ADAPTIVE URBANISM:
A PARAMETRIC MAPS APPROACH
Jernej Vidmar, Janez Koželj
University of Ljubljana, Faculty of Architecture, Slovenia
Znanstvena revija, št. 3 / leto 2015
Univerza v Ljubljani
Fakulteta za arhitekturo in
Fakulteta za gradbeništvo in geodezijo
Ljubljana, 2015
Naslov revije:
Scientiic journal, no. 3 / Year 2015
University of Ljubljana
Faculty of Architecture and
Faculty of Civil and Geodetic Engineering
Ljubljana, 2015
Title of the Journal:
IGRA USTVARJALNOSTI THE CREATIVITY GAME
teorija in praksa urejanja prostora Theory and Practice of Spatial Planning
Urednici: Alenka Fikfak, Alma Zavodnik Lamovšek
Uredniki tematskega dela: Cristian Suau,
Saja Kosanović, Carmelo Zappulla
Oblikovanje in naslovnica: Gašper Mrak
Lektoriranje: Mojca Vilfan
Prevod: Mojca Vilfan
Klasiikacija: (UDK) Renata Stella Čop, (DOI) Teja Koler Povh
Založila in izdala: Univerza v Ljubljani,
Fakulteta za arhitekturo in
Fakulteta za gradbeništvo in geodezijo
Spletna stran revije:
http://www.iu-cg.org/
Spletna stran številke
http://www.iu-cg.org/paper/2015/iu03.html
ISSN 2350-3637
Editors: Alenka Fikfak, Alma Zavodnik Lamovšek
Thematic section editors: Cristian Suau,
Saja Kosanović, Carmelo Zappulla
Design and Title page: Gašper Mrak
Slovene text proofread by: Mojca Vilfan
Translation: Mojca Vilfan
Classiication: (UDK) Renata Stella Čop, (DOI) Teja Koler Povh
Published by: University of Ljubljana,
Faculty of Architecture and
Faculty of Civil and Geodetic Engineering
Journal's Web Page:
http://www.iu-cg.org/
Currrent Issue LInk
http://www.iu-cg.org/paper/2015/cg03.html
ISSN 2350-3637
Št. 3. / 2015 IGRA USTVARJALNOSTI – teorija in praksa urejanja prostora | THE CREATIVITY GAME – Theory and Practice of Spatial Planning
Jernej Vidmar, Janez Koželj:
PRILAGODLJIVI URBANIzEM: pristop s parametričnimi kartami
ADAPTIVE URBANISM: A Parametric Maps Approach
DOI: 10.15292/IU-CG.2015.03.044-052 UDK: 711.4:004.9 1.01 Izvirni znanstveni članek / Scientiic Article SUBMITTED: September 2014 / REVISED: May 2015 / PUBLISHED: October 2015
IzVLEČEK
UVODNIK
EDITORIAL
ČLANEK
ARTICLE
RAzPRAVA
DISCUSSION
RECENzIJA
REVIEW
PROJEKT
PROJECT
DELAVNICA
WORKSHOP
NATEČAJ
COMPETITION
PREDSTAVITEV
PRESENTATION
DIPLOMA
MASTER THESIS
Velike okoljske spremembe in tehnološki razvoj, ki povzročajo hitre in nepredvidljive spremembe, silijo naša mesta v reorganizacijo in prilagoditev na vseh
nivojih. Sodobna mesta postajajo vse bolj dinamična in odprta za prihodnje spremembe, katerim pa tradicionalni operativni instrumenti regulacije
zazidave (npr. OPPN/zazidalni načrt) ne morejo slediti. Manjka jim prožnost
in odzivnost, ki bi lahko sledili hitrosti in nepredvidljivosti sprememb. Zato
potrebujemo bolj prilagodljive metode načrtovanja mest. V tem članku
predstavljamo novo metodo načrtovanja in oblikovanja, ki bi lahko izboljšala
običajne instrumente regulacije zazidave. Predlagana metoda temelji na t. i.
parametričnih kartah, ki omogočajo odprto, prožno in odzivno načrtovanje in
oblikovanje mestne zazidave. Parametrične karte preoblikujejo regulacijska
določila zazidave v neposreden (interaktiven) prostor rešitev (ang. solution
space), znotraj katerega je možno oblikovati in ovrednotiti množico veljavnih
variant zazidave. To omogoča prilagodljivo načrtovanje mest, odprto za
spremembe v prihodnosti. Da bi ocenili predlagano metodo, smo razvili interaktivno prototipno aplikacijo. Predhodni rezultati kažejo, da lahko z uporabo
parametričnih kart, ki opisujejo regulacijske pogoje zazidave, izboljšamo
proces urbanističnega načrtovanja in oblikovanja. Nakazujejo tudi, da bi
parametrične karte lahko dopolnile običajne urbanistične dokumente na
način, da ti postanejo bolj prožni in odzivni, kar bolje ustreza potrebam v
načrtovanju in oblikovanju sodobnega mesta.
KLJUČNE BESEDE
prilagodljivo, parametrično, urbanistično načrtovanje, urbanistično oblikovanje.
ABSTRACT
Immense environmental changes and technological advancement,
which are causing rapid and unexpected changes, are forcing our cities to
reorganize and transform at all levels. As modern cities are becoming ever
more dynamic and opened for future changes, the traditional operative
instruments of development regulation (e.g. master plan) fall behind. They
lack the lexibility and responsiveness needed to follow the speed and
unpredictability of changes. Thus, more adaptive city planning methods
are required. In this paper, we have presented a novel planning and design
method that could enhance traditional instruments of development regulation. The proposed method is based on parametric maps, which enable
open-ended, lexible and responsive planning of urban development.
Parametric maps transform urban development regulations into direct
(interactive) solution space within which a myriad of valid urban design
alternatives can easily be created and evaluated. This enables adaptive city
planning opened for future changes. To evaluate the proposed method,
we have developed interactive prototype application. Preliminary results
show that by using parametric maps to describe development regulations,
urban planning and design process can be enhanced. They also indicate
that parametric maps could complement conventional master plans to
become more lexible and responsive, which better responds to the needs
of planning contemporary cities.
KEY-WORDS
adaptive, lexible, parametric, urban planning, urban design.
44
Jernej Vidmar, Janez Koželj: PRILAGODLJIVI URBANIZEM: PRISTOP S PARAMETRIčNIMI KARTAMI, 44–52(148)
IGRA USTVARJALNOSTI – teorija in praksa urejanja prostora | THE CREATIVITY GAME – Theory and Practice of Spatial Planning No. 3. / 2015
1. INTRODUCTION
Rapid technological advancement, which afects nearly all aspects of our
lives, has been changing the way we live. It causes unforeseeable economic, social, and environmental changes that afect our cities, which are
consequently being reorganized and transformed. Due to uncertainty and
complexity of this process, cities »evolve in ways that are diicult to predict
and … city developments, even when planned, tend to ind patterns of
organisation that were not previously deined in the plans.« (Beirão, 2012,
p. 35). To make things worse, cities are also afected by high inertia. This
is the result of long-term efects of spatial decisions, which are ultimately
relected in physical form. Once built, transport infrastructure and buildings stay there for a long time. The form of traditional master plans, which
represent the inal instrument of development regulation, has become too
rigid and overdetermined to cope with the uncertainty and dynamics of the
changes cities face. Schnabel & Karakiewicz (2009, p. 94) state that it is »too
precise, too prescriptive, ... not making allowance for changes.«. »Conventional urban scale master plans lack the lexibility required to account for
ad-hoc changes and informal developments. They are static projections
of a possible future.« (Henn, 2014). Fixed master plans, which determine
speciic design in detail, thus hinder development of the city; they need to
be frequently amended and updated using complicated, expensive and
time-consuming procedures. To confront this intricate problem, we need
to develop new kind of urban planning instruments, which will improve
lexibility and responsiveness of development regulation (Šašek Divjak,
1999). We need adaptive urbanism, which will help cities stand up against
erratic changes that come quickly and unexpectedly. As observed by Ho
(2011, p. 5), we need master plans that will »maintain openness, allowing
for creativity instead of limiting«.
Although we have already accepted the fact that contemporary city
structure should be oriented towards unpredictable changes and continuous alternations, which take place quickly and unexpectedly (Čerpes
et al., 2001), we have not yet been able to deal with this complex issue
completely. One approach to mitigate efects of unpredictable future is to
implement one of several alternative planning systems (e.g. performance-based planning or form-based codes), which emerged over the last half of
the century (Goldstein, 2004; Hirt, 2007). These systems were developed to
provide a framework that is more lexible and responsive than traditional
zoning approach, yet with the same level of certainty (Steele & Ruming,
2012). Nonetheless, it seems that the municipalities tend to avoid changing
their planning systems due to many reasons (e.g. high migration costs, education of users or potential legal issues). If they decide to implement some
form of the alternative planning systems, they tend to do so only partially
by mixing it with the existing system (Elliott, 2008). Another way to address
this problem is to unify technological capabilities and practice methods of
urban planning and design (Pitts, Farley, & Datta, 2013). For this, »we have
to search for a new paradigm in the way of modelling urban form ... from
ixed types and pre-determined shapes of elements of the model and to introduce a concept that will be generative as much as it is analytical« (Billen
et al., 2014, p. 72).
This search has already begun in the past decade. Researchers around the
world are striving to develop new (digitally-based) urban planning and
design methods and techniques that enable more lexible and quicker
response to unpredictable changes in space and time (Batty, 2013; Beirão,
Duarte, & Stoufs, 2011; DeVries, Tabak, & Achten, 2005). Rather than
designing a inal solution (master plan), we should design its control
system (Verebes, 2013a) and let the solution gradually evolve itself through
time. According to several authors (Canuto & Manuel, 2010; Steinø, 2010;
Schumacher, 2013), parametric approaches seem to be the most suitable
to fulil this task. However, when exploring state-of-the-art case studies
(e.g. Halatsch, Kunze, & Schmitt, 2008; Verebes, 2013b; Aydin & Schnabel,
2013), one can easily observe that with most parametric urban planning
and design approaches – especially generative ones – users are expected
to interact with computer by some form of programming techniques (e.g.
scripts or visual programming). This may represent insurmountable problem, since most urban planners and designers are no programmers. Such
approaches thus require a team of programmers (or at least designers with
advanced computer literacy), which can only be aforded by the largest
and/or enthusiast practices. In addition, open-ended master plans relying
on programming techniques usually involve a lot of programming work
to provide only one-time solutions. Thus, we should aim to develop more
general and intuitive parametric planning and design methods.
In this article, we propose a novel operational urban planning and design
method based on so-called parametric maps, which interactively regulate the
form of development. They represent an instrument that directly connects
separated boundary conditions of development to its form. We argue that by
using the parametric maps method, we can establish interactive solution space within which numerous alternative urban designs can be created based on
the same (ixed) development regulations. This, when implemented properly,
can transform traditional static master plan to become open-ended system,
which is more lexible and responsive for future changes. New method allows
for: 1) quick and transparent creation of many equivalent design alternatives
under the same development regulations and/or 2) quick adaptation of
urban master plans according to changed conditions in (parts of) the city.
2. THEORETICAL BACKGROUND
To understand how the proposed method works, we irst need to explain
the diference and conlict between the two basic principles of contemporary urban planning and design practice. Local authorities (municipalities) traditionally set out the general urban plan on a top-down principle
(Pissourios, 2014), which regulates city development using 2D maps and
text documents. By deining a set of development regulations, they actually
create boundary conditions that determine the solution space within which
urban development can be designed. Each building of the development
must comply with this set of constraints (development regulations) in order
to ensure its coherence with the entire urban tissue. On the other side, urban designers take the bottom-up approach as they shape the actual (3D)
urban space by placing, spacing and grouping buildings, one by one. Here
they observe and evaluate spatial efects of the form and spaces between
buildings they are creating. The design process is thus much closer to how
45
Jernej Vidmar, Janez Koželj: ADAPTIVE URBANISM: A PARAMETRIC MAPS APPROACH, 44–52(148)
Št. 3. / 2015 IGRA USTVARJALNOSTI – teorija in praksa urejanja prostora | THE CREATIVITY GAME – Theory and Practice of Spatial Planning
cities actually evolve; as Batty (2012, p. S9) pointed out: »Cities ... evolve mainly from the bottom up as the products of millions of individual and group
decisions with only occasional top down centralised action.«.
What we are dealing with here is a complex process that takes place between two (apparently opposite) sides: on one side there are many objective
and measurable (quantitative) planning parameters and on the other side
there are more speciic and subjective (qualitative) design criteria. Although
both sides seem diferent, they are inseparably connected. As is the case
with urban planning and urban design, which are essentially the same
(Gunder, 2011). Urban designers have to develop their spatial idea in compliance with all the requirements given in the land use plan, set of other
spatial planning documents, laws, standards, and norms. These quantitative
requirements set out the boundary conditions that deine solution space
within which the urban designers need to establish high quality relationships between multitude of buildings to form a whole development. Since
there is much data involved, it is time-consuming and arduous task to
harmonize both quantitative and qualitative aspects of urban design. Use
of computer tools is thus inevitable.
Inability to efectively connect these two approaches results in rigidity of
traditional master plans. We argue that by using parametric maps conventional master plan can be enhanced to become interactive instrument, which
can better react to future needs and desires of the city. This is possible, as
development regulations in general urban plan usually do not prescribe
exact shape of development. They only set out the rules. If these rules are
described properly (e.g. using parametric maps), they can be used to actually propel the creation process instead of limiting it.
UVODNIK
EDITORIAL
ČLANEK
ARTICLE
RAzPRAVA
DISCUSSION
RECENzIJA
REVIEW
PROJEKT
PROJECT
DELAVNICA
WORKSHOP
NATEČAJ
COMPETITION
PREDSTAVITEV
PRESENTATION
DIPLOMA
MASTER THESIS
Figure 1: Metric and parametric dimensioning. Buildings are traditionally modelled using standard
dimensioning (left), while parametric approach allows for more informative dimensioning using
building’s end values (right).
rey building with a 400 m² built-up area (Figure 1). The advantage of this
approach is obvious, since urban designers have more time to explore
various possibilities directly within the desired end result, as they do not
need to calculate building external dimensions in order to achieve desired
end values.
3. METHOD DEVELOPMENT
3.1 Parametric maps
In this article, we propose a new, semi-automatic performance-based parametric urban design method that ills the gap between top-down and bottom-up principles of urban planning and design. Using the new method,
one can design urban development directly within the solution space,
using both principles simultaneously. Working inside this solution space,
a move from the design paradigm of ‘form making’ towards ‘form inding’
(Otto & Rasch, 1996, cited in: Oxman, 2008) can be made. The proposed
method is fully adapted to meet the speciic requirements of two-dimensional urban planning and three-dimensional urban design at the same time.
It is based on conventional regulation parameters (e.g. building’s height or
built-up area). Therefore, it can easily be integrated into the common urban
planning and design worklow. Since the proposed method relies on instant visual and computationally intensive information feedback, it should
be implemented as a computer application.
A parametric map represents spatial distribution of development parameters
throughout the development area. It acts as a ield of rules that regulate properties of development. The parametric map can be represented by ordinary
(RGB or grey-scale) bitmap images projected (geolocated) onto the plot
area, where each colour channel represents a selected parameter of the
development, e.g. buildings heights or built-up areas. Since bitmap images
themselves have ixed range of values (typically from 0 to 255), span of
parameters values needs to be deined as well. Span of parameters is used
to map ixed bitmap value to inal parameter value. The building volume
should then automatically be adapted based on inal parameter values at
speciic location (Figure 2). To make parametric maps method interactive, it
should be implemented as a computer application, which provides instant
visual feedback by adjusting building volumes in real-time as they are moved around in virtual 3D environment. This way an open-ended operative
control system of urban development regulation can be established.
Before continuing, we also need to illuminate the diference between the
conventional (metric) and the parametric design in the context of the
proposed method. We see the parametric urban design as a method of
modelling the development using the desired goal values, such as number
of storeys or gross loor area of the building. This represents the main
departure from traditional (CAD) methods, where each building is deined
using traditional metric dimensioning, e.g. the building usually deined by
the overall dimensions of 20 x 20 x 19 m can also be deined as an 8-sto-
Parametric maps establish a mechanism that acts on top-down principle,
as they make sure that all newly placed buildings follow the regulations of
the whole development area. They represent a solution space that adapts
each building’s volume in accordance with its location on site, thus creating
a link between a set of rules and the actual shape of the development.
Once parametric maps are set-up, buildings (of arbitrary loor plan) can be
inserted onto the plot area and adapted (in real-time) as they are moved
around the area. This way the development can be designed based on the
46
Jernej Vidmar, Janez Koželj: PRILAGODLJIVI URBANIZEM: PRISTOP S PARAMETRIčNIMI KARTAMI, 44–52(148)
IGRA USTVARJALNOSTI – teorija in praksa urejanja prostora | THE CREATIVITY GAME – Theory and Practice of Spatial Planning No. 3. / 2015
Calculation example
Because the value of building storeys parameter depends on its location on
the development area, the building location (LBx,y) – vertical projection of
building’s centroid – needs to be remapped from absolute space to parametric map coordinate space (PMx,y). This ensures that the proper values
are used accordingly to the building’s location:
LBx,y → PMx,y
Note that the building’s centroid is used only for picking up the parameter
value at certain location. In case when some of the building’s exterior walls
falls out of zoning area, the user should get an error notiication.
The next step is to ind out what is the ‘raw’ value of the parametric map (V)
at speciic location (PMx,y). Using ordinary grayscale or RGB bitmap image,
this is somewhere between 0 (PMmin = black) and 255 (Pmmax = white):
PMmin = 0, PMmax= 255
location = PMx,y
0 ≤ Vlocation ≥ 255
Figure 2: Example of parametric map for number of building storeys. Each building‘s height is adapted
according to its location – white colour of parametric map represents tall buildings and black colour
represents low buildings.
bottom-up principle without having to worry if the buildings comply with
predeined development regulations – as they actually derive from them!
This enhances the creative process and gives more time to check spatial
efects of diferent design alternatives.
3.2 Modus operandi – building heights calculation example
Parametric maps are ordinary bitmap images projected onto the plot area,
where the value of each colour channel represents selected parameter of
the building at speciic location. To keep things simple, we will explain how
parametric maps work based on a single parameter – number of building
storeys; other parameters, with the exception of land uses, also work in the
same manner.
We start with the assumption that the data about the site in question has
already been collected. Based on development goals, regulations and site
survey, which are usually carried out in advance, one can already decide
where the buildings should be high and where they should be low. Based
on this decision parametric map that represents number of building storeys (BS) can be created, where black colour represents the lowest values
(lowest number of building storeys in this case – BSmin) and white colour
represents the highest values (highest number of building storeys in this
case – BSmax), with grey shades in-between:
BSmin ≤ BS ≥ BSmax
The range limitation between 0 and 255 is the result of using standard
8-bit greyscale or 24-bit RGB colour model, which are the most widespread
among computer applications. However, if higher accuracy is needed (e.g.
when deining parameters that have more than 256 values), one could use
a ile format with a higher colour depth (e.g. 16-bit greyscale PNG format,
which provides the range of values from 0 to 65,535). However, this is not
the case here. What is important is that these ‘raw’ values (from 0 to 255)
need to be remapped to represent actual number of building storeys. For
this, parameters span needs to be deined so that it relects the lowest and
the highest values of parameter. In our example, we have decided that the
lowest buildings (BSmin) should be 5 storeys high and that the highest buildings (BSmax) should be 15 storeys high:
BSmin = 5
BSmax = 15
This way the parametric map is instructed to remap the R value of 0 to 5
(storeys) and 255 to 15 (storeys). Any value between 0 (Rmin) and 255 (Rmax) is
automatically recalculated using a simple linear interpolation:
BS = BSmin + (BSmax - BSmin) * Rx,y / Rmax
Since we are calculating end values using linear interpolation, users can
choose which colour (black or white) represents which value (high or low).
An example to calculate number of building storeys at the location with an
R value of 215 follows:
Rx,y = 215; BSmin = 5; BSmax = 15
BS = 5 + (15 - 5) * 215 / 255 = 13.43
47
Jernej Vidmar, Janez Koželj: ADAPTIVE URBANISM: A PARAMETRIC MAPS APPROACH, 44–52(148)
Št. 3. / 2015 IGRA USTVARJALNOSTI – teorija in praksa urejanja prostora | THE CREATIVITY GAME – Theory and Practice of Spatial Planning
The building at the location with R value of 215 will thus be 13 loors high
(13.43 needs to be rounded to integer). When moved around the virtual
model, selected building should be instantly adapted to the required
number of storeys according to recalculated R value at given location. The
same logic can also be applied to other parameters that deine properties
of buildings, such as built-up area or building orientation.
In addition to using parametric maps to regulate physical form of development, they can also be used to deine non-physical properties, e.g.
building’s use based on parametric map of land uses. This is of special
importance for proper calculation of urban control values as diferent land
uses usually call for diferent requirements (e.g. requirement for calculation
of parking lots for residential buildings difers from the one for oice buildings). By using parametric maps of land uses an additional beneit of having
a visual overview of building uses across the area is achieved, as each
building instantly adapts its colour when moved onto another land use in
virtual model. Since land uses are not numeric, they cannot be calculated
using linear interpolation method described above. However, by employing direct RGB-to-land use mapping (specifying which RGB colour value
corresponds to which land use) this problem becomes trivial.
Table 1: List of basic parameters. Parameters in italic are not implemented in prototype application.
Please note that this list can be extended or reduced to it the requirements of certain development/
municipality.
REGULATION (INPUT) PARAMETERS
DIRECT
PARAMETERS
INDIRECT
PARAMETERS
REQUIREMENTS
3.3 Classiication of parameters and their relations
UVODNIK
EDITORIAL
ČLANEK
ARTICLE
RAzPRAVA
DISCUSSION
RECENzIJA
REVIEW
PROJEKT
PROJECT
DELAVNICA
WORKSHOP
NATEČAJ
COMPETITION
PREDSTAVITEV
PRESENTATION
DIPLOMA
MASTER THESIS
In the previous section, we have explained the basic mechanism of parametric maps. However, not all parameters should be applied this way, as
they regulate and control the development at diferent levels. Thus, we
have deined diferent kinds of parameters (Table 1) to structure computer
software algorithms and anticipated user worklow. We have separated
regulation (input) parameters that are used to deine the shape of development directly and control (output) parameters that are used to monitor
the current state of development. In the context of input parameters, we
have identiied three sub-categories of parameters: 1) direct parameters,
2) indirect parameters, and 3) requirements. Each of these parameter categories has some speciics that determines their implementation. Control
parameters are more straightforward, as they just need to be calculated in
order to relect the state of development. Nonetheless, we propose they are
monitored at three levels: 1) for each building, 2) for each spatial unit, and
3) for the whole development.
Direct regulation parameters are used to regulate building volumes directly
at their location in model. They are required to adapt each building directly
as it is placed and moved around the virtual 3D model. An example of
direct regulation parameters are number of building storeys, building’s
built-up area, etc. Direct regulation parameters are independent of indirect
parameters and requirements, as they do not relate to any other parameter and/or calculation. However, they can be related interchangeably (one
parameter can replace another) or interdependently (one parameter afects
another parameter in the same sub-category).
Interchangeable parameters can be best illustrated by connection between building height and number of building storeys, where changing one
will also change the value of another. They both have practically the same
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
Building height
Number of storeys
Gross loor area
Built-up area
First storey height
Other storeys height
Ground loor level
Number of basement loors
Building directions (orientation)
Land use
Absolute min. distance between buildings
Allowed loor area ratio
Maximum lot coverage
Mean number of storeys
Relative min. distance between buildings
Min. distance from parcel boundary
Average size of apartment, oice, etc.
Gross loor area per parking lot /
parking lot per appartment (oice, etc.) /
parking lot per resident (workplace, etc.)
Required green area per apartment
CONTROL (OUTPUT) PARAMETERS
EACH BUILDING /
BUILDING PLOT
■
■
■
■
■
■
■
■
■
■
■
EACH SPATIAL UNIT ■
■
■
■
■
■
■
■
■
■
■
WHOLE
■
■
DEVELOPMENT
■
■
■
■
■
■
■
■
■
48
Jernej Vidmar, Janez Koželj: PRILAGODLJIVI URBANIZEM: PRISTOP S PARAMETRIčNIMI KARTAMI, 44–52(148)
Gross loor area
Floor area ratio
Built-up area
Lot coverage (built-up area factor)
Building height
Number of storeys
Building volume
Required green area
Green area
Green area ratio
Required number of parking spaces
Number of apartments, oices, etc.
Gross loor area
Floor area ratio
Built-up area
Lot coverage (built-up area factor)
Required green area
Green area
Green area ratio
Required number of parking spaces
Number of apartments, oices, etc.
Population density
Mean number of storeys
Gross loor area
Floor area ratio
Built-up area
Lot coverage (built-up area factor)
Required green area
Green area
Green area ratio
Required parking spaces
Number of apartments, oices, etc.
Population density
Mean number of storeys
IGRA USTVARJALNOSTI – teorija in praksa urejanja prostora | THE CREATIVITY GAME – Theory and Practice of Spatial Planning No. 3. / 2015
impact on the building (presuming storeys height is deined). In fact, only
one can be used, but for the sake of better user experience, we left both
options open so that the user can use whichever he or she wants. Relations
between interdependent parameters are a bit more complex, as there are
several diferent ways they can be connected.
Interdependent parameters can be best described based on connection
between number of building storeys, building’s built-up area and gross
loor area where one afects another. When e.g. number of building storeys
is changed, it suices to update only one of the two connected parameters
– either gross loor area or built-up area. For the prototype, we have set the
change of building’s number of storeys to trigger update of gross loor area,
while built-up area remains the same. Of course, it could be the other way
round – when number of storeys is changed, built-up area is adapted to the
size that keeps the gross loor area the same. In prototype application, we
have already predeined these relations as described in results section (see
prototype implementation). Nevertheless, user could also have the option
to select diferent combinations based on own preferences.
Indirect regulation parameters are used to specify performance criteria
(goal values) of the whole development area, spatial units and individual
building plots. Examples of indirect parameters are allowable loor area
ratio or minimum distance between buildings. Indirect parameters can
typically be deined for the whole development area or separately for each
urban block or subdivision of the development. Although these regulations
are not used to deine building’s volumes directly, they need to be speciied in advance as they determine constraints of the solution space. If the
development’s design comes into conlict with these pre-set regulations
(e.g. loor area ratio becomes too high), urban designer needs to be warned
about it immediately. Tool using this method should therefore automatically calculate urban control values and check if constraints are not being
followed, thus enabling designers to work in line with regulation constraints at all times.
Requirements are parameters that are typically used to calculate the quantities and diferent kinds of facilities needed to support the development.
Examples of requirements are average size of apartment (to assess how
many apartments the development will provide) or required green area
per apartment (to calculate how much green area is needed to support
the development). Requirements are typically related to land uses, as each
building use requires diferent values (e.g. number of required parking lots
in residential buildings is diferent than in public buildings, even though
they have same gross loor area). Since requirements do not deine values
that control physical appearance of development directly, they are closer to
indirect than to direct parameters; they afect some of the control parameters that change as the development is being designed.
In addition to regulation parameters, which govern the physical form of the
development, we have also classiied control parameters that are used to
monitor current state of the development as it is being designed. Control
parameters represent the data that is needed to take well-informed decisions during the design phase. Real-time calculation and display of urban
control values should thus be available at all times during the design phase.
They can be calculated in two ways: 1) directly, based on physical state of
development (e.g. loor area ratio) and 2) indirectly, based on requirements
described above (e.g. required number of parking lots). To gain higher
control over any and all parts of development, control parameters should
be implemented for each building/building plot, for each spatial unit and
as a sum for the whole plot.
4. METHOD EVALUATION
4.1 Prototype implementation
To test the proposed method, we have developed prototype application1
(Figure 3) in Maya2 (3D modelling and animation package by Autodesk),
using its script language MEL. Basic usage is very simple: the user must
irst draw a loor plan or choose it from the library of predeined loor plans
(e.g. square, circle, etc.). Once the loor plan is chosen, application generates building mass, including the loors. Once the building is generated,
users can model the development in two ways. The irst approach is to let
every building automatically adapt its volume to the parameters deined by
parametric maps in accordance with its location. When the building is then
dragged around the area, application automatically adapts its volume in
real-time based on the parameter values deined at its speciic location.
The second way is to ‘override’ chosen building parameters, which allows for
even greater lexibility of modelling the development. Once a selected building parameter is overloaded, it is not inluenced by its parametric map any
more. Using this kind of parameter overriding (user has to check parameter),
it is important to note that one is consciously moving away from the outlined
development as deined by regulation parameters. This approach can also be
employed for quick visualisation of diferent alternatives prior to creating inal
parametric map.
Any change in the parameters, whether for the whole area or a single
building, is relected immediately. Set of urban control indicators, such as
loor space area or number of required parking lots is calculated in realtime. We have implemented requirements, which are needed to calculate
indirect control parameters, as a part of land use speciication. This allows
continuous supervision over the entire development and thus promotes
performance-based urban design.
For the need of prototype implementation, we have pre-deined relations
between interdependent parameters as described hereafter. We have
chosen building height – which is the most dominant feature of the building – as our leading parameter, which should not change if not requested
explicitly. From this, we have deined the following relations: 1) changing
building‘s height (or number of storeys) adapts also its gross loor area, 2)
changing building‘s built-up area adapts also its gross loor area, 3) changing building‘s gross loor area adapts also its built-up area and 4) chang1 A demonstration video of prototype application, which shows how parametric maps
method works in real-time, is available at: http://tiny.cc/adaptive-urbanism.
2 http://www.autodesk.com/products/maya/overview
49
Jernej Vidmar, Janez Koželj: ADAPTIVE URBANISM: A PARAMETRIC MAPS APPROACH, 44–52(148)
Št. 3. / 2015 IGRA USTVARJALNOSTI – teorija in praksa urejanja prostora | THE CREATIVITY GAME – Theory and Practice of Spatial Planning
Figure 3: Screen-shots from the prototype application.
ing building‘s storeys height adapts also its number of storeys so that it
matches building height. It does not make sense to deine parametric maps
for parameters that exclude each other (e.g., when the storey height is
constant, building heights and number of loors provide the same building
property, only in a diferent way).
UVODNIK
EDITORIAL
ČLANEK
ARTICLE
RAzPRAVA
DISCUSSION
RECENzIJA
REVIEW
PROJEKT
PROJECT
DELAVNICA
WORKSHOP
NATEČAJ
COMPETITION
PREDSTAVITEV
PRESENTATION
DIPLOMA
MASTER THESIS
As an additional functionality, a prototype tool implements experimental
function to raise error warnings when the spacing between the buildings
is too small. This happens most often due to insuicient solar exposure of
buildings or when the buildings are closer to each other than the minimum
distance allowed.
4.2 Preliminary results
First internal tests of prototype application indicate that using parametric
maps method together with real-time calculation of urban control values
enhances urban design process. Increased speed and design lexibility
allow for more time to verify alternative development proposals. The main
contribution of parametric maps method is that it does not deine development’s end-state, but it rather sets out the interactive rules within which
the development will evolve. Real-time calculation of urban control values
can improve the quality of planned development, as one can take wellinformed decisions during the early stages of the design process. Authors
also observed that the usefulness of parametric maps increases with the size
of the development area, and vice versa. It seems that merging top-down
and bottom-up approaches into a single design method not only facilitates
the early stages of the design process, but it could also contribute to a more
transparent and responsible urban planning and design in general.
When testing the prototype tool we have also observed out that design
and veriication of diferent development alternatives is facilitated as it
enables rapid adaptation of buildings, all within the required criteria. This is
an essential component of contemporary urban design practice, worthy of
special attention. Since the individual buildings are interchangeable (Figure
4), the lexibility of the design, which complies with the planned development strategy, is increased and various alternatives can easily be tested.
Parameters are not used to deine the inal solution, but rather a well
performing solution space within which a designer has to ind the best possible solution also in regard to qualitative terms (Turrin, Stoufs, & Sariyildiz,
2013). This is not to say that anything goes, as working with parametric
maps method still requires experts with solid urban design knowledge.
Prototype application showed that by using the proposed method, a shift
from traditional design towards performance-based design has been made;
this (in terms of urban planning and design practice) means that one can
work with end-goal values and design the development at the same time.
Real-time calculation of urban control values alone can be one of the most
useful improvements to the existing practice, where the development is
surveyed only once complete solution is designed. Using parametric maps
in conjunction with requirements and real-time calculation of urban control
values, several variables can be removed out of the design equation, which
in turn makes it easier to solve.
We observed these main advantages when testing the prototype tool:
■ rapid and transparent design of development alternatives;
■ rapid and lexible response to new conditions;
■ quick assessment of the development area capacity;
50
Jernej Vidmar, Janez Koželj: PRILAGODLJIVI URBANIZEM: PRISTOP S PARAMETRIčNIMI KARTAMI, 44–52(148)
IGRA USTVARJALNOSTI – teorija in praksa urejanja prostora | THE CREATIVITY GAME – Theory and Practice of Spatial Planning No. 3. / 2015
Figure 4: Interchangeability of buildings. Although of completely diferent shape, all three buildings in
this example share the same height and gross loor area; thus one can easily replace another.
■ well-informed decision making;
■ reduced probability of errors.
In addition to beneits, we also found out some shortcomings of the new
method, as implemented in the prototype application:
■ preparation of parametric maps can be time-consuming;
■ errors in parametric maps or parameters span are hard to detect;
■ lack of the tool for direct modelling of the buildings;
■ redundancy of parametric maps when creating small scale developments.
5. CONCLUSION
New urban planning and design instruments should focus on coping with
dynamics of city development. They should be able to steer diverse investment initiatives and development concepts in accordance with long-term
spatial strategy of the city. In this article, we have presented an operational
method that achieves this by connecting regulation and control parameters to the inal form of the development. Prototype tool proved that
parametric maps could establish interactive solution space, which enables
creation of numerous alternative urban designs based on the same development regulations. This way the development can remain open for future
changes until the time of actual implementation (e.g. getting construction
permit). Parametric maps method, which represents operational bridge
between urban control parameters and their spatial distribution, should
thus be investigated further.
When compared to other emerging parametric urban planning and design
tools, parametric maps approach has one big advantage – it provides a
general framework that is simple and intuitive to use. This is a great advantage when compared to many other tools, which were, according to Pensa
and Masala (2014), not yet adopted in practice mainly because they are
technologically too advanced. Parametric maps represent straightforward
and transparent way of specifying parameters values visually which can be
easily understood by everyone.
Current results imply the possibility of using parametric maps as a highly
lexible and responsive instrument to regulate development. Proposed
method can be used for both, the design of urban form as well as its regulation. Parametric maps could supplement (or maybe even replace?) ixed
master plans and bring urban planning and design to a new level, which
better responds to the needs of contemporary city. This will be the main
focus of our further research, where we plan to organise several workshops
to test the proposed method in practice.
As Lemmens pointed out »One of the most positive aspects of traditional
zoning is its predictability.« (2009, p. 127), but it comes at the cost of lexibility. Yet we have observed that predictability of end result is not sacriiced
at the expense of lexibility when parametric maps method is used. However, there is one important issue that we have to solve in the future. Elliot
(2008) noted that performance-based planning system (which we regard
as the closest to the proposed method) was not widely adopted due to the
people’s desire for predictability. Not because the end-results are unpredictable, but rather as they can sometimes lead to too creative (unexpected)
designs, which might not be accepted well by local community. Perhaps
we can avoid these kinds of problems by engaging public in early phases of
design, but we need to verify this assumption before making any claim.
We can use parametric maps to graphically describe quantiiable properties
of the city. From this perspective, we can see them as a kind of city’s genetic
material, which deines numerous (mostly spatial) properties of the city.
Number of parameters directly inluences the size of solution space – the
greater the number the smaller the number of possible solutions. If all parameters are deined (not all are described in this paper) one could actually
use parametric maps to recreate virtual 3D models of existing cities. This
raises interesting question about how many and which parameters should
one deine to achieve optimal regulation level? Which are the parameters
that should be mandatory, which can be optional? This is an opened question we also need to deal with in the future.
51
Jernej Vidmar, Janez Koželj: ADAPTIVE URBANISM: A PARAMETRIC MAPS APPROACH, 44–52(148)
Št. 3. / 2015 IGRA USTVARJALNOSTI – teorija in praksa urejanja prostora | THE CREATIVITY GAME – Theory and Practice of Spatial Planning
ACKNOWLEDGEMENT
Authors would like to thank European Union (European Social Fund), which
has part inanced this research. We would also like to thank to PhD Martin
Vuk and to the anonymous reviewer for the suggestions and comments
that helped us improve the structure and clarity of this manuscript.
LITERATURA IN VIRI
Aydin, S. (2013). CODING SHAPE GRAMMARS: Hints for generating a parametric design tool for large-scale urban
renewal projects. In Cutting Edge: 47th International Conference of the Architectural Science Association
(pp. 177–186). Hong Kong: The Architectural Science Association (ANzAScA).
Batty, M. (2012). Building a science of cities. Cities, 29(6), S9–S16. http://dx.doi.org/10.1016/j.cities.2011.11.008.
Batty, M. (2013). The new science of cities. Cambridge, Massachusetts: The MIT Press.
Beirão, J. N. (2012). CItyMaker; Designing Grammars for Urban Design. A+BE | Architecture and the Built
Environment (Vol. 2). Delft, Netherlands: TU Delft. http://dx.doi.org/10.7480/a+be.vol2.diss5.
Beirão, J. N., Duarte, J. P., & Stoufs, R. (2011). Creating Speciic Grammars with Generic Grammars: Towards
Flexible Urban Design. Nexus Network Journal, 13(1), 73–111. http://dx.doi.org/10.1007/s00004-0110059-3.
Billen, R., Cutting-Decelle, A.-F., Marina, O., de Almeida, J.-P., Caglioni, M., Falquet, G., ˝... zlatanova, S. (2014).
3D City Models and urban information: Current issues and perspectives. In R. Billen, A.-F. Cutting-Decelle,
O. Marina, J.-P. de Almeida, C. M., G. Falquet, ... S. zlatanova (Eds.), 3D City Models and urban information:
Current issues and perspectives – European COST Action TU0801 (pp. I–118). Les Ulis, France: EDP
Sciences. http://dx.doi.org/10.1051/TU0801/201400001.
UVODNIK
EDITORIAL
ČLANEK
ARTICLE
RAzPRAVA
DISCUSSION
RECENzIJA
REVIEW
PROJEKT
PROJECT
DELAVNICA
WORKSHOP
NATEČAJ
COMPETITION
PREDSTAVITEV
PRESENTATION
DIPLOMA
MASTER THESIS
Hirt, S. (2007). The Devil Is in the Deinitions. Journal of the American Planning Association, 73(4), 436–450.
http://dx.doi.org/10.1080/01944360708978524.
Ho, J. K.-C. (2011). Cycle Study as the Basis of Adaptive Urbanism. In F. Neuhaus (Ed.), Studies in Temporal
Urbanism (pp. 3–6). Dordrecht: Springer Netherlands. http://dx.doi.org/10.1007/978-94-007-0937-9.
Lemmens, M. (2009). Re-zoning Alberta: Smart Regulation for Smart Growth. McGill International Journal of
Sustainable Development Law and Policy, 5(1), 115–137.
Otto, F., & Rasch, B. (1996). Finding Form: Towards an Architecture of the Minimal. Berlin, Germany: Edition Axel
Menges.
Oxman, R. (2008). Performance-based design: current practices and research issues. International Journal of
Architectural Computing, 06(01), 1–17. http://dx.doi.org/10.1260/147807708784640090.
Pensa, S., & Masala, E. (2014). InViTo: An Interactive Visualisation Tool to Support Spatial Decision Processes. In N.
N. Pinto, J. A. Tenedorio, A. P. Antunes, & J. R. Clader (Eds.), Technologies for Urban and Spatial Planning:
Virtual Cities and Territories (pp. 135–153). Hershey, PA: IGI Global. http://dx.doi.org/10.4018/978-14666-4349-9.ch007.
Pissourios, I. A. (2014). Top-down and bottom-up urban and regional planning: towards a framework for the use
of planning standards. European Spatial Research and Policy, 21(1), 83–99. http://dx.doi.org/10.2478/
esrp-2014-0007.
Pitts, G., Farley, H., & Datta, S. (2013). Research in practice: An exercise in generative urban massing. In R.
Stoufs, P. Janssen, S. Roudavski, & B. Tunçer (Eds.), Open Systems: Proceedings of the 18th International
Conference on Computer-Aided Architectural Design Research in Asia (CAADRIA 2013) (pp. 685–694).
Singapore: CAADRIA, CASA.
Schnabel, M. A., & Karakiewicz, J. (2009). Rethinking Parameters in Urban Design. International Journal of
Architectural Computing, 5(1), 84–98. http://dx.doi.org/10.1260/147807707780912994.
Canuto, R., & Manuel, L. (2010). Parametric urbanism: emergence, limits and perspectives of a new trend in
urban design based on parametric design systems. V!RUS, 3.
Schumacher, P. (2013). Free Market Urbanism – Urbanism beyond Planning. In T. Verebes (Ed.), Masterplanning
the Adaptive City – Computational Urbanism in the Twenty-First Century. New York, zDA: Routledge.
Čerpes, I., Koželj, J., Blejec, G., Fatur, M., Gregorčič, G., Jankovič, K., Matjašec, D., Simoneti, M., Struna, A. (2001).
Priporočila za urejanje naselij: zaključno poročilo o raziskovalni nalogi. Ljubljana, Slovenia: University of
Ljubljana, Faculty of architecture.
Steele, W., & Ruming, K. J. (2012). Flexibility versus Certainty: Unsettling the Land-use Planning Shibboleth in
Australia. Planning Practice and Research, 27(2), 155–176. http://dx.doi.org/10.1080/02697459.2012.
662670.
DeVries, B., Tabak, V., & Achten, H. (2005). Interactive urban design using integrated planning requirements
control. Automation in Construction, 14(2), 207–213. http://dx.doi.org/16/j.autcon.2004.07.006.
Steinø, N. (2010). Parametric Thinking in Urban Design: A Geometric Approach. In CAAD, Cities, Sustainability :
Proceedings for the ifth International Conference of the Arab Society for Computer Aided Architectural
Design ASCAAD 2010 (pp. 261–270). Fes, Morocco: ASCAAD.
Elliott, D. (2008). A better way to zone: ten principles to create more liveable cities. Washington, DC: Island Press.
Goldstein, J. (2004). zoning Alternatives: Current Trends and New Directions. University of North Carolina.
Accessed on 19.4.2014 at: https://cdr.lib.unc.edu/indexablecontent/uuid:fdd51d84-a226-420a-abe00d9d77849182.
Gunder, M. (2011). Commentary: Is Urban Design Still Urban Planning? An Exploration and Response. Journal of
Planning Education and Research, 31(2), 184–195. http://dx.doi.org/10.1177/0739456X10393358.
Halatsch, J., Kunze, A., & Schmitt, G. (2008). Using Shape Grammars for Master Planning. In J. S. Gero & A. K. Goel
(Eds.), Design Computing and Cognition ’08: Proceedings of the Third International Conference on Design
Computing and Cognition (pp. 655–673). Springer Netherlands.
Henn (2014). Adaptive urbanism. Accessed on 20.6.2014: http://www.henn.com/en/research/adaptive-urbanism.
Šašek Divjak, M. (1999). Guiding urban development on the municipal level. Urbani izziv, 10(2), 163–167.
http://dx.doi.org/10.5379/urbani-izziv-1999-10-02-002.
Turrin, M., Stoufs, R., & Sariyildiz, S. (2013). Performance-based parametrization strategies. In R. Stoufs,
S. Janssen, & B. T. Roudavski (Eds.), Open Systems: Proceedings of the 18th International Conference
on Computer-Aided Architectural Design Research in Asia (CAADRIA 2013) (pp. 519–528). Singapore:
CAADRIA, CASA.
Verebes, T. (2013a). Masterplanning the Adaptive City: presentation. Accessed on 11.9.2014: http://vimeo.
com/74103602.
Verebes, T. (2013b). Masterplanning the Adaptive City: Computational Urbanism in the Twenty-First Century. (T.
Verebes, Ed.). New York, zDA: Routledge.
52
Jernej Vidmar, Janez Koželj: PRILAGODLJIVI URBANIZEM: PRISTOP S PARAMETRIčNIMI KARTAMI, 44–52(148)