Human Activities In KnowledgeBased Architectural Design Systems

1
Human Activities In Knowledge-Based Architectural
Design Systems M. Sherif El-Attar
CAD Research Center College of Architecture and
Environmental Design California Polytechnic
State University San Luis Obispo, CA 93407,
U.S.A. Abstract
Department of Architecture College of
Engineering, Al-Azhar University Cairo, Egypt.
A knowledge-based system consists of two main
components, the knowledgebase and the inference
mechanism. The former contains the problem
solving knowledge, and the later provides the
reasoning mechanisms to control this knowledge in
finding a solution. Control mechanisms generally
depend on the 'structure' and 'type' of design
knowledge represented in the knowledgebase. This
paper focuses on the 'type' of design knowledge
used in computational design environments. The
paper discuss prototypical information in
knowledge-based design systems, and suggests a
functional type of knowledge for usage in such
systems in order to achieve a degree of
generalization in the domain of space design. A
function in an architectural space depends on the
activities that accomplish it. The paper focuses
on human activities and how they may provide the
means to semantically describe architectural
spaces in different building types. Keywords
Knowledge-based design systems activities
context prototypes architectural spaces space
function space type. Introduction In the
domain of architectural design, it is argued that
solutions rely heavily on typological
information, and that there is mental economy in
such solutions (Wade 1977) (Gutman 1972).
Typologies embody principles that designers
consider unvarying as heuristics they allow us
to apply knowledge about past solutions to
related architectural problems. Typologies
according to Rowe (1987) can be divided into
three sub-classes building types as models,
organizational typologies, and elemental types. A
building, or a part of it, may be used as a
prototype in all three regards, serving at once
as a model, organizational type, and elemental
type. Knowledge-based systems acting as a model
of human information processing try to represent
human problem solving behavior in a symbolic
computational environment. To develop solutions
of a problem, the reasoning process depends on
the rich information context of a particular
problem domain. Such knowledge is typically based
on a prototype or a schema, which is a powerful
knowledge-organizing concept (Coyne et al 1990).
A prototype is a generalization of groupings of
elements in a design domain which provides the
basis for the commencement of a design (Gero et
al. 1988).
2
Prototypical design knowledge has to represent
function, behavior and structure of a design1,
and the relationships between these. In a
prototype schema, function is what the prototype
is good for, or the goal to be achieved from it,
behavior describes what the prototype does, or
the expected reactions or responses under a
possible environment, and structure reveals the
parts and their composition, the attributes that
describe the physical features of the prototype.
Behavioral attributes are the key to matching the
physical features of the design (structure) to
function and vice versa. Figure 1 illustrates the
components of a prototype (Rosenman and Gero
1993) (Maher and Zhao 1993).
Recomposition
Decomposition
Figure 1 Components of a prototype schema
Adapted from (Rosenman and Gero 1993)
Prototypical knowledge in the field of
architectural design, usually represents classes
of design objects (e.g. building, space), however
it is also possible to represent another type of
knowledge which relates functions to other
functions and behaviors, in other words, a
functional prototype (Rosenman and Gero 1993),
which may represent a class of functions.
Mitchell (1990) also noted that there is
considerable psychological evidence that
functional classification of objects play an
important role in our thought process, and
suggested that we access our knowledge about the
uses of objects through a functional
classification scheme rather than by form.
Problem and hypothesis Typically, architectural
knowledge-based design systems utilize building
type prototypes as a source of knowledge
(building type, such as school, office building,
etc.). A system based on design prototypes as a
generalization of design elements provides a
framework for storing design experience
incorporating the functional, behavioral and
structural information associated with these
object prototypes. Modeling design knowledge in
this manner alone restricts the system to a
1 A design a any part of it may be represented
by such an organizational schema , (e.g. a
building has functions, behavior, and structure a
parts such as spaces which in tan have their
function, behavior, and structure and so forth
far each of the parts).
3
class of design objects (the type of building),
and at the same time greatly predefines the
structural components of the prototype (in this
case it could be the spaces of the building). To
extend the capabilities of a knowledge-based
design system to deal with spaces in several
building types in a more flexible manner, the
system should use information, which appears in
different building types, and at the same time
has a strong influence on shaping the
architectural space to be designed. I propose
expanding the notion of using object prototypes
that are concerned with a class of an object
(buildings, spaces), to the use of functional
prototypes that are concerned with a class of
functions. This implies that a functional
prototype has instances of functions. For the
purposes of this research the emphasis is on the
functions of 'architectural spaces', that
represent the structural blocks of many buildings
(a decomposition2). Such an argument might not be
fully accepted. However, this approach might
have its benefits in the economy of knowledge
used to design different spaces in variable
building types), system flexibility to work with
different spaces, limiting the predefinition of
spaces in a building type schema, and the
enhancement of the designer's creativity3 (by
analogy). Activities and their implications on a
space design In the domain of space design,
human activities shape the behavior of an
architectural space function, and activities link
the function of an architectural space to its
physical components (i.e. objects to be used for
a function) in a particular context.
Architectural spaces, of a building type, are the
spaces that are either internal or external and
that have a function or utility for the user or
the equipment of that building. In an
architectural space, a function is composed of a
set of activities. Without knowing the activities
that take place to fulfill that function and
their proper sequence of occurrence, the space
can not be properly designed. In practice
activities are usually requested in an
architectural program explicitly or implied
within a known space name (e.g., kitchen) that
has a known functionality. An activity is the
observable act of a human being who is
interacting with others and (or) using any object
within an architectural space (e.g., cooking,
sleeping). Activities may be related to each
other, and may occur in known related sets to
form a function. Similar activities may occur in
different spaces, located in different building
types. However, the requirements of the same
activities in different space types may not
necessarily be the same, due to contextual
differences and also cultural differences. For
the purpose of this research a homogeneous group
of people of the same culture is assumed.
Activities require tangible objects tie,
furniture or equipment), and non-tangible
environmental characteristics that the space
would provide. Non-tangible characteristics of
the surrounding environment are lighting,
thermal, acoustic, and ventilation requirements.
These qualities are provided by the elements of
the space according to the functionality required
from the set of activities. Another influence of
activities on space design is that they relate
some psychological
2 Assuming that a building can be decomposed to
space components each having its functionality.
3 Assuming that this prototypical knowledge
will map a certain functionality to different
spaces in different contexts, this may encourage
the system user to adapt solutions of past
problems or apply the problem solving process of
past problems to solve new ones --
transformational and derivational analogies--
after (Maher and Zhao, 1993).
4
need of people i.e., privacy, personal space, and
territory) to some physical attributes in an
architectural space. á To summarize the previous
discussion, activities of the building users
impose different constraints on space design á
Area and shape of space -- activities, number of
people, and the objects needed to perform the
activity, greatly affect the area and shape of a
space. á Location -- there are different
constraints that restrict the location of
activities in an architectural space depending
on the type of activity and how it is performed.
Environmental conditions -- activities occurring
in a space require suitable environmental
conditions (i.e., lighting, thermal, acoustic,
and ventilation) according to the context of the
function. á Objects -- activities may require
objects (e.g., furniture, equipment) for their
performance, such objects might have their own
behaviors that might influence the description of
a space (e.g., noise and heat emitting
machinery). From the previous observations the
following implications may be drawn á
Activities of a certain function require that the
space provides certain qualities, therefore, it
is possible to deduce a space description from
the constraints imposed by such requirements.
á Activity requirements may be generalized to be
used in different building types with similar
contexts (e.g., auditorium, movie theater,
theater) Activities can provide specific
information about different zones (which
accommodate different activities) in the same
space. á Activities of function can provide
multiple levels of details in describing a space
(e.g. in the design development stage, an
activity such as sleeping may require certain
levels of information, that are not necessarily
required during the schematic stage of design.)
á It is not necessary to predefine all
requirements (e.g., area, lighting levels) of a
space type, instead, it may be more efficient to
describe a space from the collective requirements
of all its activities. In the process of using
activities as the link between a certain function
and the physical components of a space, it is
important to distinguish the "determinants" that
might make the same "activity" similar (or
different) in different building types -- at
least four determinants or factors would be
considered firstly, the function of a set of
activities secondly, the space type which
indicates the context of the activity and the
major function of the space thirdly, the objects
that ate used to achieve an activity and their
nature and lastly, the users of the space. Such
determinants will influence, the representation
and the structuring of information about
activities, the translatability of activities
from one building type to another, and the
combination of activities that generate the
description of spaces in many building types.
5
Investigation of the hypothesis With the aim of
investigating the utility of using a functional
prototype in the description of different spaces
in many building types, an attempt is made to map
a functional prototype to some of its instance
'space functions' in two spaces, and then compare
the activities within the two spaces comprising
the same function (marking the similarities and
differences and how they resulted). If such
similarities or differences resulted from the
same reasoning process, then it might be possible
to use such a functional schema in the
description of spaces, and the issue will then be
how? As suggested previously, a 'space function'
is comprised of several activities in the
following comparison if there is more than an
activity set in the spaces to be compared, I will
refer to it as a sub-function. Issues of
comparison are the activities in the spaces, and
their influence on the layout and some of the
environmental requirements. For the issues
addressed here in this paper the focus is on a
function that can be mapped to a variety of
spaces, the function relates to communication
between two parties which I will refer to it as
"performer - audience". The function is found in
different contexts and serves different purposes
such as education (classroom, lecture hall),
entertainment (theater auditorium, movie,
theater, auditorium, opera-house auditorium),
religious (mosque, church auditorium, synagogue),
sports (stadium), business (conference room,
conference hall) and judicial (courthouse).
There are always two types of users in this
function performer(s) and audience), the
characteristics of the space depend on, the size
of the audience, the activities of the performers
and audience, and the type of performance. The
spaces to be reviewed are a school classroom and
a theater auditorium. In a school classroom the
instructor is the performer, and the students are
the audience, the former is 'teaching', and the
later are learning', both sub-functions (teaching
and learning) are composed of several activities,
as shown in (Figure 2).
In a theater auditorium the roles are clearly
defined, the activities of the performers are
variable depending on the type of theatrical
performance presented on stage, while the
activities of the audience (in this space) are
limited to watching and listening.
6
Comparison The two spaces are divided into two
categories to accommodate different roles tie.
performer, audience), and in both spaces the
number of audience is more than the performers.
While the goals of the two spaces are different
(entertainment, education) and the activities in
both spaces are not identical, the major
constraints on the design of the two spaces are
almost identical, this may be attributed to the
major function 'performer - audience' which
stresses communication (visual and auditory)
between the performer and audience (Figure 3).
The general visual constraints imposed on both
spaces are almost the same, due to the fact that
they ate related by a functional schema, the
differences in the attributes, of the spaces are
due to the activities in the 'instance' of the
function. The 'display size' is different in
both spaces, and resulted from two different
activities, but in both cases had the same
influence on the seating arrangement, and the
distance from the performance. The righting'
constraints have different effects on the two
spaces as a result of the activities and context
in the classroom (reading, writing, and
illustration) requiring homogenous light
distribution, and control on natural lighting,
while the emphasis in the auditorium is Lighting
the stage. The case is also true in auditory
constraints since the activities in both spaces
require acoustic qualities, but the instantiation
in both spaces is different due to the level of
the qualities required, and the volume of space
accommodating different size audiences. In both
spaces the performance' is carried out in a span
of time. As a result, both spaces have the
activity 'sitting', however, the type of
'sitting' is different and utilizes different
types of seats and stating arrangement, due to
differences in the sub-functions that requested
them. In the classroom the learning function
requires sitting and at the same time 'reading
and writing notes' which emphasize a certain type
of a seat with an area for writing, while in the
theater auditorium the audience are sitting for a
longer periods of time and for a different reason
(entertaining) which calls for a more comfortable
type of seating. Other similarities in some of
the design features of both spaces may have
resulted from the building code requirements
regarding the safety of large numbers of people
in one space. Such as facilities for the quick
evacuation of spaces (e.g., width of aisles, door
design, number of seats in a row), fire rating of
materials used, provision of facilities for the
handicapped, and so on. In general, the
similarities resulted from the relations imposed
by the functional prototype, while the
differences in the space attributes are due to
the instance of the functional prototype and the
activities associated with it.
7
Visiual constrains in theater auditorium
Figure 3 Visual and auditory constraints in a
classroom and an auditorium
8
Conclusion The goal of this paper was confined
to the proposition of a knowledge type that can
be used many times in the domain of architectural
space designs, namely a functional prototype. The
issues addressed here are still under
investigation, and their full implications are
not clearly defined at the present time. However,
if it is possible to model such functional
prototypical knowledge, and implement it as a
rich and flexible source of knowledge, this would
achieve a degree of generalization in
knowledge-based architectural design systems.
Bibliography Coyne, R. D., M. A. Rosenman, A.
D. Radford, M. Balachandran, and J. S. Gero
(1990) Knowledge-based Design Systems,
Addison-Wesley Publishing Company, Massachusetts.
Gero, John, Mary Lou Maher, and Weiguang Zhang
(1988) Chunking structural Design Knowledge as
Prototypes, Department of Architectural Science,
University of Sydney, NSW 2006 Australia.
Gutman, Robert (Ed.), (1972) People and
Buildings, Basic Books, Inc., New York. p.394,
400. Maher, Mary Lou, and Pang Zhao (1993)
'Dynamic Associations for Creative Engineering
Design', (in) Gero, John S. and Mary Lou Maher
(Ed.), Modeling Creativity and Knowledge- Based
Creative Design, Lawrence Erlbaum Associates,
Inc., Publisher, New Jersey. Mitchell, William
J. (1990) The Logic of Architecture MIT Press,
Cambridge, Massachusetts. Rosenman, Michael, and
John Gero (1993) 'Creativity in Design using a
Design Prototype Approach', (in) Gero, John S.
and Mary Lou Maher (Ed.), Modeling Creativity and
Knowledge- Based Creative Design, Lawrence
Erlbaum Associates, Inc., Publisher, New Jersey.
Rowe, Peter G. (1987) Design Thinking MIT
Press, Cambridge, Massachusetts. (pp.85-87)
Wade, John, (1977) Architecture. Problems. and
Purposes, John Wiley Sons, Toronto.
9
(No Transcript)