UNIVERSITY OF AMSTERDAM
GSSS
GRADUATE SCHOOL OF SOCIAL SCIENCES URBAN AND REGIONAL PLANNINGStudying Desirable Social Interactions in Design and
Assessment of Street Intersections.
Marina Vasarini Lopes UvA. 11761962
marina.vl@gmail.com Master Thesis Urban and Regional Planning
Theme: The dilemma of urban mobility, and beyond Advisor: Marco te Brömmelstroet
Second Reader: Sara Özogul
3
Abstract:
Since the 90’s and the failure of car-oriented cities to bring life quality to its citizens, the field of traffic design is in transformation. Streets are more and more understood as part of the public space system and people are being acknowledge as the most important variable in the planning process. However, a question remains: if streets are designed to provide a diversity of benefits for people, how is the traffic assessment measuring these benefits?
The field of Social Science traditionally ignored the impacts of mobility in social life, only recently starting to explore the relationships between people in transit situations, what became known as the Mobility Turn. Since street design aims to cause social transformation, social science methods must be suitable to evaluate what impacts this design is actually causing in people and society.
In agreement with the Mobility Turn agenda, this study combines the concepts of Social Interactions (from Social Science), Shared Space (from Urban Design) and Conflict Analysis (from Traffic Assessment) to provide new lenses for traffic assessment, capable of exploring the positive effects of mobility and meeting in transit. Named Desirable Social Interactions, the main goal is to define a new methodology and study the usefulness and feasibility of including it in large scale traffic assessment.
The research was divided in three parts. First, the city of Amsterdam was used as a case study to create an overview of current practices of street design and assessment. In the second part, the concept of Desirable Social Interactions was defined and operationalized through interviews with experts and survey with pedestrians and cyclists, called street users. The third part tested the new methodology in a comparative case study of two sections of the Koningsplein square, in Amsterdam.
The main outcome of the research is a first systematization of the method, defining classes of interactions and outlining an observation toolbox to assess interactions, by focusing on the behavior of street users. The experiment also gave insights on how street design with elements of shared space influences Desirable Social Interactions.
5
Acknowledgement:
I would like to thank my advisor, Marco te Brömmelstroet, not only for the guidance but also for sharing his knowledge, connections and quite few hours in interesting discussions.
I also would like to thank all the civil servants from the municipality of Amsterdam that, in a more formal or informal way, helped me in the data collection. More specifically, I would like to thank Kees Vernooij, Sjoerd Linders, Koen Schreurs, Dirk Iede Terpstra, Fadoua Boufarrah and Maarten Nulle for searching for reports and constantly answering my e-mails. I am also very grateful for the academics interviewed for this research, for brainstorming with me and developing together an enthusiasm for the subjects.
Last, a special thank you for Patrick Chaud-Brasfort, who kindly read the entire text at the last moment, to my colleagues for the University of Amsterdam that shared this journey with me and for my family, that supported me all the way from Brazil.
7
Table of Content
Abstract: ... 3
Acknowledgement:... 5
Chapter 1 - Introduction: ... 9
Chapter 2 - Theoretical Framework: ... 12
Shared Space ... 12
Shared Space: a design on a human scale ... 13
Assessment of Shared Spaces ... 13
Traffic Assessment ... 14
Conflict Analysis (CA) ... 14
Limitations of the CA method ... 16
Slow Traffic assessment ... 16
Social Interactions (SI) ... 16
Interactions in Mobility and the Mobility Turn ... 17
Problem statement... 17
Research questions: ... 18
Conceptual Framework ... 18
Chapter 3 - Research Design ... 20
Sub-question 1: What are the current traffic design and assessment methods? ... 20
Amsterdam as the Case Study ... 20
Method of Data Collection and Analysis: ... 20
Sub-question 2: Which indicators define DSI and can be included in new methods of traffic assessment? ... 21
Data collection – searching previous studies ... 22
Data collection – selecting the experts ... 23
Data collection – public survey... 23
Data analysis – a continuing process ... 24
Sub-question 3: When these new methods are included in traffic assessment, what can the outcomes be? ... 24
Choosing the Case Studies – Koningsplein ... 24
Data collection – videos footage ... 26
Method of data analysis – domain analysis ... 26
Chapter 4 - Analysis: ... 28
Sub Question 1 - The inclusion of DSI in design and assessment. The Amsterdam Case ... 28
The standard methods of design and assessment ... 28
8
Sub Question 2 - New perspective on meetings in transit: Desirable Social Interactions ... 35
Using interviews and surveys to assess DSI ... 35
Using observations to assess DSI ... 36
Operationalizing DSI – finding indicators of behavior ... 36
Defining classes of DSI – towards a practical method ... 38
Classes of DSI ... 40
Sub Question 3 - Experimenting the new toolbox for assessing DSI in traffic ... 42
Testing the methodology ... 42
Analysis of the experiment ... 45
Chapter 5 - Discussion: ... 47
The inclusion of DSI in intersections design and assessment ... 47
Is it possible to make it different? How? ... 47
What can the outcomes be? ... 48
Reflections, Potentials and Limitations ... 49
Chapter 6 - Conclusion ... 52
Chapter 7 - Bibliography ... 53
Appendix ... 57
1.1 - List of Civil Servants consulted in Sub-Question 1 ... 57
1.2 - Table with results of Sub Question 1 ... 58
2.1 – Structure of interviews with experts ... 61
2.2 – Complete Observation Table for DSI ... 64
3.1 – Sample of the description of the observations ... 65
9
Chapter 1 - Introduction:
“When people think about cities, they tend to think about certain things; they think of buildings, and streets, and skyscrapers, and noisy cabs. But when I think about cities, I think about people. Cities are fundamentally about people.”
Amanda Burden – former director of the New York City Department of City Planning (2014)
Not only are cities about people but streets are the places where city life is expressed. Public Spaces are the ones where the social animal, as described by Aristotle (384-322 v. Chr.), discovers uncertainties and finds its identity. Literature, especially related to architecture and urban design (Gehl, 2010), already revealed the importance of public spaces in enhancing citizens’ well-being, sense of belonging and health. Considering that 60 to 70% of a city’s public space should be designated for streets (Scruggs, 2015), including them in the system of public space is fundamental to understand the full potential of these areas to enhance social life. Moreover, with commuting time in Europe reaching one and a half hours per day (Eurostat, 2018), daily trips should overcome the single purpose of transportation and start to include social benefits.
Since the implementation of motorized vehicles in the 30’s and the inclusion of modern ideals of city planning (Kaparias, Bell, Biagioli, Bellezza, & Mount, 2015, p. 115), public spaces saw a segregation of functions, in which parks and playgroups were designated to leisure and streets were exclusive for mobility, guaranteeing the most efficient infrastructure for fast movement. However, especially after the 90’s, high rates of mortality, health problems and low urban quality related to existing traffic systems (Hamilton-Baillie, 2008; PPS, 2018), raised questions over the real capacity of car centric cities to deliver life quality for their citizens, culminating in a review of pre-conceptions of public space, mobility and street design.
In that same period, several new concepts started to express a priority change in street and public space planning, from car-centric to people-centric projects. Usually started by grassroots initiatives, the core idea was to subvert the segregation of public life functions by decreasing cars’ central place in streets and encouraging other modes of transport and activities. Translated in terms such as complete streets, placemaking, traffic calming and shared spaces, what started with small interventions such as placing flower baskets in the street, painting extra sidewalks or laying temporary games in the car lanes, was transformed into real policies and replicated in several contexts.
For instance, the concept of Complete Streets focuses on designs that encourage the presence of different types of street users (Benfield, 2013), such as pedestrians, local residents, cyclists and car motorists. The biggest challenge is to use the little space to accommodate adequate infrastructure for all of them. Parallel to this idea, Placemaking is strongly related to the sense of community, advocating that public spaces are the heart of city life and collaborative processes are the most efficient way of achieving space-social quality (PPS, 2018). Traffic calming, on the other hand, is more related to technical solutions to enable people to reclaim streets as public spaces. Toolboxes for traffic calming street design
10 describe strategies such as increasing sidewalks, narrowing car path and corner curbs (PPS, 2008).
Although there are already several strategies for putting people as the centre of street design, there are few tools to evaluate the real social transformations of these initiatives. The Gehl Institute, one of the best known companies working with urban design for people, developed a toolbox for public spaces evaluation (Gehl, 2010) and recently created a method of data collection and analysis, in an attempt to standardize and upscale the assessment of public spaces (Gehl Institute, 2017). However, the manual is limited to assessing the quality of the infrastructure, like width of bicycle lanes, timing of traffic lights, visibility in intersections. It does not focus in people’s perception of the places and social transformations.
Being initiated in the field of design, it is natural that first attempts of assessment are also linked to the competences of this field. Still, there is a lack of research and tools to understand broad social impacts of urban planning changing perspective. The field of social science historically ignored the influences of physical space and mobility in social relations, viewing them as given technical solutions (Urry, 2012). An interest in the influences of global mobility in society transformation only started to increase at the beginning of the XXI century, with the conceptualization of the New Mobilities Paradigm (Sheller & Urry, 2016) and the Mobility Turn. Mobility Turn provides new lenses for social science, focusing on the impacts of mobility of people, goods, economy, data, and others, in the development of contemporary societies. Aligned with this frame, scholars such as Jensen (2010 a) and Brömmelstroet (2017) are now researching the impacts of micro mobility (commuting, moving around the city) in creating social norms and individual sense of belonging.
In accordance with the mobility turn theory, this research understands transportation not only as a way of accessing well-being, but also as a well-being generator. For this reason, it intends to contribute to the debate by merging the new perspectives of street design, the most recent studies in social science, and standard methods of traffic evaluation to test new tools for traffic assessment, centered on the users of the street and not on the physical space. The main objectives are then to add to the academic debate on the role of streets and mobility in the development of people as social actors, and provide designers with examples on how to include social indicators in traffic design and evaluation. More specifically, a new concept of traffic assessment, named Desirable Social Interactions, will be defined and tested, using street intersections and case studies as research guides.
In the next chapter, the theoretical framework will present the concepts of Shared Space, Conflict Analysis in traffic assessment and Social Interactions in mobility and connect them in the context of this research. Still in chapter 2, the research question and the sub-questions will be introduced and the conceptual framework will describe the particular way each theory will contribute to answering the questions.
In Chapter 3, the methodology used to collect and analyze data will be shown, organized by sub-question. In the first phase of the study (sub-question 1), the approach for obtaining an overview of current practices in street design and assessment will be outlined, using Amsterdam as the case study. In the second phase, the main methodology, interviewing
11 experts in order to promote a theory-based brainstorm, will be presented. To answer the third and last sub-question, the methodology for creating an experiment with the comparison of two case studies in the square Koningsplein will be explained.
The fourth chapter will describe the analysis and the outcomes for each phase of the research. For the first sub-question, an overview of standard and experimental practices in traffic design and assessment will be summarized. The second part of the analysis will bring the definition of Desirable Social Interactions and two table to operationalize the method. Answering the third sub-question, de outcomes of the experiment will be presented, together with a toolbox for assessing DSI.
Chapter 5 will be a discussion chapter, with interpretations of the analysis by sub-questions and conclusions about the potentials of the new methods and the meanings of the experiment for intersections design and the academic debate. Additionally, this last part will reflect on the potentials for future research and the limitations of the chosen methodology.
12
Chapter 2 - Theoretical Framework:
Mobility is commonly understood only as the capacity to move from point A to B (Jensen, 2010 b, p. 351), considering the time and capital needed for that. However, several scholars already highlighted broader benefits of mobility, especially in its capacity to contribute to social indicators such as sense of belonging, exposure to diversities (Brömmelstroet, Nikolaeva, Glaser, Nicolaisen, & Chan, 2017, p. 2), well-being (Schwanen, 2014; Epley & Schroeder, 2014), accessibility and equity (Brömmelstroet, Nikolaeva, Glaser, Nicolaisen, & Chan, 2017, p. 3). Moreover, streets are no longer acknowledged only as places to exercise mobility, but also places of encounter and permanence (Gemeente Amsterdam, 2013; Hamilton-Baillie, 2008), functions commonly attributed to public spaces (Gehl, 2010).
This research aimed to contribute to this discussion by combining the new understandings of mobility and streets, and include them in the assessment of traffic systems, more specifically, in the evaluation of intersections. To enable this merge, the concept of Social Interaction was used to explore positive perspectives of interactions in mobility, in contrast to current negative connotations embodied by the Conflict Analysis method. By doing so, the encounter between two people at an intersection (from now on called users) is no longer considered necessarily undesirable (a conflict), allowing the evaluation of potential social benefits. With this intention, the following research question was answered:
How are Desirable Social Interactions considered in the assessment and redesign of intersections and how can this be improved?
In order to explore this new perspective in mobility design and evaluation, the theories of Shared Space as an alternative in street redesign, Conflict Analysis as a method of traffic assessment and Social Interactions as alternative lenses to analyze behavior of street users were combined. The next paragraphs will develop each theory and their interrelation. Shared Space
“[in shared spaces] Traffic rules make way for social rules” (NHL Hogeschool, 2012, p. 11)
Aligned with modern ideas of city planning from the 1930’s and the popularization of motorized vehicles (Hamilton-Baillie, 2008, p. 165; Kaparias, Bell, Biagioli, Bellezza, & Mount, 2015, p. 115), cities witnessed the segregation of traffic and leisure functions and the designation of public spaces for motorized traffic. Already in the 60’s, critics started to question the lack of interest in people in the modern model of city planning. Jane Jacobs, for instance, questioned the “abstract man” planners were considering to live in the city (Jacobs, 1961, p. 92) and advocated that sidewalks, under the “eyes of the street”, were the best place for children’s leisure, not playgrounds (Jacobs, 1961, p. 84) as the urban design trend was implementing.
The idea of City for People, assigned to the Danish urban designer Jen Gehl (2010), is a contemporary attempt to consider the human being as focal point in the planning process. The book City for People (Gehl, 2010) outlined tools for street design and parameters to evaluate public spaces (parks, squares, sidewalks) from the perspective of the built environment. The design tools are divided into five principles: (i) Design on a human scale
13
with diversity at the eye level, (ii) Existence of activities that attract people, (iii) Perceived safety, (iv) Social and environmental sustainability, and (v) Promotion of healthy daily activities. Meanwhile, the evaluation makes use of twelve categories, divided into three classes: (i)Protection: Protection against traffic accidents, against crime and violence, unpleasant sensory experiences. (ii) Comfort: opportunity to walk, to stand/ stay, sit, see, talk/listen, play/ exercise. (iii) Delight: scale, positive aspects of weather, positive sensorial experiences.
Shared Space: a design on a human scale
Shared Space is another example of the return of people-centric urban planning. This concept contributes to urban design by understanding streets also as public spaces and attempting to merge the functions of mobility and leisure (Wouterson, 2016, p. 14). The main idea behind the concept is that by removing traffic regulations such as pavement stripes and traffic lights, safety is enhanced by introducing ambiguity (Kaparias, Bell, Biagioli, Bellezza, & Mount, 2015, p. 116). According to Hamilton-Baillie, a designer specialized in this concept, traffic rules in shared spaces are based on “informal social protocols and negotiation” (Hamilton-Baillie, 2008, p. 166).
Started by the Dutch engineer Hans Monderman in the redesign of residential zones (Peters, 2017, p. 2), it also became popular in other countries such as UK (Hamilton-Baillie, 2008, p. 167) and Norway (Peters, 2017). The key idea is that people in movement can regulate themselves without the need of external traffic rules. Once the external rules are absent, users rely on spatial clues like difference in the pavement and facades to find their way. Moreover, they rely on each other to understand the space, implying that interaction between them is inevitable (Hamilton-Baillie, 2008, p. 171; Peters, 2017). The expected outcomes from this type of design is increase in safety, an improvement in flow and a rise in the number of pedestrians. (Hamilton-Baillie, 2008).
According to the Shared Space Knowledge Center, from the NHL University of Applied Science, this type of design can be implemented in specific contexts that contain key criteria. First, it has to include several functions such as space for staying and meeting, traffic and other activities. Second, it must use minimal legal traffic instruments, such as pavement stripes, level difference to segregate modes of transport or traffic lights. Third, it needs a historical environment with the potential to steer behavior. And fourth, it must be a participatory project, including residents and users (NHL Stenden Hogeschool, 2018; NHL Hogeschool, 2013, p. 8). However, these principles vary in location and each country or municipality has their own understanding of what defines a shared space (Nota, 2018), applying the concepts to varying degrees. Another relevant spatial aspect to be considered is the diversity of modes of transport making use of that space. When motorized vehicles are present and need to use the same space as pedestrians and cyclists, a bigger risk is perceived. In these situations, shared space is considered a valuable tool, with the capacity of decreasing cars’ speed and increasing real safety.
Assessment of Shared Spaces
Methods of assessing shared spaces started to be developed recently. Therefore, there is no unified methodology, vocabulary or database. Moreover, most of these
14 assessments are done by academic parties, in partnership with local governments. An example is the work done by the already mentioned Shared Space Knowledge Center, from the NHL University of Applied Sciences, in which two methods of analysis are being tested. The first one is the use of surveys and interviews based on Gehl Institute parameters (Gehl, 2010), intending to evaluate the quality of the space through the perceived accessibility, traffic safety, clarity of the space, presence of places to sit, to meet, to play and the satisfaction with the design and greenery of the space (NHL Hogeschool, 2013, p. 36).
The second method of assessment is the observation of interactions between users, applying a variation of the DOCTOR conflict analysis method (to be explained in the next section) in which the speed and distance between users before and after the interaction, as well as the type and strength of the reaction, are observed (NHL Stenden Hogeschool, 2018).
As an example, an evaluation of three locations in the municipality of Mappel (NHL Hogeschool, 2012) aimed to assess if shared spaces have the outcomes expected with the design. Both methods were used and the main conclusions were that shared spaces indeed increase safety, even if the perceived safety decreases; that satisfaction with intersections is improved and vulnerable/ disabled users have more difficulty navigationg these spaces. Traffic Assessment
Although the concept of Shared Space and the ideas of creating cities on a human scale are by now well known by architects and designers, the traffic engineering field developed in a different direction. In general, the complexity of traffic systems and the need to satisfy a range of economic and political interests rendered the process of traffic assessment purely mathematical and based on quantitative research, modeling and forecasting (Cascetta, Armando, Francesca, & Marcello, 2015). Standard methods of traffic evaluation include on-going data collection of congestion and accidents statistics, to identify desire lines and bottlenecks (Cascetta, Armando, Francesca, & Marcello, 2015; Nulle, 2018).
When the redesign of a street is on-going, further and punctual researches can be done, using micro-simulation to forecast congestion improvement (Nulle, 2018) or Cost-Benefit Analysis (Beukers, Bertolini, & Te Brömmelstroet, 2012). In-depth and qualitative assessments are usually experimental or academic in nature, for example by measuring people’s perception and satisfaction (Gemeente Amsterdam, 2018 b; NHL Hogeschool, 2013). Conflict Analysis (CA)
First used by the car company General Motors in 1971 to distinguish conflicts in motorized traffic (Jong, Gysen, Petermans, & Daniels, 2007, p. 3; Schreurs, 2018, p. 11), Conflict Analysis (from now on CA) is considered an improvement in traffic evaluation since it does not need to wait for accidents to happen to identify safety issues (Laureshyn, Goede, Saunier, & Fyhri, 2017). In the Netherlands, the DOCTOR method (see Kraay, Horst, & Oppe, 2013) has being improved since the late 60’s and consists of measuring social interactions that may result in a negative consequence by observing users’ behavior (Kraay, Horst, & Oppe, 2013, p. 13). According to the Manual for Conflict observations:
15
“A traffic conflict is an observable situation in which two or more road users approach each other in space and time to such an extent that there is a risk of collision if their movements remain unchanged”. (Kraay, Horst, & Oppe, 2013, p. 14)
The main technique for this behavior observation is “on-site observations” (e.g. intersections) assisted by video footage (Kraay, Horst, & Oppe, 2013, p. 19). For the analysis, two parameters are used to classify the severity of a conflict situation. The Time-to-Collision (TTC) consists of estimating the speed of two users on a collision route if they do not change their behavior and calculating the time left for a reaction by one of them. The Post-Encroachment- Time (PET) measures the time between the reaction of one of the users and the end of the conflict situation (Kraay, Horst, & Oppe, 2013, p. 43). By analyzing both numbers, a trained observer classifies the conflict into five degrees of severity: (i) Carefully braking and anticipating. (ii) Controlled braking or swerving in order to avoid a collision. (iii) Strong braking resulting in a near-accident. (iv) Emergency stop resulting in a near accident or even a slight collision. (v) Emergency intervention followed by a collision (Wouterson, 2016, p. 16).
Cars A and B are on a collision course based on their current direction and speed. If neither car performs an evasive maneuver, then there will be a collision: there is a potential conflict point. The seriousness of the conflict increases at higher speeds and with a smaller distance to the potential conflict point.
potential conflict point
Figure 1: TTC scheme (Jong, Gysen, Petermans, & Daniels, 2007)
16
Limitations of the CA method
As recognized by the manuals, the method is subjective, depending on the judgement of the observer. Training is the most relevant means of increasing validity, but the calibration of the observations is fundamental. For this reason, international seminars to improve parameters have been held since 1979 (Kraay, Horst, & Oppe, 2013, p. 25). Another recognized limitation in validity is its application in places where accident rates are too low. “The question then is whether the method is transferrable to those situations.” (Kraay, Horst, & Oppe, 2013, p. 20).
A third limitation, formulated by this research based on the Shared Space concept, is the understanding that all interactions among actors are problems to be avoided. If some places present a low accident risk, the model could explore why, in these situations, the behavior of users leads to less risk. In addition, it could investigate how some “conflicts” might bring benefits to traffic and users.
Slow Traffic assessment
Slow traffic, in other words the displacement of bicycles and pedestrians, is usually not considered in traffic evaluation. For instance, congestion monitoring does not assess pedestrians and bicycles. The abovementioned CA is also questioned in its relevance to measure interactions at slow speed and with less risk (Schreurs, 2018, p. 11) since the model is based on motorized vehicles and, in most of the cases, the interactions between pedestrians and cyclists are classified in the category 1 or 2, left outside of the traffic evaluation (DTV Consultants, 2017; Gemeente Amsterdam, 2016).
If not relevant from a CA perspective, slow traffic probably results in positive social interactions, with the capacity of improving social benefits beyond traffic safety. Design guides already consider it differently than motorized transit, by advising that soft traffic regulations are enough to guarantee safety and flow.
“In the national [Dutch] design guidelines (CROW6), slow traffic conflicts are basically kept outside the traffic lights. As a result, for these slow traffic conflicts, the normal priority rules apply, such as right-front and the shark teeth on the road surface. The basic idea here is that slow traffic is mutually capable of regulating conflicts.” (Schreurs, 2018, p. 20)
Social Interactions (SI)
The concept of Social Interactions (from now on SI) is usually developed in the areas of economics and sociology. In the first field, particular interest is given to how there is an (increased) likelihood of an individual to take a particular action (buy a product, commit a crime, make an educational choice) when a “neighbor” made this same action (Glaeser & Scheinkman, 1999; Mote, 2000). In Sociology, interactions are scrutinized in their relevance to creating society. As stated by Georg Simmer, “human interaction not only gave rise to society but also met a basic human need to be social” (Ritzer, 2015, p. 164). In this field, interactions are usually understood as face-to-face or through digital media conversations with relatives, friends, coworkers or strangers in the street (Ritzer, 2015).
17 Although SI in momentary and dialogue-absent encounters are present in all moments of our lives, it only appeared in sociology after Erving Goffman advocated in the 70’s that “practices that might look banal in fact carried important connotations” (Jensen, Negotiation in Motion: Unpacking a Geography of Mobility, 2010 a, p. 392). According to him, “by moving in the city among buildings, objects and people, one interacts with the ‘environment’, making sense of it and ultimately producing culture and identity” (Jensen, 2010 a, p. 389). As an example, when analyzing pedestrians walking in the street, it is possible to observe the complexity of movements and brain calculations constantly used only to avoid bumping into others (Jensen, 2010 b, p. 337).
Interactions in Mobility and the Mobility Turn
Mobility Turn, attributed to the British sociologist John Urry (Olsen & Bækgaard, 2015, p. 9), consists of the inclusion of concepts from the sociology in mobility studies, expanding the mobility concept beyond the function of movement from A to B and including its potential to create culture, identity and social norms (Jensen, Negotiation in Motion: Unpacking a Geography of Mobility, 2010; Olsen & Bækgaard, 2015). For scholars from the mobility turn, Goffman’s concepts provided a range of tools and vocabulary to study interactions in mobile situations (Jensen, 2010 b, p. 333).
Ole B. Jensen, professor and researcher of traffic planning and design at Aalborg University, Denmark, applies this terminology to alternatively evaluate behavior of street users, differing from orthodox methods previously mentioned. In his research, instead of labeling encounter of actors in a public space as conflicts, he uses expressions such as “negotiation in motion”, “mobile with” and “staging mobility” to study patterns of interactions (Jensen, 2010 a; Jensen, 2010 b; Olsen & Bækgaard, 2015).
An important contribution from Jensen is the use of “ballet” and “river” metaphors to classify their interactions (Jensen, 2010 a, p. 394). The first one describes the “semiotics from below”, in other words, the relationship between users in the negotiation process (Olsen & Bækgaard, 2015, p. 15). In the “river” metaphor, the interest is in the connection between these users and the static objects from the landscape. Here, the “semiotics from above” is used to observe how elements from the built environment, such as traffic lights, signs, urban furniture and other objects, influence behavior (Olsen & Bækgaard, 2015, p. 15).
Problem statement
With the failure of car oriented urban planning to improve social conditions, combined with the shortage of public spaces in cities, the attributions of street and mobility must be expanded, including their role as meeting places, where known and unknown people interact, learn how to deal with diversity and ultimately generate culture and social norms. In this sense, what is the duty of street design and traffic regulations in encouraging meetings, stimulating people to interact and regulate themselves in transit situations?
As shown above, there are attempts in street design to put people at the center of the planning process. However, there are not enough tools to measure if these spatial layouts are actually producing the expected social improvements. Moreover, there is an inconsistency between Shared Space and CA concepts, since the first one encourages interactions and the second one only assess meetings that should be avoided.
18 If CA uses SI to detect unsafe situations (Kraay, Horst, & Oppe, 2013, p. 14), how can positive SI be studied to explore mobility benefits beyond safety, supported by the Mobility Turn? New methods of mobility assessment must be created to measure SI that are desirable, aiming to understand and enhance the aspects of the urban environment that create not only safe, but accessible, diverse and harmonious cities.
Research questions:
Based on the theories of shared space, traffic assessment and social interactions, the following research question aimed to give an answer to the problem stated above:
How are Desirable Social Interactions considered in the assessment and redesign of intersections and how can this be improved?
Three sub-questions were elaborated to explore the main one from different angles and clearly define the steps of this research. They will be answered with diverse methods and the combination of their results will provide a response to the major question.
- Sub-question 1: What are the current traffic design and assessment methods? - Sub-question 2: Which indicators define Desirable Social Interactions and can be
included in new methods of traffic assessment?
- Sub-question 3: When these new methods are included in traffic assessment, what can the outcomes be?
Conceptual Framework
Table 1: Overview of the Conceptual Framework
The main outcome of this research was to define and experiment with the concept of Desirable Social Interactions. In an attempt to create a bridge between theory and practice, it
19 contributed both to academic debates of Social Interactions in mobility and to practitioners working with traffic design and assessment.
Conflict Analysis is framed in this research as an observation method to access traffic where SI are understood as behaviors to be avoided. The strength of the method is the simplicity in the communication of the results, promoted by the five classes of conflicts, and the existing mechanisms to standardize and decrease bias. This is why the approach for defining Desirable Social Interactions aimed to keep the same structure and methodology as CA. However, the current classes do not qualify behavior, making it necessary to absorb other theories to develop new classes.
To address this gap, the concept of Social Interactions in mobility, theorized in social science and in the mobility turn, was added. The major contribution of this concept was to provide new lenses to observe the same interactions, with similar methods, but using terms to qualify behavior instead of “quantifying problems”, and exploring their social impacts beyond safety. Moreover, the focus of the observation corresponded to the ballet metaphor, since the aim was to assess only the interactions between users, similar to CA.
From this combination, Desirable Social Interactions (from now on DSI) is defined in this research as encounters between people in mobility that required negotiation between them and resulted in positive behavior. People here are pedestrians and cyclist, considered users of the public space. In comparison to CA, DSI analyzed the same interactions, but observed from the opposite perspective, focusing on the opposite ones. Since DSI is a new terminology, the research had also served the purpose of making this definition more accurate. Furthermore, its operationalization was done by answering the sub-questions.
The Level of shared space was used as the control variable. Since the concept of shared space encourages meetings between users, it was expected that, the higher the level of shared space, the more DSI would be encouraged. The term Level was added because the majority of intersections redesigns do not include all of the elements of Shared Spaces, but include some of its elements, depending on the questions and demands of the place. To systematize this classification, three variables were used: the amount of conventional traffic regulations, the amount of functions the intersection has and the modes of transport that share the space.
Slow traffic was used as the frame of the research. It can be defined as traffic involving only cyclists and pedestrians, or designs where cars are guests and the maximum speed is 30mk/h. This frame was chosen first because, as mentioned before, CA is limited in assessing these interactions; second because cyclists and pedestrians have the biggest potential to be involved in DSI due to their flexible way of moving (Olsen & Bækgaard, 2015) and absence of a “cocoon” to separate them from external interactions (Brömmelstroet, Nikolaeva, Glaser, Nicolaisen, & Chan, 2017); and third because of practical reasons, since it is harder to observe the behavior of car users.
20
Chapter 3 - Research Design
In order to answer the three sub-questions, several qualitative methods were applied. In the first part, the city of Amsterdam was used as a case-study to explore how traffic design and assessment include SI and its relationship with having elements of Shared Space in the layout. The second part consisted of the definition and operationalization of DSI though brainstorming sections with experts. The third and final part was an experiment to test the usefulness and feasibility of the new method of traffic assessment using DSI.
Sub-question 1: What are the current traffic design and assessment methods?
In order to answer the sub-question, an overview of current design and assessment methods was done using the city of Amsterdam as a case study. The research method applied was secondary data analysis in order to test the hypothesis Traffic redesign already includes DSI but subsequent studies don’t. Newsletters and official reports produced for and by the municipality of Amsterdam were the units of analysis. For the data collection, two main sources were used, the official webpage (amsterdam.nl) and non-structured interviews with civil servants. The expected outcomes were an overview of standard and experimental practices in intersections design, and an analysis on how and in which project phase (design and/or assessment) SI were included. From the secondary data analysis, it was not possible to understand the entire process of decision-making in design and assessment, meaning a semi-structured interview with Maarten Nulle, traffic designer at the municipality of Amsterdam, was required to complement the study.
Amsterdam as the Case Study
The city of Amsterdam is usually a pioneer in technological and social innovations. For instance, the Amsterdam Action Plan for Mobility: An Attractive and Accessible Amsterdam (Gemeente Amsterdam, 2013) states that traditional traffic policies, “strongly focused on accessibility in the sense of being able to get somewhere quickly” are no longer the municipal guiding principle and the concept of “city street” should be used to combine functions of traffic and leisure in the same space.
The consequences of this municipal approach are innovative attempts to include DSI in the redesign of intersections. One example is the partnership with the Social Science department of the University of Amsterdam (UvA - AISSR) to study cyclists’ desire lines (University of Amsterdam (AISSR); Copenhagenize Design Co., 2014) culminating in intersections’ redesign in accordance with users’ behavior and the creation of a toolbox with solutions to be applied in future projects (Vernooij & Amsterdam, 2017). Another example of this innovative thought is removal of traffic lights in the Alexanderplein (Glaser, 2017) and the implementation of Shared Space behind the Central Station (Gemeente Amsterdam, 2016), showing a critical review of conventional traffic practices when they no longer match local requirements.
In addition to the appropriate innovative environment, the accessibility of secondary data through internet, and also availability and willingness of civil servants to contribute, made Amsterdam an appropriate case study.
Method of Data Collection and Analysis:
The method of data collection was through non-structured interviews and e-mail exchanges with traffic designers and modelers from the municipality, in addition to harvesting
21 the official webpage (amsterdam.nl). The selection of the projects followed the criteria of being finished approximately in the past three years and being under construction or in the design process. By snow-ball sampling, other civil servants and projects were accessed until there was a saturation of the information. In the Appendix 1.1 there is a list of the practitioners consulted.
To conduct the analysis, the projects were classified first by their Level of Shared Space (Table 2). The variables Conventional Traffic Regulations and Diversity of Space Functions represent scores from 1 to 4 and A to C respectively, being 1-A a completely Shared Space and 4-C a space with no traces of the concept. The analysis also included the modes of transport sharing the space, since framing slower traffic was relevant to this research.
Table 2: Levels of Shared Space. (created for this research)
The analysis of the inclusion of SI was done with two variables, the Phase of inclusion - Design or Assessment – and the Quality of SI – DSI (positive) or – CA (negative) (table 3). Designs that included SI are the ones where less conventional traffic regulations forced a shift of users’ attention. For example, instead of looking for traffic lights to understand right of way, users had to “read” and “communicate” with others to negotiate this right. Assessment stands for projects that included interactions in the evaluation process, before or after the reconstruction. They could be DSI, such as doing observations or interviews to understand users’ behavior and ways of interacting; or CA when meetings were considered negative.
Table 3: Inclusion of Social Interactions (created for this research)
Sub-question 2: Which indicators define DSI and can be included in new methods of traffic assessment?
The aim of the second part of the research was to explore possibilities of including DSI in traffic assessment by finding indicators to operationalize the method and make it useful and feasible for practitioners. To accomplish this goal, qualitative methods, divided into three phases, were used. First, the analysis and coding of three previous studies on SI in mobility and behavior of users in shared spaces was done to extract potential indicators for DSI and their meaning. The second and most relevant part was the conduction of semi-structured interviews with four scholars who are experts in the behavior of cyclists and pedestrians and in observation methodology. The aim here was to brainstorm, starting with the collected indicators, methods to operationalize DSI. The outcome was an observation table, a vast list
22 of possible elements to be observed. In the analysis process, this table and the reflections were used to define theoretical classes of DSI. The third phase was a public survey with cyclists and pedestrians, worldwide, to measure the relevance of the described classes.
The units of analysis in this sub-question was then previous researches in SI on mobility and behavior of pedestrians and cyclists; academic researchers in SI and observation methodologies; and cyclists and pedestrians from different countries. The expected outcome of this part was an experimental toolbox to assess DSI, similar to the Conflict Analysis format, including the classes of interaction and the indicators for operationalization.
Data collection – searching previous studies
As expected, the search for previous studies resulted in few examples of analyzing SI in mobility. The most relevant literature was in experimental methods, combining observations and interviews, to explore users’ behavior, interactions, and the influence of shared spaces. However, the majority of studies analyzed interactions to be avoided, conflicts.
Kaparias et. al (2015) focused their research on the before (not shared) and after (shared space) design of Exhibition Road in London. CA was used to observe interactions between pedestrians and motorized vehicles through videos. The analysis was made by categorizing the events in Steady Car-pedestrian (SC-P), in which cars’ behavior was not due to the interaction and Effective Shared Space (ESS), in which actors negotiated the space. The main behaviors observed were change in pace and direction (Kaparias, Bell, Biagioli, Bellezza, & Mount, 2015). Notable conclusions included that in designs in which clear priority is given to pedestrian, their willingness to give way increases; and also that shared space design decreased the number of interactions by 30%
Ole Jensen used Erving Goffman’s theory to analyze SI in motion and conducted an ethnographic observation in Aalborg, Denmark focusing on patterns of interaction between different modes of transport (Jensen, 2010 a). He observed three situations: frontal meetings, more relevant to analyzing interactions between pedestrians; orthogonal meeting, interesting between pedestrians and others vehicles; and parallel meeting, relevant to assess pedestrians and buses meetings. Dividing interaction into the ballet and the river perspectives, the main behaviors analyzed were body language, eye contact, changing direction to give room, speeding-up to join a crowd that is already crossing before vehicles start moving, taking a step back, moving parallel with the sidewalk before crossing. Jensen concluded that interactions between pedestrians and bicycles, different from pedestrians-pedestrians, are “a matter of a ‘reading’ and evaluating body language, not so much in direct negotiation but more as an estimation of the situation.” (Jensen, Negotiation in Motion: Unpacking a Geography of Mobility, 2010 a, p. 397). Another interesting outcome was the creation of the frame Staging Mobility to conceptualize SI in motion and new term to describe these interactions.
Liam Wouterson, in his Master’s Thesis named Stage fright? The accessibility of shared spaces (original in Dutch) (Wouterson, 2016), studied the impacts of different degrees of traffic regulations on the mobility of elderly people. He also combined surveys, interviews and observations to triangulate his results. In the observation method, he focused on the change in speed – stopping, anticipating (slowing down/ speeding up) and continuing without
23 changes – to analyze both cyclists and elderly pedestrians’ behavior in the shared space behind central station. His conclusions were that all users stop less in the deregulated space, showing that shared spaces promote a better flow of traffic; cyclists have clearly more patient with and give more room to vulnerable pedestrians; in addition, elderly people anticipate less, but do this more often in shared space than on the zebra crossings.
Data collection – selecting the experts
In the second part, four experts from different countries were interviewed. In this case, experts were academics specialized in studying pedestrian and cyclist behavior, use of observation methods and analysis of interactions at intersections and in shared spaces. The appendix 2.1 shows the structure of the interviews. Overall, the questions were organized in three parts: general possibilities of assessing SI, exploring the observation methodology for DSI, and questions related to their own research and technique. Since each academic has a different research approach, the results of one interview were used in the following ones and the interview structure was different for each expert.
The first interview, conducted on March 16th 2018, was with Ole B. Jensen, Professor and Researcher at Aalborg University in Denmark in the areas of Urban Theory and Urban Design. He is a researcher of Goffman’s methods of behavior analysis and responsible for the frame of Staging Mobility previously explained. The meeting was early in the research timetable because the opportunity presented by Jensen giving a lecture in Amsterdam was taken.
Sjoerd Nota, the second expert, is Traffic Advisor and Researcher at NHL Hogeschool in Leeuwarden, the Netherlands, and was interviewed on April 11th 2018. His studies are related to traffic psychology, analysis of design quality in shared spaces and observation methods to access users’ behavior in these areas.
Gerben Moerman is a professor and researcher of qualitative methodologies at the University of Amsterdam, the Netherlands, and also fascinated by the sociological aspects of cycling. The interview was conducted on May 18th 2018, when the draft of the observation table was more concrete, allowing more thorough reflection on the indicators collected up to that point then and brainstorming of new possibilities.
The last interviewee was Sebastian Peters, postdoctoral fellow in the Faculty of Landscape and Society at the Norwegian University of Life Sciences NMBU. In his paper “Sharing space or meaning? A geosemiotic perspective on shared space design” (Peters, 2017), he used ethnography to explore how users understand shared spaces and how the behaviour of some can influence others. He was interviewed via Skype on May 22nd 2018. Data collection – public survey
After the analysis of the data collected in the interviews and the operationalization of DSI in classes, similarly to the CA concept, an on-line public survey was carried out to validate the classes and assess their relevance to pedestrians and cyclists in general. A Google Forms format was used and the survey was diffused using the social media WhatsApp, Facebook and LinkedIn.
24 The survey consisted of only one question to be answered from two perspectives, pedestrians and cyclists. The alternatives were written in an accessible vocabulary and the format was kept simple to reach a broad public in different places and urban contexts. The question and alternatives will be explained in the chapter 4, as part of the data analysis. Data analysis – a continuing process
The data analysis of this phase was a continuous process. First, the three case-studies were coded and the main elements used in their observations and the respective interpretations were extracted. With these in hands, the experts were interviewed, giving new insights and presenting similar researche. After each interview, the recorded audios, as well as the extra academic papers introduced, were coded. All the data was collected in a draft of the observations table and used during the next interviews. Only after listening to the third expert was a second table drafted, containing the classes of DSI, as will be shown in the analysis chapter. After the public survey, the two final tables were organized and used in the experiment, the final part of the research.
Sub-question 3: When these new methods are included in traffic assessment, what can the outcomes be?
The third and final part of this research was an experiment to test the usefulness and feasibility of including DSI in traffic assessment. The method applied was a comparative study (Bryman, 2012) via observation of video footage in two places with different levels of shared space, framing the interactions between pedestrians and cyclists and using the two tables from the previous part of the analysis. In order to bring the research closer to practice, two case studies were chosen and a hypothesis was formulated and tested. In this last part, the unit of analysis are pedestrians and cyclists involved in SI.
The hypothesis tested was: at intersections with a higher level of shared space, both cyclists and pedestrians Anticipate and Adjust more, in comparison with a lower level of shared space. This is expected because, with more space to move and without a point that concentrates the interactions (zebra crossing), users have more room to anticipate and adjust, creating a smoother flow. Moreover, previous studies on shared space already indicated this behavior (Kaparias, Bell, Biagioli, Bellezza, & Mount, 2015; Wouterson, 2016). The relevance of this experimentation is justified by Amsterdam’s new trends of street design, where the traditional traffic norms of segregated space for traffic and leisure are being replaced, due to the lack of space and changes in urban priorities. The municipality is now encouraging the implementation of city streets, a design that can “fulfil many public functions and cater for plenty of passing and local traffic” in one place (Gemeente Amsterdam, 2018 a, p. 30).
Choosing the Case Studies – Koningsplein
Koningsplein, a square in the city center of Amsterdam, was chosen as a case study for several reasons. The selection of the case started from the analysis table from the first part of the research (see appendix 1.2). Although the Koningsplein was not recently renovated, previous studies showed that “the behavior of users gives to square a bigger level of shared space, once pavement regulations are constantly ignored and users rely on SI to navigate the
25 square” (Andriessen, 2011, p. 39). In addition, the current configuration scored relatively high, similarly to recently redesigned squares such as Leidseplein and Muntplein. The
advantage of the Koningsplein is the presence of one intersection with a zebra crossing and one without, allowing a compative study. In addition, there was the possibility of recording the videos from a building close by.
The current design was classified as 2-A (see chapter 4 – Sub-Question 1), which means that some conventional traffic regulations are present, such as pavement marks and traffic signs, but there are no traffic lights. The traffic segregation is considered soft, since the sidewalk and the road are the same color and, although there is a difference in level, the configuration of the square and bicycle parking gives rise to some ambiguity regarding where exactly the sidewalk is. Moreover, several other activities are encouraged in the area by the presence of benches, food kiosks, commerce at the sidewalk level and the flower market.
The two intersections chosen for the case study were the Heiligeweg & Singel (odd side) and Koningsplein & Singel (even side). The first one has a zebra crossing, which was considered the frame for the observation. Since the cars are banned in this part of Singel, all displacements there are Slow Traffic. The second intersection does not have a zebra crossing and the movement of users can happen at different points. For this reason, a wider area was considered in the observations, as shown in the figure 4. Car traffic is limited and the maximum speed is 30mk/h, which means that most of the traffic is also slow.
26
Figure 4: boundaries of analysis + place of video recorded
Data collection – videos footage
The data was collected by me, the researcher, at three different periods on the same day, by using a smartphone camera from the second floor of a building in the square (image 4). Five minute videos were recorded in both locations, resulting in six videos and a total of 30 minutes of data. The date and time of the recording was:
Day: 08/06/2018
1. Morning: from 8:45 to 9:00 2. Midday: from 12:50 to 13:05 3. Afternoon: from 18:20 to 18:35
The weather on that day was rainy but warm (approx. 20oC), which was an advantage in the morning and afternoon periods, because it was not raining and people were not wearing sunglasses. However, in the midday footage it was raining and several people were carrying umbrellas or wearing raincoats, compromising the observations.
Method of data analysis – domain analysis
The analysis of the material was done by observing the videos in slow-motion, at a speed of 0.250x, and describing users’ behavior using the table and classes of DSI (a sample of the description can be found in the appendix 3.1). The analysis was divided into two parts. First, two minutes of each video were observed using Spreadley technique for conducting ethnographic research (Spradley, 1980, p. 85). The movement of users was described before, during and after the interaction and, after the observation, a class was given to each user
27 involved in the interaction, a technique advised by Sebastian Peters1. Next, the texts were coded using Spreadley domain analysis, as illustrated in the table 4. All terms describing behavior were related to its DSI class, resulting in a toolbox for future observations, a table with the most common descriptions (see chapter 4, table 6).
Table 4: example of domain analysis (based on Spradley, 1980)
In the second part, this toolbox was used to analyze the other three minutes of each footage, now without the description of the movements, making the process much faster. A quantitative approach was taken, similarly to the CA method, by counting the types of DSI and drawing comparisons statistically. It is important to be aware that the focus of this experiment was to test the methodology and not to draw final conclusions, since the sample was not representative for a statistical analysis.
1“We decided to observe the square and collect interactions. It was a bit of an ethnographic research, we were
describing the meeting of people in different modes of transport and seeing how they react to each other” (Peters, Interview for this research, 2018)
28
Chapter 4 - Analysis:
After the presentation of the methodology, this chapter will introduce the analysis and results for the three parts of the research. The first section will be an overview of how design and assessment includes SI in recent traffic projects; in the second part, the theoretical model of assessing DSI in mobility will be presented; and the third and last part will explore the real feasibility of the method.
Sub Question 1 - The inclusion of DSI in design and assessment. The Amsterdam Case As explained before, the first phase of this research tested the hypothesis traffic redesign already includes DSI but subsequent studies don’t by analyzing redesigns in their Level of Shared Space and Inclusion of SI. After the collection of official reports and newsletters, the projects were organized in a descriptive and analytical table (appendix 1.2).
Overall, Amsterdam does not present projects with a fully implemented shared space. Although, several projects scored high, such as the Leidseplein, the Muntplent, the Ferdinand Bolstraat and the Koningsplein with 2-A, and the De Ruijterkade with 1-B. In addition, in all conventional street design (where the score is 4-C) all zebra-crossings in the bicycle path scored 2-C. This is a consequence of the national traffic design guideline CROW6, which determines that pedestrian traffic lights do not need to include these zebra-crossings (Schreurs, 2018, p. 20). Since the concept of Shared Space implies an increase in negotiations among users, it is expected that these same projects will include more DSI in the design and assessment.
Most of the projects included in the analysis were considered pilot ideas, in which the municipality was exploring designs and assessments outside the handbook of traffic norms. There was, however, a lack of documentation in standard redesign projects, probably because they are less well advertised and inspire less curiosity. In addition, the unique characteristics of Amsterdam city center, with a shortage of space and large variety of functions, raise questions that cannot be answered by conventional traffic rules. Moreover, as the next paragraphs will show, no assessment is done in standard reconstruction projects.
The standard methods of design and assessment
An interview with the traffic designer Maarten Nulle (2018) was conducted to understand the conventional methods of traffic design and assessment, what are the main elements in the handbook and in which situations designers make decisions not included in the book.
According to him, the decision to reconstruct a street is done by the end of its life-cycle or if pavement conditions are no longer good. The municipality permanently conducts inspections of roads, monitors traffic flow, delay and congestion and keeps data on infrastructure conditions. When a road needs to be renovated, designers use this database to check if the current layout accommodates the existing functions and remains safe. Overall, other types of evaluation, such as CA and traffic counting, are not done by default either before or after constructions and are required only when considered necessary. Another example of extraordinary research is congestion forecasting in intersections, using micro-simulation software and testing different scenarios, such as with or without traffic lights.
29 According to Nulle, the “limitation [of the software] is that you cannot predict behavior, you cannot program a car ignoring the traffic light” (Nulle, 2018).
In relation to the design, there is not one standard model, there are different requirements according to the function of the road. Depending on the type of street, residential, 30km/h road or 50km/h avenue, the handbook presents different measurements and pavement specifications. Still, sometimes the handbook alternatives do not solve the problems detected in the area and, in these cases, the designer proposes new alternatives. However, Nulle explained that going away from the handbook requires a concise argumentation on why the conventional methods do not solve the problem and how the proposed one is the best alternative. As will be explained, the redesign of the De Ruijterkade into a shared space is one example of these exceptions.
Experimental projects and the inclusion of shared space and SI
As expected, traffic projects more frequently include DSI in the design phase. Although the description of the projects did not make reference to the term interaction, the analysis showed a concern regarding improving slow traffic and a trust in users capacity to regulate themselves in some situations. In relation to the assessment, some experiments in the inclusion of DSI were detected, but still very premature and not directly addressing ways and qualities of negotiations. The most relevant attempts were by using surveys, first on the Alexanderplein, exploring users’ perception of the intersection before and after the traffic lights were turned off (Glaser, 2017), and second in the KIM project evaluation, where users were asked about what they rely on to prevent a conflict.
The table 5 presents a summary of the main projects the municipality is involved in and their inclusion of SI and level of Shared Space. From all redesigns conducted by the municipality in the past years, four main projects can by highlighted by their innovative methods of design or assessment.
30
Table 5: Overview of Amsterdam Traffic Redesign Innovative Plans
De Ruijterkade – Shared Space behind Amsterdam Central Station
The first project is the De Ruijterkade, known as the Shared Space behind Amsterdam Central Station. The main question addressed by the redesign was the confusion regarding right of way, causing several accidents.
“Because of the high traffic intensities, the feeling of priority over two conflicting bicycle routes, and the high number of crossing pedestrians, the concept of 'shared space' has been chosen as the optimal solution for safely unwinding the crossing traffic flows.” (Gemeente Amsterdam, 2016, p. 3)
The outcome was an unified gray pavement with no marks of priority (figure 5). Although it is the best-known case of shared space in Amsterdam, it scored 1-B in the analysis. This is because there is a poor diversity of functions in the area, proposed mostly as a traffic area. DSI are integral in the design, since users can only rely on each other to negotiate right of way and space for turning. In the assessment, the municipality and the DTV consultancy, conducted CA research during the first three months after reconstruction
31 (situations 1, one week after reconstruction, situation 2, four weeks after reconstruction and situation 3, 12 weeks after reconstruction), measuring classes 3, 4 and 5 as well as cause, location and mode of transport involved in the conflict. In addition, traffic flows and routes were evaluated using tracking sensors to define 'desire lines' (Gemeente Amsterdam, 2016, p. 3). According to authorities, the study was only ordered because it concerned a new traffic concept, in a busy and complex intersection (Gemeente Amsterdam, 2016, p. 3). The final report concluded that non unexpected behavior occurred and, in 38 hours of video analysis, only 25 conflicts in the classes 3, 4 or 5 were detected . Moreover, the new redesign fulfilled the expected requirements of better regulating transit and decreasing the number of accidents(Gemeente Amsterdam, 2016, p. 3).
Figure 5: De Ruiterkade. Shared Space behind CS (photo Mats van Soolingen)
Kleine Infrastructurele Maatregelen – KIM project
Winner of the Tour de Force Innovation Prize 2017 (Nationaal Fiets Congres, 2017), the KIM Project started with a partnership between the municipality of Amsterdam, the department of Social Science at the University of Amsterdam and the Danish company Copenhagenize (University of Amsterdam (AISSR); Copenhagenize Design Co., 2014). The goal was to study cyclists’ behavior and apply it to the redesign of intersections, through the analysis of desire lines. The outcome was an initial study of 10 intersections in Amsterdam, the redesign of three during the first stage and the creation of a toolbox to be implemented in future redesigns (Kees Vernooij, 2017).
The analysis of the project showed that, although the intersections as a whole are not shared spaces (4-C), the bicycle paths redesigned are considered 2-C, due to the lack of traffic lights for pedestrians. The inclusion of SI was done preceding assessment, since the desire lines represent interaction according to the “river perspective”. By incorporating the interactions of users with the physical elements of the environment in the design, it also
32 includes SI. In the evaluations after constructions, further methods of assessing SI were implemented, also in an experimental character. A survey with users in three of these intersections evaluated the satisfaction with the area as well as the techniques used to avoid conflicts with others. The final report showed that 57% of respondents felt an improvement in the intersections, the majority did not experience conflicts and, when they did, more than 50% looked at other users to avoid them, instead of using pavement marks (Gemeente Amsterdam, 2018 b). Further evaluations of this project were the study of cyclist distribution and flows, using cameras and heatmaps in the before and after scenarios (Gemeente Amsterdam, n.d. [c]).
Figure 6: before and after situation of Mr. Visserplein (source: Beeld en Data)
Binnenring – Inner Ring Project
The Inner Ring is an on-going project of redesigning the streets traversing the city center. The ring is composed of the streets Sarphatistraat, De Weteringschans and Marnixstraat, and the squares Alexanderplein, Frederiksplein and Leidseplein (Gemeente Amsterdam, n.d. [d]). The project is divided into several sub-redesigns, each one with different questions, resulting in diverse levels of shared space and inclusion of SI. Overall, the main objectives of the redesigns are to increase the capacity for bicycles and improve bicycle flow. For this purpose, the three streets were redesigned according to the concept of "cars are guests", which means a maximum speed of 30km/h and the entire road being shared by car drivers and cyclists.
