Procedia Social and Behavioral Procedia -Social and BehavioralSciences 00 (2011) 000–000 Sciences www.elsevier.com/locate/procedia Conference Title (Transport Research Arena– Europe 2012) Planning of the Athens metropolitan cycle network using participative multicriteria GIS analysis Dimitris Milakisa*, Konstantinos Athanasopoulosa, Evangelos Vafeiadisa, Konstantinos Vasileiadisa, Thanos Vlastosa Sustainable Mobility Unit, National Technical University of Athens, Heroon Polytechniou 9, Zografos Campus 157 80, Greece Abstract Although the intention to use the bicycle in Greek cities is relatively high the modal share is still quite low. This is mainly due to lack of infrastructure. The Greek Ministry of Environment in recognition of this deficit funded a research for the development of the metropolitan cycle network of Athens with a view to include it in the new master plan. The paper presents the methodology and results of this research. The aim was to plan a network that will cover daily activities, while providing safety, directness, convenience and attractive routes. Strategic choice was to avoid a strictly top-down approach; hence the routes were selected by applying participative multicriteria analysis. Our methodology is a way to give a meaningful role to end-users and involve them in important decisions for the city’s future. Moreover their participation was helpful to determine what is important from the cyclist point of view concerning cycle routes, increasing the possibility to use them. © 2011 Published by Elsevier Ltd. Selection and peer-review under responsibility of TRA 2012 Keywords: Athens-Greece, metropolitan cycle network planning, participative multicriteria analysis, route choice analysis 1. Introduction In Athens, although the number of cars has exceeded its capacity, resulting in traffic saturation, pollution, noise, delays and degradation of public space, transport planning remains auto-oriented. The increase of traffic and parking space availability is presented as obvious solution, leading to a systematic reduction of the space that normally belongs to pedestrians, cyclists, and public transport. It is quite * Corresponding author. Tel.: +30-210-7724213; fax: +30-210-7722752. E-mail address: milakis@mail.ntua.gr Author name /Procedia -Social and Behavioral Sciences 00 (2011) 000–000 /Procedia -Social and Behavioral Sciences 00 (2011) 000–000 interesting that the length of highways increased by 580% between 1990 and 2007 (Eurostat, 2010), while the recent funding of 7.4 billion Euros, Greece has received from the European Cohesion Funds for the transport sector (2007-2013 period), were allocated mainly to new road constructions (59%) and only 19% to rail (MoE, 2008). Very little money regards projects of sustainable mobility in urban areas. As a result the increase of GHGs emissions from road transport was 71% in Greece during the period 19902006 and only 27% in EU-27 (EEA, 2009). The main victims of this car-oriented culture are the most vulnerable road users such as children, elderly and people with disabilities. That also deeply hurt the social cohesion and public character of the city, and negatively affected the number of its visitors. Today, car is unable to exert its undeniable charm in Athens, just because it does not fit and its functional benefits are voided. Keywords of the current needs are to save space, reduce energy consumption and pollution complying with specific commitments to the international community. One of the most popular alternatives is cycling. The bicycle attracts more and more European citizens with its charm, as they discover a different lifestyle, pleasure, health, joy, vitality and freedom. The everyday routine acquires color, becomes more friendly, gentle, and the city looks more beautiful. For many people cycling means socializing and new friendships. A recent survey in one of the largest Greek cities (Volos) found that the Greeks now recognize these advantages and they express a high intention to use the bicycle for their daily activities (Papavasileiou et al., 2010). It is also clear that due to the lack of infrastructure citizens are very reluctant to manifest this intention. The Greek Ministry of Environment in recognition of this deficit funded a research regarding the development of the metropolitan cycle network of Athens with a view to include it in the new master plan of the city. The planning of the metropolitan cycle network was a major challenge for our research team. The extent of the study area, the variety of parameters that should be taken into consideration, the unique urban and transport characteristics of Athens and the strategic choice to avoid a strictly top-down approach defined an interesting and challenging research landscape. The methodology developed in response to these challenges, resulted, for the first time in Athens, in a comprehensive proposal for a metropolitan cycle network consisting of 36 major routes (226 km in length). The structure of the paper comprises a review of the literature on metropolitan cycle networks planning (chapter 2), the application of our methodology and the results (chapter 3), and the conclusions in chapter 4. 2. Background While the bicycle is quite popular in medium-sized cities around the world, it fails to attract large shares of trips in big cities. The obvious reason is that trips in major cities are longer and as a result the car becomes a more popular choice. This in turn leads to increased traffic congestion and reduces the sense of safety for cyclists. The systematic solution given to this vicious circle problem is the combined use of bicycle and public transport, and the construction of large scale metropolitan cycle networks that makes cycling a safe, pleasant and fast option. The cycle network planning methodologies can be divided into two groups: supply-and demand- based (Rybarczyk & Wu, 2010). The first focuses on designing cycle infrastructures in all those roads that cyclists may be using, with particular emphasis on the safety and ease of movement. The second adjust network planning on the demand for cycling trips. To this end, stated preference surveys are used (Hopkinson & Wardman, 1996). The recent advancements in recording, mapping and analyzing the spatial attributes of the preferred cycle routes using GPS and GIS software (Menghini et al., 2010) have fostered more recent methodologies, which seek to integrate these two approaches (Larsen & Geneidy, 2009; Rybarczyk & Wu, 2010). GIS in conjunction with multicriteria analysis is one of the most frequently used methods, which offer ease of procesing and quantifying qualitative variables involved in Author name / Procedia -Social and Behavioral Sciences 00 (2011) 000–000 / Procedia -Social and Behavioral Sciences 00 (2011) 000–000 both approaches. It is also easier now to simultaneously assess factors that describe the road characteristics and the route preferences of cyclists. Multicriteria methods also provide an advantage that has yet to be exploited in cycle network planning. They can "open" the scientific procedures to the public (Stirling, 2004), affect the choice of routes and finally synthesize the apparently contradictory requirements of a scientific approach and a process open to the public. A typical method is the Multiattribute Utility Analysis (Merkhofer et al., 1997; Schwartz & Eichhorn, 1997) in which the public participates in the formulation of evaluation criteria and the definition of their weights. An even deeper analysis of the public views is offered by the Deliberative Mapping method (Davies et al., 2003). A common feature of all methods that combine multicriteria analysis and public participation is that they seek consensus regarding the weight of each criterion, avoiding to use average values. In the case of the Athens metropolitan cycle network we applied multicriteria analysis with deliberation and the help of GIS software regarding the quantification and mapping of the road environment characteristics, and the cyclists’ routes preferences. In this way we succeeded to gain a deeper insight into cyclists’ choices and plan the network accordingly. It is also very important that cyclists felt that they contributed decisively to the design of the network and therefore they promoted the research results to the public through their websites, web forums, newspapers and public discussions. 3. Methodology application and results The starting point of the research was the investigation of the European experience on issues regarding metropolitan cycle network planning. Eight European cities, which have developed such networks, were examined (London, Copenhagen, Dublin, Berlin, Amsterdam, Barcelona, Paris, Zurich). We analyzed the urban and transport framework of each city, focusing on the planning principles, the technical characteristics and the accompanying promotion policies (e.g. bike-sharing systems). The results formed a useful reference point for our research, especially for important decisions such as the shape of the network, the road types and categories, the basic design principles and the types of infrastructure. In the next phase, we developed a methodology leading to a proposal for a legible, functional, safe and pleasant metropolitan cycle network. Strategic choice was to avoid a strictly top-down approach and therefore the views expressed by experienced cyclists were integrated in specific methodological procedures contributing substantially to the final outcome of the project. In this way we tried to give a meaningful role to end-users and involve them in important decisions for the future of the city. Moreover their views were helpful to determine what counts for them, increasing the attractiveness of the network. Specifically, our methodology comprises four successive steps (figure 1). Fig. 1. The methodological steps of our research 3.1. 1ststep: Identification of the key metropolitan poles The aim of the first step was to identify the major urban poles of Athens (producing and attracting most of the daily trips) to become the nodes of the metropolitan cycle network. Eight metropolitan poles were detected based on criteria as (a) the spatial allocation of specific land uses (retail, offices, education, Author name /Procedia -Social and Behavioral Sciences 00 (2011) 000–000 /Procedia -Social and Behavioral Sciences 00 (2011) 000–000 leisure), (b) the O-D matrices of the two most recent travel surveys (AM-DPMG, 1997; AUTO, 2007), and (c) the hierarchical network of centers in the Master Plan of Athens (OJHR, 1985). The zones connecting the metropolitan poles are the areas along which the cycle routes scenarios were developed (see figure 2). Our objective was to make all possible connections between the major poles, unless topography did not allow it. Finally, 14 zones were identified. A maximum width of 1.5km was attributed to each zone, thus ensuring that cycle routes scenarios would meet the criterion of directness for all couples of poles. Then, specific categories of urban activities with a special interest for cyclists, were analyzed in order to examine whether they are included in the 14 initial zones. According to the results the coverage varies between 50-94%, which was very satisfactory, as critical activities such as universities, shopping centers, administration and entertainment zones show high rates (see table 1). Table 1. The percentage coverage of activities related to bicycle by the proposed cycle routes zones. Urban activities Universities Entertainmentzones Cultural poles Shoppingcenters (malls) Linear centers Parks/openspaces Sportsfacilities Ministries/ Publicservices Hospitals Publictransport stations % of cycle related activities included in the cycle routes 94% 69% 83% 86% 72% 65% 50% 82% 55% 54% zones Athens Piraeus Fig. 2. (Left) The 8 metropolitan poles, the 14 zones where cycle routes scenarios are developed and the spatial distribution of the educational and recreational activities. (Right) An example of four cycle routes scenarios between the centers of Athens and Piraeus. 3.2. 2nd step: Development of cycle routes scenarios The aim of the second step was the specification of the cycle routes scenarios within the spatial context of the 14 zones defined in the previous step. It is quite obvious that in a broad zone of 1.5km there exist more than one alternative. So, we followed some basic planning principles, relating to safety, directness, comfort and quality of urban environment in order to define the initial cycle routes scenarios. Additional scenarios derived by (a) the Greek Cyclists Organization early proposals (Greek Cyclists, 2008) and (b) a workshop with 10 experienced cyclists organized by the research team. In this workshop cyclists were Author name / Procedia -Social and Behavioral Sciences 00 (2011) 000–000 / Procedia -Social and Behavioral Sciences 00 (2011) 000–000 asked to evaluate the research team scenarios or even suggest the routes that they use for their commuting trips. Finally, 46 cycle routes scenarios in 11 zones (about 4 scenarios/zone) were identified. In the remaining 3 zones there was no need for alternatives because the advantages of one scenario were really strong (e.g. along the shoreline or a stream). 3.3. 3rdstep: Evaluation of the cycle routes scenarios The aim of the third methodological step was the evaluation of the cycle routes scenarios and the selection of the optimum ones for each zone, based on specific criteria. In this way a first network scheme was designed. Multi-criteria analysis with the participation of experienced cyclists in (a) the criteria formulation and (b) the weight-giving process was applied in order to identify the optimum routes. More specifically, the steps followed were: • Selection of the evaluation criteria. • Quantification of the criteria per cycle route scenario using GIS software. • Determination of the criteria weights and application of multicriteria analysis per zone. Identification of the optimum cycle routes. • Configuration of a draft plan for the metropolitan cycle network. 3.3.1. Selection and quantification of the evaluation criteria The research team proposed the main group of the evaluation criteria and then asked the experienced cyclists to comment, reject or add some more. The general principles, which were used for the initial selection of the cycle routes scenarios such as safety, directness, comfort and quality of urban environment, were discussed. 11 criteria were finally selected: • Energy Consumption: The energy consumption for a cyclist is principally affected by two parameters: the length and the slope of the route. We calculated this by dividing each route into sections with different slope. Energy consumption is also affected by the coefficient of friction of the surface material. However, this calculation was beyond the scope of this research. ÅC = ÓLiSi [Li: length of route section i (km) | Si: slope of route section i (%] (1) • Junctions Density: It describes the level of safety and continuity of a cycle route. It was calculated as the quotient of the number of nodes and the length of a route. We defined a node as the intersection of a route with a road of equal or higher hierarchical level. The signalized nodes was taken into account at 50% level, given that cyclists in roughly half of them will pass during the green period. JD = number of nodes/route length (2) • Traffic Intensity: It describes the pressure and discomfort that a cyclist feels due to the traffic. The calculation is based on the hierarchical level of the road. Three levels of traffic volume were defined (low| 10 -medium| 5 -high| 0). Thus, motorways and primary arteries correspond to high traffic volume, secondary and collector roads to medium and local roads to low. TI= ÓLiTi /ÓLi [Li: length of route section i (km) | Ti: traffic volume level of route section i] (3) • Traffic Speed: It refers to the average speed of vehicles on a route and therefore the feeling of insecurity caused to cyclists. Three levels of traffic speed were defined (low| 10 -medium| 5 -high| 0). The calculation is based on the geometrical characteristics and the congestion level of the road. TS= ÓLiSi/ÓLi [Li: length of route section i (km) | Si: traffic speed level of route section i] (4) Author name /Procedia -Social and Behavioral Sciences 00 (2011) 000–000 /Procedia -Social and Behavioral Sciences 00 (2011) 000–000 • Legibility: It describes how easily a route is imprinted in the cyclists’ mind due to its form. High legibility makes the route more friendly and the cyclist less dependent to technical assistance (maps, GPS). It is calculated as the quotient of the number of directional changes (deviation from a straight line) and the length of a route. L = number of direction changes (breaks)/route length (5) • Natural Environment: It describes the degree of presence of natural elements (e.g. stream, coastline, forest, urban park) along the route. Three indicator levels were defined (Important contact| 10: route in landscape or in large parks, medium contact| 5: route beside urban parks or streets with dense hedgerows, little contact| 0: route along a road with minimal presence of green spaces). NE=ÓLiNi/ÓLi [Li: length of route section i (km) | Ni: degree of presence of natural elements in route section i] (6) • Built Environment: It describes the quality of urban-architectural environment along a route. Three indicator levels were defined (High quality| 10: route in high quality urban-architectural environment, medium quality| 5: route along railway lines, tramlines, pedestrian streets, traffic calmed roads, low quality| 0: route in low quality urban-architectural environment). BE=ÓLiBi/ÓLi [Li: length of route section i (km) | Ni: quality level of the urban environment along route section i] (7) • Activities coverage: This criterion regards the number and type of urban activities that a route serves. The calculation was based on the urban activities identified in the first step. We examined how many of them fall into a buffer zone of 250m from the route. Then the various categories of urban activities were weighted in order to provide a bonus on those routes serving activities of high interest for the cyclists. The higher weight was given to universities and the lower to hospitals. • Centrality: This criterion regards the number of urban centers that a route serves. We first identified the location of all municipalities’ centers and then examined how many of them fall into a buffer zone of 500m from the route. We decided to extend the buffer zone from 250m to 500m as urban centers exert greater attraction to cyclists given the higher intensity of activities. • Accessibility to urban parks: The criterion concerns the number of urban parks (suitable for cycling) that a route serves. • Accessibility to public transport stations: The criterion describes the number and type of public transport stations that a route serves. We first examined how many of them fall into a buffer zone of 250m from the route and then the stations were weighted based on their hierarchical level (metropolitan, regional, and local stations). The values of all the above criteria were calculated for 46 cycle routes scenarios through GIS software (ArcGis 9.3). The final values were transformed into 1-100 scale in order to allow comparisons between them (see table 2). Table 2. The values of the 11 evaluation criteria (1-100 scale) for the 4 cycle routes scenarios of the Athens-Piraeus zone (see fig. 2). Cycle routes scenarios in Athens - Piraeus zone Length ( km) Energy Consumption Junctions Density Traffic intensity Traffic speed Legibility Natural environment Built environment Activities coverage Centrality Accessibility to urban parks Accessibility topublic transportstations AP1 10,3 70 66 83 68 55 42 36 66 50 0 81 AP2 7,89 88 65 0 0 100 0 100 58 57 0 15 AP3 9,01 81 53 67 59 48 23 61 65 57 0 58 AP4 9,19 79 62 83 56 61 27 35 55 64 0 81 Author name / Procedia -Social and Behavioral Sciences 00 (2011) 000–000 / Procedia -Social and Behavioral Sciences 00 (2011) 000–000 3.3.2. Criteria weighting and multicriteria analysis One of the most critical issues in multicriteria analysis is the weighting of the various criteria that form the evaluation basis. In our case we chose to attribute weights from the prospective users of the network, in the framework of a participatory planning approach. More specifically, we established a working group comprising 3 members of the research team and 10 experienced cyclists. We defined as experienced cyclists those who used the bicycle as the main transport mean for the last two years. Since there is no data available regarding the socio-economic profile of cyclists in Greece, we proceeded to a preliminary web-survey among regular cycle users (428 participants). The results showed that bicycle users in Athens are predominantly young (<44 years, 90%), highly educated (at least a bachelor’s degree, 76%) men (83.7%). So we chose a representative sample for our working group consisting of 8 men (80%) and 2 women (20%), most of whom were under 44 years old (4 members 25-34, 5 members 35-44, 90% in total) and highly educated (8 members had at least a bachelor’s degree, 80%). The weighting methodology was completed after two meetings and included the following steps. Firstly, the research team presented and explained the evaluation criteria one by one. Then a discussion session followed, between the members of the research team and the cyclists in order to clarify the logic of each criterion. In the next step, the research team distributed a printed form regarding the application of weights on a 1-10 scale, and a document summarizing each criterion, while detailed instructions were provided for the completion of the form. Based on the results of individual weighting, we applied multicriteria analysis, which revealed the optimum cycle route for each zone. In the next workshop (one day after) the optimum routes were presented to the experienced cyclists in order to make a direct connection between their weighting choices and the results on the map of the city. A discussion ensued on all the optimum routes and then cyclists were asked to decide together, based on the visual analysis of the routes, the final group weights (see table 3). A member of the research team was the coordinator of the debate. Under the new “group weights” multicriteria analysis was re-applied providing the final optimum route per zone and a first network scheme (see table 4 and figure 3). It is particularly interesting that the most important criteria for cyclists are energy consumption, junctions density, legibility and centrality, while the criteria relating to the traffic intensity and speed were assessed lower. The less important criteria are related to the natural environment, access to urban parks and public transport stations. A wide range of those criteria is consistent with other studies. Among these the most important are a preference for continuous provision of bicycle facilities and routes with moderate slope (Stinson & Bhat, 2003; Titze et al., 2008) as well as routes with fewer cross-streets, no parallel parking and fewer traffic controls (Stinson & Bhat, 2003). According to Sener (et al., 2009) traffic controls and cross-streets have a negative effect mainly on females and inexperienced cyclists, while traffic volume exert a similar impact on cyclists, especially when it is accompanied by noncontinuous bicycle facilities. In our study the character of the network seems to influence significantly the criteria preferences. Cyclists seem to prefer the metropolitan routes to offer fast, direct and safe movement to the detriment of qualitative characteristics of a cycle route like the attractiveness of the urban environment. On the contrary, according to Dill & Gliebe (2008) when the travel purpose of the cyclist is recreation, then the distance minimization and the avoidance of hilly terrain are the lowest priorities. As a cyclist (male, 33) stated in our workshops "For recreational purposes I would not choose the metropolitan network, but a city park" while a second one (female, 28) noted, "The best for me is to reach my destination with closed eyes, effortlessly and safely”. These options premium cycle route scenarios in central arteries and downgraded longer routes in secondary axes even if the urban environment there was better (see figure 3). Straight routes in relatively short distance from major urban centers were also selected by priority. Author name /Procedia -Social and Behavioral Sciences 00 (2011) 000–000 /Procedia -Social and Behavioral Sciences 00 (2011) 000–000 3.4. 4thstep: Spatial coverage assessment and final selection of the metropolitan cycle routes After the identification of the optimum cycle routes per zone we proceeded to an overall assessment of the metropolitan cycle network in terms of spatial coverage of the overall metropolitan area and the determination of the implementation priorities. Firstly, we identified the areas that the initial cycle network, as defined in step 3, did not cover. We applied buffer analysis in a 500m zone along the network, which means that a cyclist could approach a network branch in about 4 minutes with an average speed of 17km/h. We found 5 such areas (see figure 3) and examined whether any of the cycle routes scenarios that had initially been rejected would be appropriate to serve these areas. If such a scenario did not exist, then we designed a new cycle route. Eventually 22 complementary segments were added. The final network consists of 36 routes totaling 226km in length and allowing the majority of the Athenians to reach a network branch in less than 5 minutes. In the final step the cycle routes were categorized according to their implementation priority, based on: (a) the aim to obtain, right from the first phase, an integrated and functional network, fulfilling the criterion of continuity (or arteriality according to Marshall, 2005), (b) the geographical position-significance of a route in relation to the whole network (e.g. central or regional), and (c) the geographical position-significance of a route in relation to the neighboring routes of the network. Finally, 24 cycle routes of 180km length included in phase A and 12 routes of 46km length in phase B (see figure 3). Table 3. The final weights of the evaluation criteria, given by the group of the experienced cyclists. Group weights 10106 5 9 3 6 5 7 13 Table 4. The results of the multicriteria analysis for the Athens-Piraeus zone, based on the cyclists’ personal weights (left) and group weights (right). The fourth scenario was finally selected in this case, even if the first one was considered as the optimum in the first day of the workshop (according to the personal weights). AP1 57 59 58 53 58 59 58 55 60 53 61 AP2 40 49 37 37 47 57 49 52 41 47 58 AP3 53 55 53 48 54 57 54 52 55 49 59 AP4 55 58 55 52 58 60 57 55 58 53 62 4. Conclusions The promotion of cycling in Greece is not just a transportation challenge. The conditions in Greek roads are so hostile to human presence that cycling policies should take the form of a broader strategy for culture, security, health, quality of life, and economic development. However political decisions and planning practice remains car oriented. The decision of the Greek Ministry of Environment to include a metropolitan cycle network in the new Master Plan of Athens is therefore very important. The planning of Cycle route scenarios in the Athens-Piraeus zone Cyclist 1 Energy Consumption Cyclist 2 Junctions Density Cyclist 3 Trafficintensity Cyclist 4 Traffic speed Cyclist 5 Cyclist 6 Legibility Naturalenvironment Cyclist 7 Built Cyclist 8 environmentActivities Cyclist 9 coverage Cyclist 10 Centrality Accessibilityto urban parks Group Accessibility topublic transportstations Author name / Procedia -Social and Behavioral Sciences 00 (2011) 000–000 / Procedia -Social and Behavioral Sciences 00 (2011) 000–000 the network was a challenge for our research team as cyclists are still relatively few and Athens is characterized by a very complex land use and transport system (Milakis et al, 2008). We developed a four-step methodology seeking to combine the technical prerequisites of planning with the preferences of future network users. Thus, we identified specific zones where cycle routes shouldbe developed by analysing the urban structure of Athens. Then we turned to experienced cyclists to evaluate the planning criteria and their weights for the cycle routes scenarios. Through this balanced approach of public participation and multicriteria analysis we succeeded to design a functional network (covering most of the cyclists’ activities) that corresponds to the preferences of the end-users regarding the cycle routes characteristics. This methodology is useful for cities trying to introduce and prioritize cycling infrastructure because it focuses on where cyclists would prefer to cycle, making such investments more succesful in attracting users. It is also very important that cyclists felt that they contributed decisively to the design of the network and therefore they promoted the research results to the public through their websites, web forums, newspapers and public discussions. What is particularly noteworthy is the fact that cyclists prefer main arteries with straight, legible tracings and gentle slopes. They do so, despite the high traffic volumes and speeds, as they think that these axes would facilitate their movement and should be redesigned and become more human-friendly. These attitudes and preferences affected the shape of our cycle network, which is radial and oriented towards the center of Athens. It is particularly interesting that this is the shape and characteristics that predominate in most European cities’ networks (e.g. London, Copenhagen, Berlin), highlighting a common design approach, and possible common preferences of cyclists. Athens Piraeus Fig. 3. The initial network scheme after the application of the multicriteria analysis in 46 cycle routes scenarios (left) and the final map of the metropolitan cycle network after the spatial coverage assessment (Phase A -continuous line, Phase B -dashed line). Acknowledgements The Organization of Planning and Environmental Protection of Athens funded this research in the framework of the new Master Plan for Athens. We would like to thank George Nathenas, Transport Planner for his contribution in shaping some of the cycle routes scenarios and the 10 cyclists who participated in the workshops. Author name /Procedia -Social and Behavioral Sciences 00 (2011) 000–000 /Procedia -Social and Behavioral Sciences 00 (2011) 000–000 References Athens Urban Transport Organization (AUTO) (2007). O-D Household Travel Survey. Athens: AUTO. Attiko Metro S.A. (AM-DPGS) (1997). Metro Development Study. Athens: AM-DPGS. (in Greek) Dill, J., & Gliebe, J. (2008). Understanding and measuring bicycle behavior: A focus on travel time and route choice. Portland, OR: Oregon Transportation Research and Education Consortium. EEA (European Environmental Agency) 2009. Transport at a crossroads. TERM 2008: indicators tracking transport and environment in the European Union. Luxembourg: OOPEC. Eurostat 2010. EU energy and transport in figures. Luxembourg: OOPEC. Greek Cyclists (Non Governmental Organization) 2008. Cycle routes in Athens. Marshall, S. (2005). Streets and Patterns. London: Spon press. Official Journal of the Hellenic Republic No 18A-N.1515 (1985). Master Plan and Environmental Protection Programme for Greater Athens. Athens: OJHR. (in Greek) Papavasileiou, D., Milakis, D., & Vlastos, Th. (2010). Car dependence or appetence? Examination of attitudes towards sustainable mobility in the Greek case. Proceedings of the 12th WCTR, Lisbon. Sener, I. N., Eluru, N., & Bhat, C. R. (2009). An analysis of bicycle route choice preferences in Texas, US. Transportation, 36, 511 – 369. Stinson, M. A., & Bhat, C. R. (2003). An analysis of commuter bicyclist choice using a stated preference survey. Transportation Research Record, 1828, 107 – 115. Titze, S., Stronegger, W. J., Janschitz, S., & Oja, P. (2008). Association of built-environment, social- environment and personal factors with bicycling as a mode of transportation among Austrian city dwellers. Preventive Medicine, 47, 252 – 259. MoE (Ministry of Environment, Physical Planning and Public Works) 2008. National Strategic Reference Framework, Accessibility Improvement Programme 2007-2013. Athens: MoE. (in Greek) Davies, A., Burgess, J., Eames, M., Mayer, S., Staley, K., Stirling, A. & Williamson, S. (2003). Deliberative Mapping: Appraising Options for Addressing ‘the Kidney Gap’, Wellcome Trust’s programme for novel methods in public consultation (2000 – 2003), Final Report. Hopkinson, P. & Wardman, M. (1996). Evaluating the demand for new cycle facilities. Transport Policy, 3(4), 241 -249. Larsen, J. & El-Geneidy A. (2009). Build it, but where? The use of geographic information systems in identifying optimal location for new cycling infrastructure. Paper presented at the 89th TRB Annual Meeting, Washington D.C., USA. Menghini, G., Carrasco, N., Schuessler, N. & Axhausen, K.W. (2010). Route choice of cyclists in Zurich. Transportation Research A, 44(9), 754–765. Merkhofer, M., Conway, R. & Anderson, R. (1997). Multiattribute Utility Analysis as a Framework for Public Participation in Siting a Hazardous Waste Management Facility. Environmental Management, 21(6), 831-839. Milakis, D., Vlastos, Th., & Barbopoulos, N. (2008). Relationships Between Urban Form and Travel Behaviour in Athens, Greece. A Comparison with Western European and North American Results. European Journal of Transport and Infrastructure Research, 8(3), 201-215. Rybarczyk, G. & Changshan, W. (2009). Bicycle facility planning using GIS and multi-criteria decision analysis. Applied Geography, 30, 282-293. Schwartz, M. & Eichhorn, C. (1997). Collaborative Decision Making: Use of Multiattribute Utility Analysis to Involve Stakeholders in Resolving Controversial Transportation Issues. Journal of Advanced Transportation, 31(2), 171-183. Stirling, A. (2006). Analysis, participation and power: justification and closure in participatory multi- criteria analysis. Land Use Policy, 23, 95-107.