improving the design of urban underground space in metro stations using the space syntaxmethodology
DESCRIPTION
This article explores the potential use of the space syntax methodology for evaluating user wayfinding, orientation and visibility in urban underground space. Two case studies from the Brussels-Capital Region are presented: the Bockstael metro station and the Anneessens premetro station. The use of the space syntax method is proposed to produce measurable or ‘hard’ parameters on design aspects that might otherwise be perceived as qualitative or ‘soft’ design aspects. Four space syntax analyses are conducted: an axial analysis, an orientability analysis, an isovist analysis, and agent-based modelling. A comprehensive, but effective, method using space syntax is developed to optimise design and renovation alternatives. We conclude that space syntax can provide a tangible contribution to the qualitative design of urban underground spaces.TRANSCRIPT
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Improving the design of urban underground space in metro stations using the space syntax methodology
Frank van der Hoeven1, Akkelies van Nes1
TU Delft, Faculty of Architecture and the Built Environment Netherlands [1]
Abstract
This article explores the potential use of the space syntax methodology for evaluating user wayfinding,
orientation and visibility in urban underground space. Two case studies from the Brussels-Capital Region are
presented: the Bockstael metro station and the Anneessens premetro station. The use of the space syntax
method is proposed to produce measurable or ‘hard’ parameters on design aspects that might otherwise be
perceived as qualitative or ‘soft’ design aspects. Four space syntax analyses are conducted: an axial analysis, an
orientability analysis, an isovist analysis, and agent-based modelling. A comprehensive, but effective, method
using space syntax is developed to optimise design and renovation alternatives. We conclude that space syntax
can provide a tangible contribution to the qualitative design of urban underground spaces.
Key words
wayfinding; orientation, visibility; underground; metro; station; Brussels; space syntax; design; urban underground
space; soft modes
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Introduction
Wayfinding, orientation and visibility are key user requirements in the design and construction of underground
spaces. However, wayfinding and orientating are more difficult underground because of the lack of reference
points, such as landmark buildings, and the absence of direct sunlight. Visibility is often hampered by labyrinth
like corridors characterised by a “hyper-accumulation of signs, media, symbols, lights, materials, displays, and
proportions” (Bélanger, 2006). These conditions influence the spatial legibility and social safety of underground
stations.
Discussions on design issues and problems regarding underground spaces began in the 1980s (Carmody, Huet
and Sterling, 1994). The spatial configuration of an underground space may even influence crime. Incidence
of crime in metro stations is not just a matter of organisational measures but of situational measures as well
(López, 1996). There “is a need for a more systematic approach to the design and assessment of quality of
underground spaces so that a better quality of underground spaces can be obtained” (Durmisevic and Sariyidiz,
2001).
Poor wayfinding, orientation and visibility in underground public spaces are not necessarily the result of
deliberate planning choices or a lack of either design skills or knowledge. The design and construction of
underground structures are ruled by several, sometimes conflicting, factors. There is a gap to be bridged
between the traditional engineering sciences and the architectural design disciplines. “The design of
underground infrastructure is often dominated by civil engineers but the aspect of social safety requires a
different view point that can be best realised by the involvement of an architect” (Bosch, 2011).
Argumentation in the engineering sciences is often based on ‘hard’ quantitative evidence, visualised through
complex diagrams and graphs that may appear quite solid in the eye of decision makers. Designers, however,
tend to illustrate their vision through ‘soft’ evidence, such as architectural renderings of situations ‘before
and after’ or ‘with and without’. Expert opinion is required to judge which design is preferable and much of
that judgement may sound like only ‘opinions’ to non-designers. Tool sets that can successfully measure these
‘soft’ parameters are now available. Space syntax is such a collection of tools, but it is still rarely used in the
field of tunnelling and underground space technology. The purpose of this article is to demonstrate the use of
these tools in this field. This article describes the outcome of a study commissioned for the renovation of two
underground (pre-) metro stations in the Brussels-Capital Region, Belgium. In addition, the applicability of space
syntax analysis to improve wayfinding, orientation, navigability and visibility in underground spaces is explored.
The Brussels public transport authority STIB commissioned the Yellow Design Foundation to conduct a
feasibility study for the upgrade of the (pre-) metro stations. The Yellow Design Foundation is an independent,
multidisciplinary and interregional platform for research and information on design and visual communications,
based in Brussels, Belgium (Yellow Design Foundation, 2012). Yellow Design subcontracted the space syntax
analysis to TU Delft. The Brussels case provided the authors a strategic opportunity to apply the space syntax
method and test its applicability in an environment described as “urban underground space (UUS)” (Bobylev,
2010).
This article briefly discusses the main features of the space syntax method. It describes the two Brussels
underground (pre-) metro stations that were analysed: Bockstael and Anneessens. It presents an effective
and simple method that is applied to optimise the two Brussels stations. The six-step research methodology
consisted of the following: mapping the underground urban space (1), performing visibility, axial, isovist and
agent-based analyses (2), evaluating the outcome (3), reconfiguring the floor plans (4), re-analysing the improved
floor plans (5), and conducting a side-by-side comparison of the original and improved floor plans (6). The article
concludes with a discussion of the applicability of space syntax for evaluating and improving the design of urban
underground spaces.
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
The space syntax method explained
The space syntax method used to evaluate the Bockstael and Anneessens stations was developed by Bill Hillier
and his colleagues at the University College London (Hillier and Hanson, 1984). Over the past three decades,
Hillier and his team have applied space syntax to urban studies and to complex buildings. In the last decade,
numerous improvements have been made to various types of spatial analyses and software development. The
evaluation of design proposals for the Tate Britain in London, 2002, provided a prominent showcase of the use
of space syntax for buildings (Dursum, 2007).First, a short overview is given of the history of relevant Dutch
spatial policy, including its main objectives of urban compaction and more liberal spatial planning. The following
section introduces seven significant types of urban developments at the rural-urban fringe and analyses three
regions showing different patterns of urbanisation. Finally, the most important findings are summarised and
evaluated in light of relevant policy objectives – not to evaluate the planning policy in the strictest sense, but to
identify future challenges for policymakers, urban planners and designers, on local, regional and national scales.
Figure 1
Use of space syntax in the case of Tate Britain
Space syntax is used in assessing and rearranging the interior spatial structure of complex buildings, such as
offices, retail (shopping malls), hospitals, museums, railway stations and cultural buildings. However, there are
currently no references found in scientific literature for the use of space syntax on urban underground spaces.
Space syntax is based on three concepts: the convex space, the isovist field and the axial line. Convex maps
are used for analysing buildings and the public spaces between buildings. Convex space is defined as: “all points
within a space that can be joined to all others without passing outside the boundary of the space” (Hillier 1988).
In urban analyses, the convex space analysis has been replaced by the point depth and the all-lines analyses.
No significant convex space analysis software improvements have been made since the 1990s. An isovist field
represents the panoptical view a person has from a given point in an urban space. It is used for orientation
or wayfinding in the urban fabric. Initially the isovist analyses were conducted manually. Now, both one-point
and all-points isovist analyses can be conducted using Depthmap, an open source application developed by
University College London. An axial line represents the longest sight line one has in an urban space or building.
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
It represents the way human beings move in lines through streets and roads, or rooms and corridors. During
the past two decades, the axial line has been the basic spatial element in the methodology and theoretical
development of space syntax in urban studies.
Figure 2
Axial lines, convex space, isovist (van Nes, 2012)
The main premise behind these three basic spatial elements is that human beings move in lines, interact in
convex spaces and experience changeable panoptical views when moving around in a built environment. As such,
it can be instrumental to test the requirements of undeground urban spaces. The space syntax methodology
has been verified during decades of research; consequently, the case studies presented in this article do not aim
to verify the tool set. Instead, the aim of this study is to draw conclusions on the spatial configurations of two
specific (pre-) metro stations and find generic leads to promote the use of space syntax in future research and
design of urban underground spaces.
The Bockstael and Anneessens stations in Brussels,
Belgium
In 2012, Yellow Design received a commission to evaluate two underground stations and advise on their
renovation. The commission included the Bockstael metro station and the Anneessens premetro station. The
Brussels premetro is a light rail system that was built with the option to upgrade to a full metro system in the
future.
Typically, a design consultancy conducts a number of analyses to obtain a better understanding of the qualities
and problems of the spatial configuration, the use of materials, the programming, the load-bearing structure,
and the uses of a building, facility or space.
The types of analysis and the methods used may differ from office to office as they are linked to the unique
architectural styles and design concepts that the designer or the design consultancy embraces. Our experience
in the field of underground space technology has shown weaknesses in analysing wayfinding, orientation and
visibility. We suggested that Yellow Design conduct a number of space syntax measurements to gain detailed
insights in these areas. The objectives of this research could then be defined using the following four research
questions:
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
1. Can space syntax be effectively used to assess wayfinding, orientation and visibility in urban underground
space?
2. If yes, what is the current state of wayfinding, orientation and visibility in the two Brussels (pre-) metro
stations?
3. What intervention(s) can be proposed for the investigated Brussels (pre-) metro stations to improve
wayfinding, orientation and visibility?
4. Which generic approach can be applied to these and other underground stations to investigate and improve
wayfinding, orientation and visibility?
Bockstael
The Bockstael station is situated along metro line 6 (Roi Baudouin - Simonis) in the Brussels municipality Laeken.
The metro station is connected to an underground train station with the same name. The Bockstael train
station is situated along rail line 50 (Brussels - Ghent). Rail line 50 is an above ground rail line covered with a
deck structure at Bockstael. The Bockstael metro station is located underneath the train station. The metro
station was inaugurated in 1982 and includes two tracks and two side platforms. The train station has a similar
configuration. The Bockstael metro station includes two intermediate (ticket) halls or mezzanines (north and
south) between the street-level and platforms. An additional passageway, which provides access to the metro
station, runs underneath the train station. The train station is located at the north side of the metro station.
Figure 3
Bockstael site
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Figure 4
a-b-c Bockstael cross-sections
Anneessens
The Anneessens premetro station is part of the North-South axis that stretches from the Bruxelles Nord
railway station, via the Brussel Midi (high-speed train) railway station to the Albert premetro station. The
Anneessens station is located in the centre of the municipality of Brussels, underneath Boulevard Anspach and
next to the Brussels Stock Exchange. It is served by the premetro lines 3 and 4. The station was inaugurated
in 1976 and includes four tracks, one island platform and two side platforms. The Anneessens premetro station
has an intermediate (ticket) hall or mezzanine between the street-level and platforms. The public area of the
mezzanine is divided into a northern and a southern section.
Figure 5
Anneessens site
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Preliminary analysis
Ideally, public transport users in an underground station are able to easily find their way from the various
entrances to the platforms below and from the platforms back to street-level. The design of such an ‘ideal
station’ reduces the number of required direction changes and angular deviations along that route as much as
possible. Along their way from street-level to the underground platform, the public transport users in this ‘ideal’
station experience clear overviews of the spaces they move through, receive clues as to major directions, while
they are aware of other users of the same space. Four measurements were performed on the Bockstael and
Anneessens (pre-) metro stations to assess the status of such ‘ideal’ requirements: an axial (all-lines) analysis, a
point depth (visibility) analysis, an isovist (intervisibility) analysis of all points, and an agent-based modelling of the
station. The results of the initial analyses indicated that the Bockstael station performed poorly. An experienced
space syntax analyst could easily spot the shortcomings of the northern mezzanine that links the metro station
with the train station. The removal of the physical separation between the two underground stations seemed
to be a promising solution. This solution was verified by additional analysis performed without this separation
wall. Yellow Design followed the advice to remove the separation wall and incorporated this design solution
in the proposal for the renovation of the Bockstael station. It was documented as a classical before-after
comparison (Figures 6.a-b) and reported to the public transport operator who seemed willing to implement the
solution.
Figure 6
a-b Bockstael station: before-after removal of the dividing wall between train and metro station (Leemans et al., 2012)
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
After the commission was concluded, the authors used the Bockstael and Anneessens cases as subjects of
further exploration into the application of space syntax to urban underground spaces. This additional work is
described in the following paragraphs.
Optimising the mezzanines
The key approach in the Bockstael case was to rethink the spatial configuration of the underground station
by questioning the appropriateness of the non-structural elements in the station. The wall that separates the
metro station from the train station is an obvious example. However, both stations contain numerous elements
that are not part of the load-bearing structure, but play a vital role in the way that users experience the interior
of the station. The following steps were developed and applied to evaluate if an underground urban space can
be improved in wayfinding, orientation and visibility:
1. map the underground urban space and make a distinction between the publicly accessible areas and the non-
accessible areas
2. use this floor plan to conduct axial, visibility, isovist, and agent based analyses
3. evaluate the outcome of the analyses and identify the ‘weak’ areas of the urban underground space
4. reconfigure the floor plan of the urban underground space by removing non-structural elements to improve
the ‘weak’ areas
5. use the improved floor plan to conduct the axial, visibility, isovist and agent based analyses again
6. evaluate the results based on a comparison of the original and improved floor plan results
The aim of the floor plan reconfigurations is to provide short and clear routes that minimise changes in direction
and angular deviations and to avoid spaces that are not required in the routing of metro users. In some cases
this means that the urban underground space should be expanded, in other cases its size can be reduced. Both
(pre-) metro stations have ‘mezzanines’ at level ‘minus one’ and the actual platforms are located one level below
at level ‘minus two’. The configuration of the platforms is simple and straight-forward and offers only few
opportunities for optimisation. The mezzanines, however, are divided by many walls that provide enclosed spaces
for staff, storage, shops, toilets, and (in the case of Bockstael) demarcate areas for different transport modes
(metro and train). These walls force metro users to unnecessarily change directions and make angular deviations,
while they block the view at many points along the route. The floor plans of both mezzanines, together with
their optimised versions, are provided in this paragraph. The Bockstael station has two mezzanines, a northern
mezzanine (shown right in the drawing) and a southern mezzanine (shown left in the drawing). The right
mezzanine connects the metro station with the train station. The halls of both stations are separated by a
non-load-bearing wall. The left mezzanine is a relatively small circular space that limits the overview a metro
user would otherwise have. The circular space connects to a ‘passerelle’ that provides access to the stairs and
escalators leading to the platform. A round information booth is placed in the centre of the circular space. In the
optimised floor plan, the right mezzanine is reduced in size along the straight line that links the two entrances
at street-level. The left section is expanded along the straight lines between the two entrances at street-level,
and between the two entrances at street-level and the ‘passerelle’. The information booth is removed. The
gate-lines in both stations were also removed. These gate-lines should be placed at street level were the real
entrance to the station is located. They should be as accessible as possible.
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Figure 7
a-b-c Bockstael mezzanines, original floor plan, weak points and optimised floor plan
The mezzanine of the Anneessens station is divided in to a northern section (right) and a southern section (left).
In between, there is a large section used for other purposes. The public areas of the Anneessens mezzanine
are fairly optimal, with the exception of the entrances from street-level. These entrances are blocked from
both view and direct access by additional walls that force metro users to make unnecessary direction changes.
In the optimised mezzanine, the walls are removed. This results in minimal expansion of the public space in the
mezzanine.
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Figure 8
a-b-c Anneessens mezzanines, original floor plan, weak points and optimised floor plan
These improvements to both stations were suggested only after performing the first round of space syntax
analyses and identifying the problem areas. It is more logical to present the floor plans side-by-side in this
paragraph to highlight the differences.
Axial, orientability, isovist, and agent-based analyses
The floor plans of both stations are analysed in their ‘original’ and ‘optimised’ states. The results of the axial,
orientability, isovist, and agent-based analyses for both stations are shown side-by-side to provide a clear
comparison of the original and the optimised floor plans. In both stations, the publicly accessible areas of the
mezzanines are located north and south of the platforms. This spatial configuration makes it possible to display
the floor plans of the mezzanine and the floor plans of the platform(s) in one single plane. If the public areas of
the mezzanines were located directly above the platforms it would have complicated the space syntax analyses,
requiring a more advanced tool set to work in three-dimensions. (A three-dimensional space syntax tool set is
available). The analysis results are explained in the text accompanying the floorplans. The axial, orientability and
isovist analyses results are presented in orange and red to indicate the highest (or ‘good’) values, while light blue
and dark blue indicate the lowest (or ‘bad’) values. The agent-based modelling results are interpreted based on
the agent density. Grey indicates no agents; blue means few agents; and orange and red indicate many agents.
Axial (all-lines) analyses
The first analysis discussed in this article is the axial analysis. The axial analysis is based on all the possible straight
walk lines inside the (pre-) metro stations. Such possible walk lines are drawn as straight lines in the floor plan
before conducting the analyses. The all-lines analysis shows accessibility properties. The higher integration of
an axis, the more it is coloured in red, indicating the fewest direction changes to all other axes in the station.
Conversely, the lower the integration, the more the axis is coloured in blue.
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Bockstael
Figure 9
a-b: Axial analysis of the Bockstael metro station
Figures 9.a-b show the axial analyses results for the Bockstael metro station. The routes from the mezzanines
to the platform are clearly the most integrated axes in the Bockstael metro station. These routes can be easily
found and understood with few direction changes. However, the Bockstael train station platforms are very
segregated spatially (in the drawing top-right, coloured in blue); passengers have to pass underneath the tracks
to reach the other platform. Routing is thus rather complex. This same passageway serves as an additional, but
far-from-optimal, entrance to the Bockstael metro station.
The optimised floor plan (b) shows improvements in the right mezzanine as a result of removing the wall
between the metro and train stations. The section that serves as entry to the train station is no longer blue,
but green. The overall central axis (red) now extends from the right mezzanine to the left, all the way to the
end of the station. The axis is no longer blocked by the information booth (the smaller circle to the left). The
adjusted shapes of both mezzanines do not seem to significantly affect the outcome of the axial analysis of
Bockstael. The modifications affect the routing only marginally.
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Anneessens
Figure 10
a-b: Axial analysis of the Anneessens metro station
Figures 10.a-b show the axial analyses results for the Anneessens premetro station. The left section of the
mezzanine is clearly the most integrated space in the current configuration (a), followed by the centre platform.
Minimal changes in the mezzanine floor plan (b) have a significant effect. The axial analysis is based on hierarchy.
The most integrated lines are shown in the diagram as red. Because of the changes to the right part of the
mezzanine, the centre platform increases in importance and becomes among the most integrated parts of the
floor plan. All street-level entrances improve as well. The left section of the mezzanine remains a well-integrated
part of the station, although the colour in the diagram changes from red to orange.
Orientability (point depth) analyses
A point depth analysis shows the degree of direction change from each point in the analysed space to all other
points. How many times does someone have to change direction from a given position in the station if he or
she wants to oversee the whole station? The point depth analysis divides a space in to grid cells and calculates
how each cell relates to all other cells in the grid. Obstacles, such as columns or fences, increase the topological
depth between various cells (Turner, 2007). Less point depth is desired for optimal orientability. The point depth
analysis is useful to determine where the most and least orientable areas are located in the underground facility.
It describes the variation in the spatial properties within large continuous spaces. The method allows for the
testing of how the overall outline of the station, as well as the placement of columns, announcement walls,
fences and advertisements, affect a station’s degree of orientability.
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Bockstael
Figure 11
a-b: Point depth (orientability) analyses of the Bockstael metro station
Figures 11.a-b show the orientability analyses results for the Bockstael metro station. The results of the original
floor plan analysis (a) are consistent with the axial analyses. The platform and the left and right mezzanines
of the current floor plan (a) are those spaces in the Bockstael station where metro users can best orientate
themselves. The yellow, orange and red colours indicate a higher level of visibility. The corridors to the staircases
that lead to street-level perform poorly, as does the passageway underneath the Bockstael train station. The
improved floor plan (b) shows clear increases in orientability at both mezzanines, exactly where required, at the
point where the metro users have to change direction.
Anneessens
Figure 12
a-b: Point depth analysis (orientability analyses) of the Anneessens premetro station
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Figures 12.a-b show the orientability analyses results for the Anneessens premetro station. The current
Anneessens station floor plan (a) performs rather well, particularly compared to the Bockstael station. The
colours red, orange and yellow clearly dominate. The most orientable spaces inside the Anneessens station are
the platforms and the left section of the mezzanine. Among the few segregated spaces in the station are the
staircases that link the station to street-level. The segregation of these staircases is the result of the poor
configuration of the interior space of the mezzanine. In the Anneessens station optimised floor plan (b), the
walls that separate the staircases from the mezzanine are removed. Although this intervention is small, the
impact is significant, particularly in the left section of the mezzanine.
Isovist analyses, all points
An isovist is: “the set of all points visible from a given vantage point in space and with respect to an
environment” (Benedikt, 1979). An isovist visualises a human’s panoptical view from a particular perspective in a
built environment. The panoptical view boundaries are defined by walls and free standing objects, such as trees,
bushes and statues, located within a built environment. When moving around in built environments, the shape
and size of the isovist changes. It is thus possible to visualise the sequences of scenes or panoptical view arrows
from particular points along the movement routes. The isovist analyses are useful for analysing the degree
of visibility of the panoptical view of a room from a specific point, and how new interventions will increase or
decrease existing isovist views. Using graphical analyses, the Depthmap software is able to calculate the degree
of integration of each point or isovist root related to others in a built environment. In Depthmap, spaces in a
built environment are rasterised by a grid. One can choose how fine-grained the grid can be. The more fine-
grained the grid is, the more time consuming the analyses will be and, hence, the more exact the results will be.
Each point for the visibility analyses is taken from the centre of each grid cell. While the axial analysis focused
on sightlines along a passenger’s route (wayfinding), the isovist purpose is to analyse the level of intervisibility,
that is, how people can see each other and how others can see them. In research on space and crime, the
intervisibility between buildings and streets contributes to perceived and conceived safety in urban areas (van
Nes & Lopez 2010). The fact that someone can be seen does not mean that someone else can walk in a straight
line towards that person. Passengers can see in a straight line but they are not allowed to cross the tracks or
jump over fences. These differences between accessibility and visibility determine the difference between the
point depth/axial analysis and the isovist analysis.
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Bockstael
Figure 13
a-b: Isovists analyses of the Bockstael premetro station
Figures 13.a-b show the isovist analyses results for the Bockstael metro station. They indicate that the tracks
and, to some extent, the mezzanines are the most intervisible spaces in the station. The isovist analysis of the
original floor plan (a) shows a great deal of blues and greens. The wall that separates the metro station from
the train station appears to play a significant role in this. The information booth in the left mezzanine has a
negative effect on the intervisibility of the station. The optimised version of the floor plan illustrates that there
is room for improvement. Expanding the left section to a triangular space, removing the circular booth, reducing
the right mezzanine in size and integrating the metro and train stations have an overall positive impact on both
mezzanines, which even extends to the platforms.
Anneessens
Figure 14
a-b: Isovists analyses of the Anneessens premetro station
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
Figures 14.a-b show the isovist analyses results for the Anneessens premetro station. The image is
predominantly coloured red and orange, representing the highest values. Metro users appear to have a perfect
overview of their surroundings, resulting in a high awareness of other people who use the space. The staircases
and escalators that lead to the street-level are an important exception here; they are all coloured in blue,
representing the lowest values. The floor plan optimisation proposal to remove a small number of non-load-
bearing walls improves the situation. A station that performs well is even capable of becoming excellent.
Agent-based modelling
The recently developed Depthmap software agent based modelling is based on how people actually orientate
themselves in buildings and urban areas. Through empirical testing of how people move through virtual
environments with strange angles, significant correlations between actual human behaviour and the results
from the all-lines analyses and point depth analyses were found. The least angular deviation from one’s direction
plays a role in how people orientate themselves through built environments (Conroy 2001). Therefore, the
results from this analysis can be useful in estimating how urban spaces and buildings will be navigated in the
future in addition to how they were navigated in the past. Likewise, it is possible to investigate how large
crowds will behave in a given area, or how people orientate themselves from a given point in different time slots.
People, in this case metro users, are represented by so-called ‘agents’.
Bockstael
Figure 15
a-b: Agent based modelling of the Bockstael metro station, original floor plan (a) and optimised floor plan (b)
Figures 15.a-b show the agent based analysis results for the Bockstael metro station. In the original floor plan
(a), the tracks appear to generate the highest number of agents. The utilisation of the original floor plan is not
well balanced. There are significant areas that are not used (grey). In the optimised floor plan (b), the areas that
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
attract the most agents have expanded to include the platforms and significant parts of the mezzanines. The
areas not used by agents (grey) have been reduced.
Anneessens
Figure 16
a-b: Agent based modelling of the Anneessens premetro station
Figures 16.a-b show the agent based analyses results for the Anneessens premetro station. In the original
floor plan (a), the ticket hall generates the highest number of agents, followed by the platforms, with the
centre platform performing better than the other platforms. In the optimised floor plan (b), the left part of
the mezzanine generates even more agents. The overall distribution of the agents throughout the station has
been improved. The areas without agents (grey) have been reduced significantly. The placement of the gate-line
where metro users check-in and check-out is clearly important. It is at the (removed) gate-line where the large
improvements occur.
Conclusions
Space syntax can be effectively used to explore user wayfinding, orientation and visibility in urban underground
space. The spatial configurations of the two Brussels stations were ideal for the space syntax analyses. It is
possible to investigate more complex multi-level spaces but a more advanced tool set would be required to do
so. However, the fact that space syntax has not been used to analyse urban underground spaces is not because
of methodological or technical limitations.
The ease with which metro users find their way, orientate themselves, and become aware of others in the
underground station differs between the Bockstael and Anneessens stations. The Anneessens station is well
organised spatially, with clear and straight routings. It performed rather well in the analyses. The Bockstael
renovation would require significantly more effort to reach a similar level of quality.
The configuration of the mezzanine level of both stations presents room for improvement. The ‘raw’
construction of the mezzanines offers ample opportunity and sufficient space to enhance wayfinding,
orientation and visibility. This potential is not being fully utilised in the Bockstael and Anneessens stations
because of numerous non-load-bearing walls that provide enclosed spaces for staff, storage, shops, toilets, and
(in the case of Bockstael) demarcation of areas served by different transport modes (metro and train). The
position of the gate-line is an important factor as well.
Frank van der Hoeven, Akkelies van Nes
Improving the design of urban underground space in metro stations using the space syntax methodology
Tunnelling and Underground Space Technology, Volume 40, February 2014, Pages 64–74
http://dx.doi.org/10.1016/j.tust.2013.09.007
The spatial configurations of the mezzanines can be improved by both enlarging and reducing the space that
is available to the metro users. The key objective here is to provide straight routes and sight lines between
the street-level entrances and the staircases that lead from the mezzanine to the platforms. Because the
‘mezzanine’ is a common feature in many underground (metro) stations, this conclusion provides a generic
suggestion for improving other underground stations as well.
References
Bélanger, P., 2006. Underground landscape: The urbanism and infrastructure of Toronto’s downtown pedestrian network. Tunnelling and
Underground Space Technology, Volume 22, Issue 3, May 2007, Pages 272-292 http://dx.doi.org/10.1016/j.tust.2006.07.005
Benedikt, M. L., 1979. To take hold of space: isovists and isovist fields. Environment and Planning B: Planning and Design, Volume 6, Pages 47–65.
http://dx.doi.org/10.1068/b060047
Bobylev, N., 2010. Underground space in the Alexanderplatz area, Berlin: Research into the quantification of urban underground space use.
Tunnelling and Underground Space Technology. Volume 25, Issue 5, September 2010, Pages 495–507. http://dx.doi.org/10.1016/j.
tust.2010.02.013
Bosch, J., 2011. Social safety un underground metro stations; lessons learned and implemented in the Amsterdam metro, the Netherlands. ITA-
AITES World Tunnel Congres 2011, Helsinki.
Carmody, J., Huet, O. & Sterling, R., 1994. Life Safety in large Underground Buildings: Principles and Examples. Tunnelling and Underground Space
Technology, Volume 9, Issue 1, Elsevier, Exeter, Pages 19-29. http://dx.doi.org/10.1016/0886-7798(94)90005-1
Conroy Dalton, R., 2001. The secret is to follow your nose. In: J. Peponis, J. Wineman, and S. Bafna, editors, Proceedings Space Syntax. 3rd
International Space Syntax Symposium, Atlanta. http://discovery.ucl.ac.uk/1023/
Durmisevic, S. & Sariyildiz, S., 2001. A systematic quality assessment of underground spaces – public transport stations. Cities, Volume 18, Issue
1, February 2001, Pages 13–23. http://dx.doi.org/10.1016/S0264-2751(00)00050-0
Dursum, P., 2007. Space Syntax in architectural design. In: Proceedings Space Syntax. 6th International Space Syntax Symposium, Istanbul.
Fakhrurrazi, R. & van Nes, A., 2012. Space and Panic. The application of space syntax to understand the relationship between mortality rates and
spatial configuration in Banda Aceh during the tsunami 2004, In: M. Greene, J. Reyes and A. Castro, editors, Proceedings Space Syntax. 8th
International Space Syntax Symposium, Santiago de Chile: PUC. http://www.sss8.cl/media/upload/paginas/seccion/8004_1.pdf
Hillier, B., 1988. Against enclosure. In: N. Teymus, T. Markus, and T. Woaley, editors, Rehumanising housing, Pages 63–85. Butterworths, London.
Hillier, B. and Hanson, J., 1984. The Social Logic of Space. Cambridge University Press, Cambridge.
Leemans, A., Ivkovic, M., Duponcheel, M. & van Lierde, B. (2012). Rapport Station Bockstael. Yellow Design Foundation, Brussels.
López, J.J., 1996. Crime prevention within Metro Systems. European Jornal on Criminal Policy and Research, Volume 4, Issue 4, Pages 113-119,
Springer. http://www.dx.doi.org/10.1007/BF02736717
van Nes, A., 2012. The one- and two-dimensional isovist analyses in Space Syntax. Pages 163-184. In: Nijhuis, S., van Lammeren, R. & van
der Hoeven, F. editors, Exploring the Visual Landscape; Advances in Physiognomic Landscape Research in the Netherlands. IOSpress,
Amsterdam. http://dx.doi.org/10.4233/uuid:fe6698ae-045c-436b-945b-c61c4b0437e4
van Nes, A. and López, M., 2007. Micro scale spatial relationships in urban studies. the relationship between private and public space
and its impact on street life. In: A. S. Kubat, editor, Proceedings Space Syntax. 6th International Symposium, Istanbul. http://www.
spacesyntaxistanbul.itu.edu.tr/papers%5Clongpapers%5C023%20-%20VanNes%20Lopez.pdf
Turner, A., 2007. UCL Depthmap 7: From isovist analysis to generic spatial network analysis. In: A. Turner, editor, New Developments in Space
Syntax Software, Istanbul Technical University, Istanbul. http://www.vr.ucl.ac.uk/events/syntaxsoftware07/turner.pdf
Yellow Design Foundation, 2012. Mission de consultance dans le cadre de la rénovation des stations Bourse, Bockstael et Anneessens. Brussels.