Conference PaperPDF Available

Parametric Urban Data Structuring and Spatial Query - Advanced data mapping and selection methods for parametric modelling environments

Authors:

Abstract and Figures

This paper presents a method for organising urban data inside the CAD environment into a hierarchical structure, which promotes the ease of transferring information between all available urban elements, from streets to buildings passing by the plots and blocks. This is done using parametric methods that map the urban data using the available CAD and GIS records. Finally, the paper presents a couple of example scenarios where such methods are most needed and how much they could facilitate more detailed and complex data to be accessed, compared, and analysed.
Content may be subject to copyright.
Parametric Urban Data Structuring and Spatial Query
Advanced data mapping and selection methods for parametric modelling
environments
Abdulmalik Abdulmawla 1, Sven Schneider1, Reinhard Koenig12, Martin Bielik1, Ekaterina
Fuchkina1
1Bauhaus-University Weimar 2Austrian Institute of Technology
1{abdulmalik.abdulmawla|sven.schneider\reinhard.koenig
\martin.bielik\ekaterina.fuchkina}@uni-weimar.de 1
This paper presents a method for organising urban data inside the CAD environment into
a hierarchical structure, which promotes the ease of transferring information between all
available urban elements, from streets to buildings passing by the plots and blocks. This is
done using parametric methods that map the urban data using the available CAD and GIS
records. Finally, the paper presents a couple of example scenarios where such methods
are most needed and how much they could facilitate more detailed and complex data to be
accessed, compared, and analysed.
Keywords: Data Query, Urban Geometry, Spatial Mapping.
INTRODUCTION
Nowadays, urban designers must operate with more
complex data, increasing their tasks’ difficulty. The
extraction of basic spatial information depends
mainly on the individual’s technical knowledge.
Moreover, as the process that generates the project
data is challenging to document, more project data
are deprived of any guidance to be understood,
reused, or further developed quickly.
Furthermore, Since the advancement of
different computational methods and the spread of
parametric modelling environments, urban
designers have generated many urban data that,
once well organised, could increase efficiency,
reduce the time of repetitive tasks, and save a lot of
resources.
Urban data from GIS to CAD
Geographical Information System (GIS) studies well
summarised the nature of urban data. They are
divided into two main categories, Geometrical
(Locale) and Attribute. Geometrical data is simplified
into three main types: Points, Curves, and Polygons.
At the same time, Attributes are the features carried
inside the geometrical types, like a point carrying
information about a tree’s location and
specifications, as in figure 1.
The structure of information inside the attributes
also matters. For example, a curve geometry could
represent a street with two names with different
lengths. This means that the key “Address” attribute
will have its value as a list of two items with the name
of each part. And if there is another key for the
“Length” of each part, the list of values should also
be in an order corresponding to the names of each
part.
Using this structure, any type of information
could be stored. Even analytical information like the
area of visibility at a specific location. Or a result of
empirical data gatherings like the number of
pedestrians passing at a specific date and time.
Moreover, if all the geometry types in a file are
Volume 2 – Co-creating the Future – eCAADe 40 | 277
numbered, connections could be made between
them by storing their number ids inside each other.
For example, a Block of an ID=5 could host several
Plots with IDs= {110, 111,112}.
Utilising this data structure, generating any GIS
information inside the CAD environment is possible.
But CAD files are not that easy to organise or
maintain. Due to the difference in use for each
software. Since the advancement of Computational
design and the growth of parametric modelling
platforms like Rhino and Grasshopper, the amount r
of information generated inside a CAD file has grown
exponentially. Since Rhino provides a location
dedicated to storing the user information, any
geometry stored in Rhino could follow the same
structure as standard GIS data.
Data Selection (Spatial Query)
Exporting information from a database is called a
query, and since we are dealing with geometric data,
we will call the selection of urban data Spatial Query.
The spatial query is more like an algorithm. It is a
set of rules that define the process of selecting,
manipulating, and extracting the data. This paper
considers data mapping from one geometry to
another as part of the spatial query.
There are mainly two types of Spatial Query:
Attribute-based query and Location or geometrical-
based query. Attribute queries, on the one hand, are
direct searches for features in a GIS database. For
example, extracting a value based on a specific key
embedded with a shape.
Location queries are what spatial queries are, is a
process of asking a question to be extracted from the
spatial geometry data. For example, a simple
distance search like the neighbours of a plot. Or
finding out the number of shops overlooking a street
is a more complex query that requires multiple steps.
Spatial data linkages
With the growth of open geodata, urban designers
have easier access to GIS data supplied by many
sources. Parametric modelling environments like
Grasshopper and a Plugin for the CAD software
Rhino are popular tools. Grasshopper has many add-
ons that handle shapefiles, the most common form
of online GIS data, like reading, mapping, and writing
inside Grasshopper.
But with such data inside a parametric
environment, many data querying tools are still
missing from the GIS software to be integrated
inside Grasshopper. Some of the most basic
methods for spatial data mapping are about how to
understand urban data structuring and Geometrical
relations.
Plots as Link
Berghauser Pont et al. (2019), Lim et al. (2015), and
Vialard (2012) have outlined the importance of the
linkage between the different elements of urban
morphology. They highlighted the importance of
Blocks, Plots, or both in facilitating the connection
between the streets and the buildings.
Bobkova et al. (2019) argue that we must study
urban form to develop and sustain life in the urban
environment. They consider both Streets and
Buildings as the main layers of urban form. Streets
facilitate the accessibility to different urban spaces
Figure 1
Types of Urban
Database (Top)
Types of Selection
methods (Bottom)
278 | eCAADe 40 – Volume 2 – Co-creating the Future
hosted mainly inside the city buildings. In a recent
study, Bobkova et al. (2021) added that some
properties of the plots, like their frontages,
correspond to essential qualities such as the ability
to operate as an interface between street and
building.
This is also highlighted in other urban
morphology studies. Hillier (1996) and Kropf (2009)
have also differentiated urban elements into two
main categories, the buildings and the spaces
between them. And by studying both categories
together, one should better understand how they
affect life in different urban areas.
Modifiable Areal Unit Problem (MAUP)
Openshaw (1984) and Wong (2004) have pointed to
a common problem in any spatial analysis process in
Physical Geography. Since different studies could be
applied to the same area over time, some problems
in the analysis units would need to be addressed.
They collected them under the Modifiable Areal Unit
Problem (MAUP) using two main problems the
zoning effect and the scaling effect.
The zoning effect is when the boundary of the
analysis area changes. Like if a study group selected
another boundary to measure the density of the
same region. The Scaling effect addresses any
change in the unit design or type. As if a new plot
division was added to an existing dataset.
Both problems are related to changes in the real
world and the digital data collected. And since our
control over the real world is impossible, we should
focus more on how to structure the data so that once
any of the effects appear, the data can be easily
adapted.
METHODOLOGY
This paper systematically organises urban data to
access information needed to solve specific spatial
problems quickly. The method consists of a
collection of mapping techniques that record the
data in a CAD file once and then store them in a
location corresponding to the type of information.
This is done using parametric modelling inspired by
the work of Lee and Hardy (2005). Specifically, the
topological relationship among urban geometry
features like ensuring shared geometry and
considering interfering features.
We create a hierarchical data structure to link
urban geometry (from the street, over the block, and
plot to the building). These links between the
different scales are bi-directional to access the plots
that belong to a street and the streets that belong to
a plot. This data can include analysis information, like
visible buildings to streets, local building densities,
active building frontages, etc. We will demonstrate
the method in the case of Weimar, using Data from
OSM and GeoPortal from Thuringia, Germany (GDI-
TH).
All analysis was calculated inside Rhino and
Grasshopper using DeCodingSpaces Toolbox by
Koenig et al. (2018) as it has a collection of methods
for urban data generation and evaluation.
In summary, we propose:
Mapping urban geometry in a specific
hierarchical structure using the parametric
environment.
Recording urban data relations and embedding
them in each urban element. This should help
the users quickly rebuild the exact spatial
hierarchy when the files are reused.
Spatial data query in the parametric
environment while considering MAUP effects.
Figure 2
Geometrical
features of
buildings
Volume 2 – Co-creating the Future – eCAADe 40 | 279
Organising the city geometry
The method we used for dividing the city geometry
into layers with simple elements was first adopted by
Dennemark et al. (2017). Figure 3 shows our
approach to the hierarchy connecting the geometry.
We start with the street network, passing data
through Blocks that contain the Plots, which contain
Buildings with street frontage information.
Streets
In our study, the streets mainly measure accessibility
using the nodes and linkages in the space syntax
spatial configuration methods (Hillier 1996a). This
simplified our focus only on the segments of the
streets as a network. Each street segment is carefully
placed between the plots and buildings to avoid
intersections.
Also, as presented in the case study later, the
streets are numbered and named to differentiate
them from each other when their information is
passed to the block, plot, or buildings. This is used
mainly to keep the data structure from being
randomised in Grasshopper every time we reload
the data. Also, since each street has an ID, it could be
used to quickly access their information on any other
platform such as Mapbox, as we will show in the case
study later.
Blocks
The block definition is not a straightforward process,
and some blocks could include pedestrian paths in
them. They are making the borders of the block
challenging to define. Moreover, some building
designs can include passages like tunnels or bridges
under or over their buildings. This adds to the
complexity of making a generalisable method for
block definition.
We found the best practice to define the streets
first as the public area is the easiest to clarify. Then
define the rules for dealing with the semi-private
semi-public spaces.
For example, if the study is about the frontage
access directly from the street, then the building
volume above the tunnel could be excluded in the
geometry but kept as an attribute. And if the density
of the buildings is more crucial, then the opposite
could be done. For solutions where both elements
are with the same value, we recommend creating a
parallel layer to maintain the original data structure.
The order of the segment lines constructing the
Block edges was recorded. This made the transfer of
the street frontage information much more
straightforward. Moreover, the street intersections
were transferred to each Block edge, and the edge
which faced this intersection was divided at the
mapped point. This made the transfer of all
surrounding streets’ information much more
accessible.
We have also found that extracting the blocks
from the plots is easier since they are primarily used
in defining land use. And therefore, if we eliminate
Figure 3
Mapping Hierarchy
from streets to
buildings passing
by plots and blocks
Figure 4
Mapping Street
intersection to the
Block to facilitate
the transfer of
multiple streets on
one building
frontage
280 | eCAADe 40 – Volume 2 – Co-creating the Future
all the public street plots, we will have the remaining
islands as the plots that are within one block.
Plots
The data from the GDI-TH had all the land in the city
covered with plots that define the land use. Even
streets and forest areas as well as hiking roads.
Since we can define all public areas, all the streets
and plazas were excluded, and the street blocks were
generated from the outlines of the rest.
All plots had their ID and which Block they
existed in. They also should have street direct front
access. Therefore, we look for island plots and merge
them with neighbouring street-overlooking plots.
This is governed by the buildings that are in them. If
multiple plots contain a building, it will be difficult
for the algorithm to connect it to all the plots.
Therefore, we prefer merging those plots and
treating them all as a new larger plot. If no buildings
were found, it is merged with the plot with the
closest distance to the street.
The edges of the plots were also ordered, and
the information on their relation to the Block edges
was documented. In other words, each plot edge has
the information of all the block edges it touches. And
since the Blocks were generated from the plots, any
edge that is not touching a block edge could be
identified as a not frontage edge and might not have
a relation to the streets. This made it easier to
differentiate between internal plot edges and
frontage edges by looking if plot edges have any
recorded information on the corresponding
attribute.
Buildings
All three-dimensional building geometries were
reduced to their footprint polygon surface to reduce
the file size. We found surfaces are more
straightforward to transfer than polyline curses. This
is mainly due to the problem of transferring
information about the open areas inside the
building. A building could have many polyline
curves inside it, increasing the complexity of
attaching information to one single object, as is the
case of the actual building data.
All buildings were grouped by containment
inside the Plots and Blocks. This made the transfer of
information between them straightforward.
Therefore, each building has information regarding
which street it overlooks, which edge has this
specific street ID, and which buildings are directly
attached to each edge.
Building functions were recorded inside each
building attribute as two types of lists. One list for all
the functions with direct individual access from the
street level. And another list for the rest of the
functions in the upper levels accessible through the
same main building entrance.
Frontages
As discussed in the previous section, the information
about the frontages is already embedded in each
building. Nevertheless, we will discuss some
Figure 5
Mapping Buildings
to their direct
Streets and reverse.
Volume 2 – Co-creating the Future – eCAADe 40 | 281
conditions regarding mapping the frontage
information.
Starting with the definition of a Frontage. In this
case, we use the term defined by Kickert (2016) as the
Active frontages. Specifically, we define how
accessible a frontage is to the public and the type of
visibility and permeability of this frontage. We will
discuss in the second case study how this
information is used.
The information on the GDI-TH platform also
included the address of the building. Both the name
of the street and the number of the building were
available. This will be beneficial information for our
second case study that will be presented later, as we
have observed that the information about its
address can define the entrances of each building.
Hence the entrance of a building with multiple street
frontages is on the address street name.
NOTES ON DATA MAPPING AND QUERY
When querying data inside the parametric
environment Grasshopper, some worth mentioning
conditions we found.
Information Hierarchy
The direction of mapping information from streets to
buildings was most efficient. This method
guarantees that the data organisation reduces the
complexity of spatial queries.
Moreover, since the inner objects like Buildings
and Frontages have all the information about their
containment, all the data structure could be reversed
so that each Street could have a list of all the
information regarding its relation to all the buildings
that have its ID.
Visibility and Direct access
When mapping buildings to their streets, some data
conflicts might appear. Specifically when choosing
to map all visible buildings to the streets. It will
produce a significant duplication of street IDs,
resulting in difficulty defining their relation to the
buildings.
For example, one building could have one direct
street frontage and at the same time is visible to 5
other street segments within a 100-meter radius. The
only way to specify the streets would be through the
research objective.
Figure 6
(Left)Mapping
visible buildings
onto a street
(Schneider et al.
2017)
(Right) Mapping
building functions
through their
frontages (adopted
from Bielik 2021)
Figure 7
(Left) Mapping
frontages to
dedicated street
segments
(Right) Mapping
Frontages to shared
street segments.
282 | eCAADe 40 – Volume 2 – Co-creating the Future
Nevertheless, we recommend using the direct
frontage method due to its simplicity for beginners
and wide use for applications that do not include
large visibility areas.
MAUP for Building Functions
The problem of mapping building functions is
outlined in Bielik (2021). Using our method, we
found that some notes must be considered once we
introduce more details to the model.
First, different options might affect the analysis
result when mapping the functions to the street. But
the case is a little different when using Frontages.
As frontages are more defined by their street
location, as we mentioned in the case of the building
entrance. More methods could be used to map the
location of each frontage to a dedicated part of the
street.
Another case when is when mapping the
number of functions in a building. We found some
functions occupying more than one building.
Therefore, each function will require a separate
identification to avoid duplications in their numbers.
Case study 1
In this case study, we presented a structured set of
data in a CAD file for a group of students to design a
small neighbourhood and analyse it in the context of
the city of Weimar in the seminar on Parametric
Urban Design and Analysis (PUDA). The provided
data on the city was created using the methods
described above. It consisted of 4 main layers for the
street network, blocks, plots, and buildings. Each
layer has the corresponding geometries with IDs,
and embedded attributes that describe each object
like the area, the number of storeys, and the number
and type of functions inside. The dataset also
includes precomputed and analysed information
about the performance of each object, like the
betweenness centrality of a street segment or the
local density of the area surrounding each building.
Using the tool mineR (Abdulmawla et al. 2018),
the students could extract information about the city
immediately while still being inside the same design
and analysis environment of Rhino/Grasshopper.
They could also reduce the computational effort by
embedding the design information inside the CAD
file.
We developed a process for combining various
data from the student’s group in a shared file. As
shown in figure 8, it is possible to upload the data to
an online map platform like Mapbox. This allowed a
simplified and more general for of presenting the
resulting information.
While working with the data, the students
required some tutoring to understand the process of
extracting and recording information to the CAD
model.
Case study 2
For a research study, our objective was to find
relations between different groups of functions
within the active city centre of the city. We used the
method to map the active building frontages to their
dedicated segment of the streets inside the historical
centre of Weimar, Germany.
To analyse the betweenness centrality of each
function frontage, we had to find the part of the
street segment that each frontage overlooks. This is
crucial to capture the through movement that the
betweenness centrality indicates. The final model
was generated using more information based on the
method to connect each frontage ID with its
corresponding street segment length.
Figure 8
Street information
passed from our
data model to an
online mapping
platform.
Volume 2 – Co-creating the Future – eCAADe 40 | 283
Based on the processed data, matching the
information about the visibility with the accessibility
of each frontage information was straightforward
using matching the IDs of the frontages in the
buildings. We used the direct frontage mapping
shown in Figure 9, where we dedicated a frontage
for each function type and connected this
information to the dedicated street segment.
For further analysis, we used the statistical t-test
to find relevant relations between each function. We
used sub-groups of functions (Fig 9) of the whole
function group inside the city centre. The idea is to
compare each sub-group with the rest of the group
to show how different the mean of each sub-group
is from the rest. Next, we use this information to
compare the sub-groups with each other’s (Fig. 10)
to find functions that share the same location
characteristics. We found three different types of
functional clusters inside the city centre.
Figure 9
location
information for all
the seven sub-
groups of functions
inside Weimar city
centre
Figure 10
(Left) a plot of the
metric
betweenness
centrality of clothes
function sub-group
(purple) and the
rest of the group
(grey)
(Right) Boxplot of
the BC at radius
R300 to compare
the relation
between the mean
of the cloth to the
rest of the group
284 | eCAADe 40 – Volume 2 – Co-creating the Future
DISCUSSION
Although the tools and methods we used are well
defined, a few notes need to be addressed: First, the
complexity of the data required for analysing the
connection between geometries requires a large
computer memory. And it grows exponentially with
increasing the study area. This could be solved by
dividing the study area into smaller sub-areas that
are easy to join after the analysis.
Second, updating the geometry might require the
embedded data to be double-checked for changes.
For example, we found that one street block had to
be cut into two during the study. This resulted in a
new street segment, two new blocks, and new plots
and building sides exposed to the street.
Third, the complexity of some geometries was
manually modified to run the analysis efficiently. For
example, a plot from the GDI-TH had a plot inside
another, which created a plot with no accessible
street front. Therefore, the inner plot was excluded.
CONCLUSION
Using parametric CAD environment methods can
facilitate further advanced spatial data analysis. As
shown in this paper, linking the elements of urban
geometry results in more possibilities for spatial data
analysis.
The presented data structure facilitates easy
access to link the available urban geometries and
related data (like land uses and local density) for
either urban design or analysis-based tasks. It makes
it possible to manoeuvre through urban data easily.
Moreover, since the whole process is developed
in a parametric CAD environment
(Rhino/Grasshopper), any processing step is easily
accessible for data extraction or manipulation. This
gives advanced users more flexibility to tailor the
tools for their needs for any custom investigations.
In the case studies, we exemplified the methods,
such as (1) extracting and recording street segment
information in the data structure and (2) identifying
similarities and differences in location characteristics
between functional groups.
This presented method facilitates the ease of
exacting information and reduces repetitive
computational tasks by recording their results.
However, since these applications were not used in
cross-disciplinary fields, once shared with the
community, we could further develop the tool for
broader integration.
REFERENCES
Abdulmawla, A., Schneider, S., Bielik, M. and Koenig,
R., 2018. Integrated data analysis for parametric
design environment-mineR: a grasshopper
plugin based on R.
Bielik, M., 2021. Form-Activity-Movement
Interaction Model. Doctoral Dissertation.
https://doi.org/10.25643/bauhaus-
universitaet.4397
Bobkova, E.; Marcus, L.; Berghauser Pont, M.;
Stavroulaki, I.; Bolin, D. (2019) ‘Structure of Plot
Systems and Economic Activity in Cities: Linking
Plot Types to Retail and Food Services in
London’, Amsterdam and Stockholm. Urban Sci.
2019, 3, 66.
https://doi.org/10.3390/urbansci3030066
Bobkova, E., Berghauser Pont, M. and Marcus, L.
(2021) ‘Towards analytical typologies of plot
systems: Quantitative profile of five European
cities’, Environment and Planning B: Urban
Analytics and City Science, 48(4), pp. 604–620.
DOI: 10.1177/2399808319880902.
Berghauser Pont, M., Stavroulaki, G., Bobkova, E., Gil,
J., Marcus, L., Olsson, J., Sun, K., Serra, M.,
Hausleitner, B., Dhanani, A., & Legeby, A. (2019).
The spatial distribution and frequency of street,
plot and building types across five European
cities. Environment and Planning B: Urban
Analytics and City Science, 46(7), 1226–1242.
https://doi.org/10.1177/2399808319857450
Berghauser Pont, M., 2018. An Analytical Approach
to Urban Form. In Teaching Urban Morphology
(pp. 101-119). Springer, Cham.
Dennemark, M., Schneider, S., Koenig, R.,
Abdulmawla, A. and Donath, D., 2017. Towards
a modular design strategy for urban
Volume 2 – Co-creating the Future – eCAADe 40 | 285
masterplanning. Sharing of Computable
Knowledge!, p.485.
Hillier, B., 1996a. Space Is The Machine: A
Configurational Theory Of Architecture.
Hillier, B., 1996b. Cities as movement economies.
Urban design international, 1(1), pp.41-60.
Kickert, C.C., 2016. Active centers–interactive edges:
The rise and fall of ground floor frontages.
Urban Design International, 21(1), pp.55-77.
Koenig, R., Beilik, M., Knecht, K., Abdulmawla, A. and
Fuchkina, E., 2018. New Methods for Urban
Analysis and Simulation with Grasshopper-
Using DeCodingSpaces-Toolbox.
Kropf, K., 2009. Aspects of urban form. Urban
morphology, 13(2), p.105.
Lee, D. and Hardy, P., 2005, July. Automating
Generalization–Tools and Models. In 22nd ICA
Conference Proceedings, A Coruña, Spain.
Lim, L., Yang, T., Vialard, A., Feng, C. and Peponis, J.,
2015. Urban morphology and syntactic
structure: A discussion of the relationship of
block size to street integration in some
settlements in the Provence. The Journal of
Space Syntax, 6(1), pp.142-169.
Pratt, M., 1999. Asking the right questions.
https://www.esri.com/news/arcuser/0309/files/r
ightqueries.pdf
Vialard, A. (2012) ‘Measures of the fit between street
network, urban blocks and building footprints’,
In: Eighth international space syntax
symposium, pp.8101:1–8101:16.
Wong, D.W., 2004. The modifiable areal unit
problem (MAUP). In WorldMinds: Geographical
perspectives on 100 problems (pp. 571-575).
Springer, Dordrecht.
Openshaw, S., 1983. The modifiable areal unit
problem (Vol. 38). Norwich: Geo Books.
Parametric Urban Design and Analysis – PUDA
(2022). Online Learning Platform for the
Bauhaus University-Weimar. Available at:
https://otp.uni-weimar.de/courses/parametric-
urban-design-analysis-puda-open/ (Accessed
15 June 2022)
286 | eCAADe 40 – Volume 2 – Co-creating the Future
ResearchGate has not been able to resolve any citations for this publication.
Thesis
Full-text available
This dissertation investigates the interactions between urban form, allocation of activities, and pedestrian movement in the context of urban planning. The ability to assess the long-term impact of urban planning decisions on what people do and how they get there is of central importance, with various disciplines addressing this topic. This study focuses on approaches proposed by urban morphologists, urban economists, and transportation planners, each aiming the attention at a different part of the form-activity-movement interaction. Even though there is no doubt about the advantages of these highly focused approaches, it remains unclear what is the cost of ignoring the effect of some interactions while considering others. The general aim of this dissertation is to empirically test the validity of the individual models and quantify the impact of this isolationist approach on their precision and bias. For this purpose, we propose a joined form-activity-movement interaction model and conduct an empirical study in Weimar, Germany. We estimate how the urban form and activities affect movement as well as how movement and urban form affect activities. By estimating these effects in isolation and simultaneously, we assess the bias of the individual models. On the one hand, the empirical study results confirm the significance of all interactions suggested by the individual models. On the other hand, we were able to show that when these interactions are estimated in isolation, the resulting predictions are biased. To conclude, we do not question the knowledge brought by transportation planners, urban morphologists, and urban economists. However, we argue that it might be of little use on its own. We see the relevance of this study as being twofold. On the one hand, we proposed a novel methodological framework for the simultaneous estimation of the form-activity-movement interactions. On the other hand, we provide empirical evidence about the strengths and limitations of current approaches.
Article
Full-text available
Plot systems (also referred to as "property", "parcel", or "lot") are generally recognised as the organisational framework of urban form that contributes to the economic performance of cities. However, studies that link the spatial form of plots to economic data are limited. The paper builds on the theory of Webster and Lai, which argues that the process of urbanisation is aligned with increased subdivision of property rights (increased division of land into plots, for example) due to the process of economic specialisation that is typical of cities. The aim of the paper is to test this theory by analysing whether there is a correlation between: (a) the shape and structure of plot systems, which are classified as types based on three plot metrics (size, compactness, and frontage index) and (b) economic activity, measured as the concentration of retail and food activities per plot. The paper will use statistical analysis to relate plot types to economic activity in three European cities (London, Amsterdam, and Stockholm). The results provide empirical support for our initial hypothesis and Webster and Lai's theory, which states that plots of smaller size, more regular shape, and smaller frontage generally correspond to a higher concentration of economic activity in cities.
Conference Paper
Full-text available
In this paper we introduce mineR-a tool that integrates statistical data analysis inside the parametric design environment Grasshopper. We first discuss how the integration of statistical data analysis would improve the parametric modelling workflow. Then we present the statistical programming language R. Thereafter, we show how mineR is built to facilitate the use of R in the context of parametric modelling. Using two example cases, we demonstrate the potential of implementing mineR in the context of urban design and analysis. Finally, we discuss the results and possible further developments.
Article
Full-text available
The paper discusses the relationship between the syntax of street networks and the differentiation of the size of urban blocks in a sample of small towns and settlements. The argument is in four parts. In the first part it is demonstrated, through design games, that the differentiation of streets by integration is linked to the differentiation of blocks by size. In the second part, it is shown that in a small sample of towns in the Provence, small blocks are not associated with more integrated streets but are distributed throughout the street network. The demonstration is based on an original method for studying the block size in relation to street integration. In the third part, the historic evolution of these particular towns is shown to involve the rationalisation of their integration core: integrated streets become better aligned and wider, and reach more directly into all parts of the town. However, the historic relationship between integration and block size is also based on mixture rather than a linear pattern of association. The final part uses these findings to advance a speculation about the origin of the syntax of these towns as compared to the syntax of the smaller settlements that Hillier and Hanson characterise as ‘beady rings’. This leads to a discussion of some of the abstract syntactic generators originally presented in The Social Logic of Space. In short, the final section of the paper argues that the lack of linear association between small blocks and integrated streets, in this particular sample, points to the emergence of gradually more complex generators of town form, generators which presuppose the ideas of the urban block and the street. These act upon the seeds of prior small aggregations, generated by simpler rules of adjacency.
Article
Full-text available
The diversity and complexity of human settlements is reflected in the range of ways we try to understand them. The richness of subject matter presented by cities has given rise to an equal richness in methods of investigation. Even within a single field such as urban morphology, there are different approaches with different terms of reference. The challenge raised by the diversity is not how to select between the different views but how to combine and co-ordinate them. The purpose of this paper is to undertake an initial critical analysis of different approaches to urban morphology in an effort to meet that challenge. The first aim is to identify the range of different phenomena taken as the object of urban morphological enquiry. The second is to identify an aspect that is common to all the approaches and that can be used as a reference key to co-ordinate different views in a rigorous way. The ultimate goal is a composite view in which the different approaches support each other to provide a better understanding of human settlements.
Article
The importance of the plot (also referred to as ‘property’) as one of the fundamental elements of urban form is well recognized within the field of urban morphology. Despite the fact that it is often described as the basic element in the pattern of land divisions, which are essential as organizational frameworks for urban form, studies offering comprehensive descriptions and classifications of plot systems are quite scant. The aim of the paper is to introduce a classification of plot systems into typologies based on five European cities, in order to distinguish particular spatial differences and similarities in terms of their plot structure. The proposed typologies are developed using unsupervised k-means cluster analysis based on numeric attributes derived from central theories in urban morphology. The introduced typologies are essentially configurational, allowing collective systematic properties of plot systems to be captured. Numeric attributes include plot differentiation (or plot size), plot frontage and compactness ratio, corresponding to essential qualities of plot systems such as the capacity to carry differences in space, the ability to operate as interface between street and building and providing a framework for evolution of built form over time. All three attributes are translated into configurational measures in order to capture the context of the plot system, rather than the parameters of individual plots. The combination of these deductively defined variables with algorithmically defined classification methods results in seven plot types that can be used to scale up traditional urban morphological analysis to whole city regions and conduct substantial comparison of patterns within, but also between these regions. Further, it also makes it possible to describe commonly recognized plot patterns and discover new ones.
Article
Typologies have always played an important role in urban planning and design practice and formal studies have been central to the field of urban morphology. These studies have predominantly been of a historical-qualitative nature and do not support quantitative comparisons between urban areas and between different cities, nor offer the precise and comprehensive descriptions needed by those engaged in urban planning and design practice. To describe contemporary urban forms, which are more diffuse and often elude previous historic typologies, systematic quantita-tive methods can be useful but, until recently, these have played a limited role in typo-morphological studies. This paper contributes to recent developments in this field by integrating multi-variable geometric descriptions with inter-scalar relational descriptions of urban form. It presents typologies for three key elements of urban form (streets, plots and buildings) in five European cities, produced using statistical clustering methods. In a first instance, the resulting typologies contribute to a better understanding of the characteristics of streets, plots and buildings. In particular, the results offer insight into patterns between the types (i.e. which types are found in combination and which not) and provide a new large scale comparative analysis across five European cities. To conclude, a link between quantitative analysis and theory is established, by testing two well-known theoretical propositions in urban morphology: the con-cept of the burgage cycle and the theory of natural movement.
Book
Since The social logic of space was published in 1984, Bill Hillier and his colleagues at University College London have been conducting research on how space features in the form and functioning of buildings and cities. A key outcome is the concept of ‘spatial configuration’ — meaning relations which take account of other relations in a complex. New techniques have been developed and applied to a wide range of architectural and urban problems. The aim of this book is to assemble some of this work and show how it leads the way to a new type of theory of architecture: an ‘analytic’ theory in which understanding and design advance together. The success of configurational ideas in bringing to light the spatial logic of buildings and cities suggests that it might be possible to extend these ideas to other areas of the human sciences where problems of configuration and pattern are critical.