of insti­tutions or companies looking for premises in the cramped inner areas of European cities. Many of those are willing to move into existing building stock, if the building fabric can be adapted to their needs.

To match supply and demand while at the same time respecting the constraints imposed by cultural heritage, it is necessary to bring them to­gether and to effectuate feasibility studies in the shortest possible time. This will sametimes occur in delicate circumstances, in particular when preserva­lion depends on the proof that a building has an economical chance of survival. The feasibility study should be served by tools allowing the various partners to communieale on the level of the match between supply and demand, translated in terms of spatial organisation and building constraints.

2. Problems with matching:

The matching of a building and an organisation can be seen as a specific space planning problem in

Open House International Vol. 21 No. 1, 1996 31

which characteristics of a demanding party (an organisation with a housing problem looking for new or existing appropriate premises) are compared to the actual supply (spatial characteristics of exist­ing or newly designed buildingsin stock). In match­ing organisations with existing buildings (as com­pared to designing a new building) we are con­fronted with some specific problems. Spatial con­structions of existing buildings are more or less rigid as compared to newly to design bu ildings. Each building has its own spatial structure and spatial characteristics with both fixed and more flexible parts which make this particular building more or less suited for a specific organisation.

In the process of matching supply and demand it is necessary to take correct decision regarding the match within a short space of time. Next to actual fitting organisational functions to spatial units of the building in a generated layout it is, in order to analyse and evaluate the proposed geography of the building, also necessary to easily retrieve the bottie necks or discrepancies given this specific arrange­ment of organisational functions over the spatial units of the building. To match supply and demand of buildings, various approaches are possible.

3. Approaches to the matching problem:

While interest in artificial intelligence ap­proaches to the matching of buildings and organisa­tions is growing and use of more conventional algorithmically based methods is most widely spread and therefore favoured b y some, these methods, in generaL have some significant draw­backs. The imperfections concern a severe restrietion of available information on both build ing and organisation that is actually used by these pro­grams, the transparency of the processing of this information as well as the quality, flexibility and format of the generated output, the proposed build­ing layout. We think that a less "heavy" approach can be more cost and time efficient and solve some of the fundamental problems encountered by the two other types of methods. Although the tool we developed (ROP : the Dutch abbreviation of Space Optimisation Package) was originally developed for other problems, because of their alikeness to the abovementioned matching problems we expect this

Matchmaker

tooi to be potentially suitable for these quandaries as well. We tested this by applying the tooi in an actual case.

4. A graphic and interactive reordering tooi : ROP

4.1 Background of ROP Development:

In the past years, our research unit at the Eindhoven University of Technology has developed and tested a graphic-based reordering tooi and applied it to space planning problems of large governmental buildings, both existing and new.

This graphic-based reordering tooi in fact is a generic tool, and therefore can be used for a large range of ordering problems for weighing, ranking and evaluating objectives, to have a synthetic view of the combined basic preferences and differences of the involved parties as for example in a jury wise evaluation and ranking of alternative proposals.

The proposed tool is the computer based, graphic and interactive version of the data and association matrices. The well-known paper version of particularly the association matrix (see figure 1) for a long time has been recommended for the inventory of existing relationships between functions or activities of an organisation in the preliminary phases of design. However, as long as these instru­ments could only draw static information on paper, they were much too ineffectual and found little real application in practical space planning problems.

4.2 Functioning of ROP:

The graphic based reordering tool consists of two components; one registering and displaying relationships between entities of a single set (in building design entities can be the activities or spaces of a building, in an organisation the func­tions or activities of employees), the second contain­ing a record of relevant criteria for each of these entities (in building design a nurnber of charader­istics of the activities or spaces as e.g. constructive, climatological or layout properties; in an organisation the brief with its specific requirements of functions or employees). Scale of measurement of these criteria can range from nomina! to ratio scales. In ROP there is no restrietion as to the nurnber of relevant criteria that can be inserted into the ma­trices. The registered data are translated into a graphical format and subsequently can be rear­ranged interactively.

The developed tooi is connected by sub-routines to a computer aided design package, within which the grouped entities of the matrices are translated

Matc.'unaker 32

into plans and attached data bases.

The experience gained by us in the past years corresponds for a large part with the needs outlined in the problem definition. However, the heavy constraints imposed by cultural heritage as ad­dressed in this paper ask for a redesign and elabo­ration of these tools.

5. Case : Applying ROP to a monumental court house:

In this case the reordering tooi was applied to matching some general facilities (as a restaurant with its employees, meeting room facilities and a post office) to a monumental court house.

5.1 Registration and optimisation of organisational and building information :

5.1.1 Registration :

The first phase is to describe the organisation (the demanding party) as functional units which can be made corresponding with spatial units. The prescription of spatial needs and additional require­ments, also called the brief of an organisation, can take place in both quantitative and qualitative manners. In our tooi the initia! numerical or Bool­ean values expressing organisational requirements are translated into sirnple graphical symbols as dots or squares, whose sizes are in accordance with the original val u es they represent, and restored . in the cells of respectively the triangular association ma­trix (revealing existing relationships between all distinct functions of the organisation) and the rect­angular multi-criteria matrix (exposing all relevant

Functions, organisation parts

D central postroom

R central postroom

D chef restaurant

R chef restaurant

D kitchen

R restaurant

D restaurant

D large meeting room

R large meeting room

D small meeting room

R small meeting room

D waitresses

R waitiesses

R kitchen

Area, m1 Reiations

. Fig. 1 Matrix of relationships between organisational functions.

The values in the cells indicate the strength of the relationship with higher values repcesenting strong relations.

Open House International Vol. 21 No. 1, 1996

requirements, Wishes and ~~~~~~~~~~~~~0~~~~-i=n=Uo=n=~~d=•=h~i=n~n=n=~=m~o~~~~~~~~~~~~~~~ preferenCeS Of the fundiOnS FuncUonal properties Assodation matrix

.. . as to the offered space). In this case there were some fourteen entities with, next to the area requirements, a total of sixteen criteria con­sidered to be of importance in generating a layout (see

c: ë 5 r ·~ Ë ij :§ .§ -0 ?:-0 .~ ë - - -~ g. ·g § .~ ~ .2 ~ .g >- ~ ·~~§- I

· · · ~:Z~1:gfiy 414i>. -9 9 9 ·~ 1 ·~ ~ ~ ~ ta ~ ~ ~ ~ c ~ ~ ~ 0.. ~ ~ :c c ~ .§ .~ tiO t 2,0 ~ .~ j ~ j Û ~ ct ct t3 ê ü tt :t' J: ~ :> è5 Function Area o - N M ..,.

figure 2).

number of criteria involved. : • • • • . •• ; •• ; • • ~:~~~;~ses u~--·- · ·:- ~::.:: ::-~~: :~x Mostly, only area require- )--------------.:__ ______ __: _________ _ mentS Of the fundions and R-kitchen 14

1 D-<:entr~l-posrroom z R l 0 ------ 0 -centra -postroom

the relationships between _,

D-large-meeting-room 8

them are considered in the creation of a layout. With these programs it is not pos­sibie to include other infor­mation that has a possible or k.noWR consequential ef­fect or is even decisive in the generated layout. Store­houses demanding a specific construction, while construc­tive characteristics cannot be ~--------------------------------------------~ incorporated or computer Fig. 2 rooms that are not allowed to be placed in the southern (sunny) wing of a building while the program is not equipped to include information on required or

Organisational data from the brief appear in random order. On the right in the topmost part of the figure, the association matrix shows the interrelationships of functional units. Larger dots express stronger relationships. On the left, the functional units are scored according to functional properties. The list of functional properties is accompanied to its right by a bar chart expressing required areas. In the \ower part of this figure, an alternative view of the relationships is seen as 'walking lines'.

preferred location of functions. While a number of relevant criteria is of a more general nature, other criteria are more specific to the present-day match­ing problem. Especially with matching of monumen­tal buildings some constraints and charaderistics will be unique for this situation. It is therefore essential that the program is flexib le as to the number and type of criteria attended to.

A second major drawback of the algorithmic programs is that there is no feedback on quality and correctness of the input. In the generated lay­out the area of a room is, for instance, ten tirnes as large as it should be because of a typing error. Not every mistake in the input has to be directly discernible in the resulting layout, although it did have effect. Because of the graphic format in which the information is presentt~d in ROP, incorrect in­put is immèdiately noticed and can be easily cor­rected.

Open House International Vol. 21 No. 1, 1996 33

Parallel to descrihing the organisation in a tri­angular relationship matrix and a reetangwar multi­criteria matrix, we also describe the monument in spatial units and distance relationships between these spatial units. The distance relationships be­tween these spahal units can be generated directly within the matrix, but it is much easier, more re­alistic and self evident to generate this automaticaUii.. from the draughted plan. The possibility to do t1W is part of the existing tooi (see figures 3, 4, 5 appearing on next page).

Registration and typing in the constraints and relevant charaderistics originating from heritage preservation objectives, expressed in levels of authorised intervention can also be clone directly into the rectangular matrix of the reordering tooi or retrieved from the draughted plan. In order to be able to match organisational constraints to authorised building interventions the databases of

Matchmaker

Original plan of the monument Spatial connections through corridors

Fig. 3 Fig. 4

Original plan of the monument drawn in a CAD program. The various spaces in the building that can be used for the spatial arrangement of functional units of the organisation are numbered. The courtyard, nr. 12 could, after roofing in, also be used to place one or more of the functional units. Spaces with numbers 16, 18 and 13 and 17 are actually located on an entresol floor above the spaces with respective numbers 8, 14 en 7, 11, 15.

Lines connecting the various spaces. The lines are chanelled through the existing corridors of the building. The lengths of these lines therefore correspond to the actual 'walk.ing' distances between spaces.

to such, mathematically correct, reconfigurations. As the matrix is re drawn on the computer screen after this movement, immediate visual feedback is given and the effectiveness of the manipulation can di­rectly be assessed. In this method, complex cogni­tive processes in the analysis and interpretation of information are thus operationalised into tasks of

both organisation and building have to be compa­rable; they have to comprise the same relevant criteria. When these building interventions can be expressed at the level of the spatial nnits we get two data-matrices of the building that are homoge- Characteristlcs of the monumenta1 building

neaus W ith those of the Functlona1 properties ~--~--------------------------------------------

Associ~tion matrix

organisation. The triangular ma tri x revealing rel a tion­ships between spatial nnits of the monument, the rect­angular matrix containing their properties and con­straints regarding alteration in construction. Both matri­ces are also presented in a graphical format, with small graphical symbols in the · matrix cells (see tigure 5).

5.1.2 Analysis :

While analysing the data in ROP, the layout of the information can be rear­ranged by moving a se­lected part of the matrix, containing one or more of its rows and columns, to a new position. The matrix construction lends itself well

Matchmaker

c - ~ 00 Q ~ ~'~~ .S ~ ~ ~ § § ~ ~ ·; -~ ti~~cu.,:c2 ..!! 1

-o_ g"e .Zv;.2·~c:o ?:~ g, '*'*'* -~'*:0~2-ë~>.~~ ~ B; öj 'ij öj ~ "i ·-a ·-& ~ 2 ë ~ ·~ t ~ ~ .~ 1

~ 5 j Û 5 ~ ~ 8 6 Ö ~ :Ë' ~ ~ $ Q 1 Function • 8 • • • • e e e . , e . •space- 1 • e • • e e e e • •space- 2

e e e ·• e e . e . e • . 'space- 3 e • • e • e • e e • 1 space- 4

. e e e e e e e e e e e e . :•pace- 5 e e e • • e e e • e • • e • 1 space- 6 e e e • e e e • • e • • 0 • 1 space- 7 e • e • • • e • e • • e • e 1space- 8 e • • • • • • e e e •• space- 9 • • • • • • • • • • a e • e 1space-10 • • • • • • • • • e • • • • e • •space-11 • e e e • e e e•space-12

e e e e e • e e • • e e e•space-13 • e • • • • • • e e • e • • 'space-14

• e e e e • e e e • e 'space-15 • • • • e • e e • • •' space-16 • e e e • • e e • • • :space-17 • • e • e • e e • • •,space-18 • • e e • e • e e • • 1 space-19 • e e e e e e e • • e ,space-20

Fig. 5

· The manurnental building expressed in a matrix format. In lines 1 to 20 the different spaces of the monument are enumerated. On the left hand part some 16 functional properties of the di­verse spaces are given (larger dots expressing higher scores on the properties). In the associa­tion matrix, on the right hand part, the size of the dots is an indication of the actual distance between spaces in the monument (see figure 4). The largest symbol corresponds to a ·walking distance' of less than 20 metres, the middlemost symbol with distances between 20 and 40 metres and the smallest dot expresses a distance. of 40 up to 60 metres. Empty cells indicate that spaces are further apart than 60 metres.

34 Open House International Vol. 21 No. 1, 1996

visual discrimination of the Organisational data, clustered individ ual graphic elements, f------F-u-nc-ti-on-a-1 -p,-o-pe-rt-ie_s __ _.:::. ___ _::.:..::::_=~::..::....------A-ss_o_ci_at-io_n_m_a_tr_ix---1 the recognition of graphic 1--------------.,__------~----------l pattems and interactive pat­tern forma ti on, tha t are highly developed in people (see figure 6).

The tooi herewith offers the possibility of an interac­tive clustering of entities, with a cluster reflecting dif­ferent entities of comparable spatial needs and/ or addi­tional requirements. As this rearrangement is clone manually, the user deter­mines or has to determine the importance of the differ­ent criteria involved. Devia­tions from obvious actions (actions generating a coher­ent pattern) not only pro­vide in1ormation about the subjective importance of those criteria but also can or have to be elucidated in the

R-smaii-meeting-room 140

I D-c~ntra1-postroom ------,__..!a2 R--central·postroom

D

R-large-meeting-room 12

D

D-large-meeting-room 11

a 3 D-chef-restaurant

a 4 R-chef-restaurant

as R-waltresses

a a R-kitchen 7 diSCUSSiOn between the Vari- D-restaurant 9 0 0-kitchen 8

ous partners involved. With ·Fig. 6

algorithmic programs, the The same data as in figure 2, but clustered. Items 13, 12, 14, 5, 7 and 6 now form a cluster of user has no or little insight functional units with strong interrelations as expressed by large dots in the triangular matrix, as on the weight or effect of are items 1, 2, 3 and 4 respectively. The areas bar chart is supplemented by shadings expressing the different criteria. functional grouping (see also figures 7 and 8). The qualilalive matrix is rearranged according to Weights are fixed or have to scores to obtain qualitative profiles. We can for example see that the units 1 and 2 have a similar

profile and form a qualitative group, as do units 3 and 4. Within the larger group ranging Erom be determined in advance item 13 to item 6 we find a subgroup 'restaurant' with items 7 and 6 and a 'kitchen' group (14 and intermediate changes in and 5), each with its own requirements. By rearranging columns, we find groups of requirements weights are difficult to ac- with similar profiles concerning functional units. The graph expresses the obtained spatial con­complish or not allowed. centration of functional groups.

When allowed the effect of changes in weights is not directly visible but only in the final layout plan. As matching a (large and complex) organisation with a (extensive and complex) building can be a prolonged process it is necessary to be able to examine and discuss intermediate results. Small, hardly notiçeable, errors or slightly unfavourable al­locations occurring in an early phase of the matchmaking process often can have far reaching, dramatic consequences on the final floor plan lay­out. As the layout is generated manually in ROP, corrections can be made immediately while the subroutines in the CAD program provide the user additional information as to adequate allocations for a specific fnnction.

Thus, with the electronk version of this instru-

OpeD House International Vol. 21 No. 1, 1996 35

ment visualisations can not only be used for pre­sentation of summarising irtformation at the end of the processing and analysis of data, but also to commnnicate, explore and support decision making throughout the analysis.

5.2 Matching:

In the next step, the actual matching takes place, where clusters of functions, or individual fnnctions are assigned to the available spatial units (see Figures 7 and 8).

In this step the actual positions, properties and constraints of existing spaces in the monument are compared to the optimised and clustered spatial needs of the end user. This phase of the matchmaking process is, as the previous ones, clone

Matchmaker

'

e

e

.. -- _,,

Imported functional units, expressed as areas various parties, obtained by these negotiations.

Unes 1, 2

=

R-waltresses D-waitresses R-kitchen D-kitchen

Lines 3, 4 Lines 13, 12, 14, 5, 7, 6

~~:~~~-iet-restaurant R-chef-restaurant

13 12 14

5

Fig. 7

Unes 8, 9, 10, 11

With algorithmic programs the format of the output often is unclear, as e.g. when the resulting arrangement is pre­sented in a numerical form; lists of spatial units with their functions or inhabitants. This format certainly is not suited to infer practical implications as distances and walking routes between related functions or to bring forth functions residing in adjacent spatial units. The out-

Functional units as imported into the CAD program from the matrix program, diagrammati­cally expressed as areas from the brief, with corrections introduced by constraints of the monument (spatial grid, depth of building). The four groups (each with its own shading) correspond with the functional groups in figure 6.

put is also impractical and not realistic when functions are e.g. completely enclosed by other functions and not hordering the facade of a building, when

manually. This offers the possibility to include other considerations than those directly deducible from the registered facts . Next to the intrinsic inforrna­tion, the registered information laid down in the matrices, the person in charge of the space planning can also take extrinsic information into account, information conceming building and organisation that is known by the planner although it is not comprised in the data matrices. The planner often

storehouses are not located in the basement of the building but on the main floor or when spaces are jagged or the length-width ratio is out of range. The generated layout is not flexible when changes in the contiguration of spaces (exchanges) or small adap­tations are difficult or impossible to attain.

In the last phase, the feasibili ty in terms of space, building fabric and costs can be appraised by

already has a coarse layout in mind, partly based on previous experience and partly on his communication with persons being well known with both building and institution. This expertise he can neatly add to or fit into the present-day matching problem with its particulars and specific require~ents.

Plan of the monument with spatial arrangement of functional units

There are various ways in which to view the consequences of the matching, one of the most interesting being the discrepancies' views, a display again based on the graphical matrices. This view obtained after evaluation of the generated layout, in a clearly structured way, reveals all deviations between demand and supply both regarding re­lationships as well as concern.ing addi­tional spatial requirements (see figure 9). It therewith enables comni.unication and stimulates negotiation between in­volved parties. This leads to rearrange­ments, possibly with some modifi­cations of the original input from the

Matchmaker

==-·~"A-'""~~

Fig. 8

Plan of the monument with a first spatial arrangement of functional units within the building. Some functional units had to be split to fit, while some existing spaces are unoccupied.

36 Open House International Vol. 21 No. 1, 1996

c:: -~ ~ "' ·a .. ~ 0

-~ ... -;; E c::

-~ ~ V 0 U)

"' <

c:: -~ -;; "' ·a .. ~ o-.

..... a 0 ·-a "' c - ::1

~ u.

-N~·~~~~~~~N~.~~~ . ---------

..

ee 88 22 ==~~ g_s~~ ~~ ë ~tE ~ E~ ~ 5 ~ ~~~~ -~ -~~~~ 1'\'1'\' ' ' -':' -':' ':' Cl o:: Clo::O::ClO::ClCl

•

e eeeeee ~eeeee2 öoöoOo0o 0o 0oOo c cccccc

-~ ~ &r~ ·t -~ ·t eeeeee e cV cV :::==== C.OC.OfO"l"l f'l:l fQ :;;:;; eeeee 66~666~

]... I § uoqe:>o[-a311S3Jd • • • • • e e e e e • • • • ;; OZIS-a8qSOJd • • • • • • • • ••••• e e e e -~ UO!l0JOJOP·JD-(0h"l • • • • • ••••• e e e e ~ '" "!A • • • • eeeee• • • • ~ JOJIUO> l 1?tDTIJ • • • •••• e e e e e e e

0 OU013ÁH • • • • • ·•••• • • • • • • ~

~ ÁJ1?A!JJ • ·••····· • e ..... ............ "t; 8 UTUOIPunJ·jD-(0h"l • • e e •• • • • • ••••••• ~ UO!l1?1UêJ!JQe e e e • • • • •• e e e e e e e 0 Álp•d • ::>•.. • ••••••••••• ~ Á!Ojes-an:I• • • • • • e e • • . . . . -; a;,ueuaJlf!1?1.D·Jo-JaAêYJ• • • • • e e • • • • • •

c:: ~uawd!nb;)-jQ-fêJAêYJ • • • • • • e e • • • • • • -~ ÁI•J•s-ÁJe[3rnge • • • • • • • • • •

~ ooueqmlS10• e • • • • e e • • • ~ UO~JO.USUOJ e e • • • • • e • e •

Open House International Vol. 21 -No. 1, 1996

-~M~~~~~~e=~~~~~~

l 1" " ' ""1" " " ," " " '"":" " '" "~" .··.···:·-,-. : ·: ">:1 ë -illt±fi~ltl ! ... ~ E

~ • l ~ .. ~

•i -: ' ' ) ~-1 c - - - - - - - - - -- -- - ~' t ' U)

- NM~ ~~~~~ ~= ~~~~~~ <

+ eee et: e e

ee 8888888 88~ë ~~~~~~~ -~ ~ ~~ ~EEEEES c::

c &~~~~~ ëë ~~~iti~ ~ E~~iillcc ~ ~~~eeee~ E ü~~~~ ~ ~~~~~~~===== = C2- ·~ = -vv -- ~~··•• • C:: ~~~~~ ~~~~~ ~~~~~~~~ OE ó~ ó~~ó~óó~óó~óóó~ -0 .

~~~N -~o- ~~M~~ o ~ ~ ·~

§ ~~~ ~~~~ ~~~~~~~~b ~ ~ ~~~~~~ ~ ~~~~ ~ ~~~~~~ c E a..a..a..a..a..a..a..a..a.. a..a.. a..a..a..a.. a.. a.. ;:J ~ ~ ~ ~ ~ ~ ~ ~ t ~ ~ ~ ~ ~ ~ _~ ~ ~ ] ~rnJ~.~~~ • .[]]~ -;

uoqe:>o(-aS!I S<Ud • § az!s-aS!ISêJJd 't:

UO!~ eJOJêJP·JO·jêJAêJl

MêJ~I\.

(O~UOJ-reW!JJ

OU0!8ÁH • •

....... ·• . . ··-.

••• • Á:Jl?/\!Jd

8u!'JO!lJunj· jD-(aMl ..... ... . . .

uop1?JUa!JQ • • • • Á!Ded •::>

ÁIOfOS·all:l • • êJ)U1?UêJlU!1?\D•jQ-fêJAêJ1

lUOWdmbi>-jO•fOh"l ÁIOfOS·ÁJe[8mg • •

oou eqm !SIQ e e

•••• ••••

. ... •• ...

uo qJru!SUO::J • • !. • ~ - Jll • .__ <

uoqe:>o[-a8!1S"'d • • • • • • • e . e e e e • • • e a-z!s-aSqsaJd • ••••• • e e e e e e • • • e

U0~1?JOJap-JO·fêJAêJ1 • • • • • e e • • • • • • •

MêJ!I\. • e • • ••• e • • • • ••

[OJ1UO:>-leUl!l::Je ••• • • •• e •••••••• Olla!8ÁH • • • • • • •• • • •• ÁJ"'!ld •••••••••••••••••

8UfUO~unJ·J<>-f<MêYJ . • • • • • • ••• e e • e e e •

UO!le lUOllQ • • • •. e •• e •. Ál!oed•::> • ••••• • ••• • ••• • • e

ÁlafOS·OJI:I • • •• e e e e e • •• e •• e e a;,ueua1U'!1?W·Jo-JaAa1 • • e e • • • • • • • • • • • • •

lUOJWd!t).ba-jo-J~MêYl e e e e • • • • • e • • • e ÁIOJOS·ÁJe[8mg • • •• e e . . . • . . . . . . .

oou eqmlS!O e e • • • • e • • • e • uoqon.QSUO::J • • e e • • • • e . e e e e e • .

37

c:: ::1 --0

>< "f .. E

"' .. 'ü c:: "' "' .. o..::: .. ... .. .. V 0.. -~ 0 Q ~

ë .. E ::1 c:: 0 E U) .. t: .. 0.. 0 .. 0..

-;; c:: -~ ti c:: ::1 ""

~ <.n i- QJ I Q)

..9 . ~ ~ ~ -5 .~ -5 >, ::1 0 u c::.9"- '.:::c:: 5l E ta ..._ u ·- <.n 0 u EO.. o c::c:: "' '"' Q) :s ..._ ..._

;s:: <11 -~ .... ~ 0 ...

"';; =] ~ .~ ~ ..-:..v;ta .c c.n .t:: o ~ o E .~ Q) ~ .9 ·t; o..è;' ~ !; E <11 x.~ c ~ 1)1) ~ :;

- ta - c ·-f: ~ fr c: ·- ~ ~ 0.. ..9 '"'" Q) ::g ta "'0 u <11 ·- 0.. c:: -mg :; ~ _g~ ... § ·= Q) ~ Ln

·.::: 0.. g: ..0 ~:a "' ~ ~ ·~ .~ -~ Q) ~ 2 ;a = ~ -~ ~ Ë O Q> ~ ~Q).Z.~

;:. O"'~ _c .c<l.l -~ ·.::::: ..._ u - 1)1) >-. !:: ;a 0 -~ >-. 'i: .2 "' ::~ "' -c.aEo.. E '" ~ <11 -c o E "' -c o ..<: <l.l -<11 QJ C: .... - '"'"Ö C: 'ü ta 0 ~ .. ~ ..0 Q)

;~! ~0':; ~ Q.. ~ - ;<.n c: ~ ~ ·s;: ~ E .~ ._g ~ u -;;: .-::: tao :; o c::' e ~ ·g :; <11v E lll~ ~ E . =>=>=' aJotaCC() ·=~o.c c.. ....... e ta u- o "'0 :::: aJ c: aJx<lJ c: ~ -5 ] - .z 'i: -5 ~ &~ 8~"'0-:; -;at-... ~.c <11<11 c:: E .c"' (1) - ~ ·t: tac.n -E

·0 ~ .c ~ ~ -~ ~ ~ c: <.noooo~<lJ-C!_ ~ :.2 Cl. -~ Cl ~ ~ Q) 0.. 0 - ·- <1J ~ <1J ._Eo,.u'"'"' .C ~ta -C: ~~ c:Z'

0\ ~ ~ -~ .g ~ ;; ..g ~ bÎl

ti <11 c:: ' <11 <11 - E <11 -= .:2 "2 ~ -= -~ 0 -= E ö.. 8 ..._ 1"3 ::::

·- u .-:... 0 0.. 0 "'0 o~~ c: c.n..oc:

..!::: ~ "'0 -~ )( Q) - -= c: Q) <1J t; 'i: .z ..... c

-~ "'0 Qj -~ -:; ..._ [j -- 1Vi C::E 0 E t .g- - ..2 tib <1J ~ ~ "E u E u ... :> ·.::: § o ;c o <11o :.;:::=E "''_ .... ..._ ~ta c: ö ta 0 ::::: :s c:J ·-1 ..0 2 -~ -; o- ..<: -c - Q) '2 !: :s ~ : ~ "' ..<: ::1 .. o- L.. c --·- - E r- .a o..c:: -m <11 .-:. ><E 0 0 ·.:::: c: -5 c: 'i: 0 - ... ~o o - u Q) - c.. ·.:::: c: ·- ta

..c § (/) -~ ·- m e ~ - ·= <1J ~ "'0 -~ Q) ta c:: ,. .c "'<l.I C::;:.>, 0 -~ 1:- "' ~ .. ·.::: <11

.c::...:. "' ::l e.o ... .c: -~ ~ë ..:- --;; 0 ::= r--C::""' QJ .C ;>..._ ta . ::~ o E -~ <11 o ::1 ~ _..._:1 '"'" Q)~ O"~ "' o c:: o.. ... -- <11 E ·= c.noota t:..c::=' e!.'2 E.::::-2 ~ - c:: "' ::1 .... c:: ·- 0 .s 0 <11 ~ o .:2 § ... E

..<: c:: !ll - 0.. c:: ~ .9 ·2 -~ --;; (Ü ~ ~ 0 ~ ::1 c:: ·.::: c:: 'öi:l -- c.n - tataO ..._

.~ '2 ~ ~ ~ '-E 2:! 0 - ta ·- 0 c ta <.n "'t>eS <11<11 ::1--~ ""'-.c..c..._c-] o ~--..:- <1J~ ·- Q) · - ..._ ..._ ..._ E c.. : -5 -:;oo~ ::::s e ". .... o.. ;c"' o.. C:: o.. 0> 0 <.n .c .~ .9: 0 -<l.l ~t>e"' .-:::~ E,.

=ö <11 .s -~ -;; c:: <11 § -c ·;;: Vi :I 0 ..<: ·­·- ·x <1J 0" ·= - ti E .... <11 -5 "' c:: c::

0 "0~ ·-::1 ~(:~.9 ~~<1:1 ..._ -:; 'ö :: bO -~ bO ·2 .::" ..... 0.. 0 .s ~ ~ :I~

Matchmaker

e

e

analysis of the type and importance of the discrep­ancies and the consequences of their pursued solu­tions. Once a compromise has been attained, pre­liminary proposals can be designed and laid down in terros of drawings. The spatial designs can then again be translated into matrix views and evaluated, including the comparison of renovation costs.

6. Summary of the comparison between ROP and algorithmic methods :

Some principal differences between our graphic interactive approach and methods that are based on algorithms can be summarised as follows :

Most of these methods use the so called SABA algorithm. Distinguishing a more coarse stacking and a more refined blocking phase in the genera­tion of a layout these methods automatically gen­erate a layout, while in contrast in ROP a layout is generated in a manual and interactive manner.

In ROP the user can base the decision he makes partly on extrinsic information, facts that are not recorded in the brief of an organisation or the description of the building but are nonetheless known by the user. Algorithmic approaches per definition only use the information that beferehand has been recorded into the program. In addition, the user often has no insight in the manner in which the algorithm weighs the different criteria.

Because of ·the manual generation of a layout, in ROP the user has control during the whole process and can correct or give preferenee to spe­cific solutions. With the automatic generation of a layout, control of the process is mainly determined by the algorithm, with the user having no or only a very restricted pos'sibility to interfere during the process. The planner normally can only make some manual changes in the stacking or zoning of func­tions after the stacking phase is finished.

In order to support the user in decision-mak­ing during the process, our method is centered on a graphic and clearly structured presentation of the information; other methods mainly present informa­tion in a numeric format.

Another supportive aspect of our approach is the immediate feedback of (the consequences of) achons taken by the user; algorithmic approaches on the other hand often only present the final results (layout) at the end of the process.

In a study by Li (1993) in two case studies (in

Matchmaker 38

the first a design was made for a new kindergarten, in the second an organisation was matched to an existing twenty story building) the working and obtained results of ROP were compared to those of SPACE, a program in which the generation of a layout plan is based on the SABA algorithm. In the second, matching case, it was found that the ob­tained geography (both ln stacking as well as in blocking) in ROP was significantly better than those obtained in SP ACE, as based on distance between related functions . These results are in accordance with those found by Scriabin and Vergin (1975).

7. Additional options of the graphic interactive reordering tooi :

The ROP package has a flexible design appear­ing from:

In the graphic reordering tooi, information can be presented irt different views, all having their own specific graphical format. Next to the triangular matrix of relationships, the connections between entities can e.g. also be presented in a bubble dia­gram or organogram. We found that different us­ers prefer different formats while optimising the datasets. The program can easily be extended with additional formats wished for.

The tooi saves all actions taken by the user. This creates the possibility to return to previous configurations of entities and to analyse the order­ing process itself. In our project this ordering pro­cess of individual users in an experimental context indeed is stuclied for development of additional features and an increased potentiality and user­friendliness of the tooi.

The matrix of relationships can be automatically searched for the non-planar graphs K3,3 and KS. Planarity is a prerequisite for the direct translation of a graphical matrix into a ground plan. A non­planar graph can be solved by bridging over or tunnelling under crossings, introducing intermediale nocles at crossing points (as e.g. a hallway) or changing the original pattem of relationships which caused non-planarity.

A genetic algorithm is going to be implemented which generates "intelligent" solutions of both rect­angular as well as triangular matrix. These solutions in the near future can be presented as new steps in the process of the analysis. The user will keep the possibility to refrain from this supportive op­tion or to make changes to the options presented by this genetic algorithm.

Open House International Vol. 21 No. 1, 1996

Once the organisation is described in the format of the graphical matrices it can be matched to nu­merous buildings. The same is true for the build­ing, once described, all potentially interested organisations can be compared or matched to it.

9. Condusion :

The graphic interactive approach to matching buildings and organisations has a number of advan­tages as compared to existing methods, such as more control of the user over the process, clearly structured views beneficia! to communication be-

Because of the modularity of the CAD compo- tween ·partners involved and revelation of subjec-nent in the tooi, with additional macro's, floor plan . .. ·five preferences. Case studies with ROP however layouts can be evaluated for concurrent purposes as have shown that because of an often too large in determining "space syntax" axiality maps, deter- degree of freedom of the user or p lanner in the minatien of (shortest) routes between spaces, ere- decision making process, additional "guiding com-atien of duality graphs, analysing layout on possible puter support" is to be recommended. problems conceming orientation, etc.

8. Disadvantages and deficienci~s ~f ROP :

One of the major drawbacks to the program is that it requires attention and active involvement of the user throughout the complete process. It is not possible to insert the data, push one button and piek up the results after a little while.

A second drawback in the matching is the sometimes too large degrees of freedom a user has. Although feedback is given after allocating functions to spaces it would be helpful if the user would be given additional support in the creation of floor plan layouts, especially before making actual deci­sions regarding the best allocation for a specific function or the sequence in which to place the different functions. In a number of matching experi­ments we noticed that users often interchanged already placed functions or even reshuffled large parts of the settled configuration during the match­ing process. Advice on e.g. placing functions with heavier constraints first, as regarding the involved building, possibly would reduce retracing one's steps during the creation of a building layout. A possible future extension, helping to reduce reshuf­fling, could be to create a function which highligh ts best still available spahal units regarding specific criteria considered important (as e.g. which of the still empty rooms is at least twenty square meters, has power currency, and scores at leas t four on privacy).

Bibliography:

Adams, W. and S. v. Kol. Gebouw-indelingen met behulp van MiniCad+. VCA-Mededelingen, 1991, november, pp. 20-22.

Adams, W. Supporting decision making processes: A Graphical Interactive Approach. PhD thesis, Eindhoven University of Technology, 1991.

Daru, R. Evaluation by graphics. In: A. Faludi, H. Voogd (ed.): Evaluation of complex po licy problems. druk. Delftse Uitgevers Maatschapij, Delft, 1985, pp. 89-107.

Daru, R. and W. Adams. Interactive graphic heuristic procedure, a technique for the salution of matrix problems. In : F. Kimura, A. Rolstadas (ed.), Computer Applications in Pro­duction and Engineering, Elsevier Science Publishers, Amsterdam, 1989, pp. 783-807.

Daru, R. Beyond algorithms and visualisations. In : L Beng Siong (ed.), Second IES Information Technology Conference on Ad­vances & Development in lnformation Automation, Singapore, Institution of Engineers, 1991, pp. 133-140.

Daru, R. On-Orthodoxe Facility Planningstechnieken. (in Dutch) VCA-Mededelingen, 1991, oktober, pp. 8-15.

Li, J. P. ROP and Space: comparison of the building layout pro­grams w ith two case studies in architectural d esign. Eindhoven University of Technology, Institute for Continu­ing education, 1993.

Scriabin, M and R.C. Vergin. Comparison of computer algorithms and visual based methods for plant layout. Management Sci­ence, vol. 22, no 2., 1975

Authors' Address:

Eindhoven University of Technology, Department of Architecture, Building and Planning, Eindhoven, The Netherlands.

---···8····---

Open House International Vol. 21 No. 1, 1996 39 Matchmaker