Mobility hubs in Salzburg and it’s neighbouring municipalities

Inhaltsverzeichnis

1 Introduction

2 Motivation and State-of-the-Art
2.1 Definition and characteristics of a “Mobility Hub”
2.2 “Mobility Hubs” in the present literature
2.3 Methodological approaches in the literature
2.4 Selection of mobility hubs in the literature
3 Methodological Approach
3.1 Methodology to identify Mobility Hubs
3.1.1 Data Acquisition
3.2 Data preparation
3.3 Potential Analysis for Mobility Hubs
3.3.1 Creation of Service Areas around service stop locations
3.3.2 Integration of data into service areas
3.3.3 Integration of frequency data, service lines and intervals
3.3.4 Calculation of suitability values
3.3.5 Reintegration into ArcGIS Pro and identification of mobility hub locations

4 Detailed Planning and Design for a selected Mobility Hub
4.1 Constructional design and bike parking facilities
4.2 2-D drawing of the area based on traffic planning principles and regulations
4.3 On-site-inspection and survey of existing infrastructural equipment and furniture at the existing hub Europark
4.4 3D-Visualization of the Mobility Hub Europark

5 Conclusion and Outlook

References

Table of Figures

Tables

Story Map

Abstract

Originating from present service stops, combineOBUS analyses potential locations for innovative mobility hubs as places of connectivity. Mobility hubs do not simply serve as waiting points but offer different services: depended on the surrounding conditions that deal as parameters of suitability, “upgraded” stops for public transit also offer possibilities and amenities to obtain realtime-information about waiting times, WIFI-Access, store or rent bikes or store goods in smart lockers. In this way, the motivation to multimodal forms of mobility shall be fostered within the city of Salzburg and its surrounding municipalities.

A multicriteria approach was applied for the analysis, being based on different aspects that are important basis for the capability of installing a mobility hub. The analysis is based on service areas of 300 meters from each stop. For these service areas, inhabitants, workplaces, stores and shops, the cycling network, bus intervals, passenger frequency, number of interchangements as well as further points of interest (educational institutions, sights, hotels,…) were set as parameters for the suitability of stops as mobility hubs. The overall suitability for setting up a mobile hub for each service stop was calculated and displayed on a map. In addition, the 15 service stops were chosen as suggestion for establishment, taking also spatial distribution into consideration. The identified mobility hubs show good ratings in all applied parameters, so the intended differentiation of equipment for each stop was evitable.

Keywords: mobility hub, potential analysis, multicriteria approach

1 Introduction

Present service stops are analysed towards their potential as dealing or being upgraded to “mobility hubs” – public transit stops with additional functional and service facilities - keeping in mind that service stops of the future should not only serve as waiting points, but offer additional functionalities: To keep the bike safely locked when using combined transportation, to receive updated passenger information and the potential for having smart lockers for storing or receiving goods nearby.

A high share of motorized private transportation (45% modal split) lead to congestion, noise and air pollution while on the other hand share of public transport is moderate at about 15% (Stadt Salzburg, 2017). Negative impacts of private car use can be met by promoting the eco-friendly and multimodal mobility as competitive alternative.

The aim of promoting multimodal mobility, especially the combination of walking or cycling and taking means of public transportation derives from the concept of compact cities, also referred to as cities of short distances. The implementation of mobility hubs follows the idea of this concept that “urban activities should be located closer together to ensure better access to services and facilities via public transport, walking and cycling, and more efficient utility and infrastructure provision”(Kii and Doi 2005). The compact city is characterized by a relatively high density and mixed land-uses and is related to a range of concepts about urban form, traffic and liveability (Haase, Bauer et al. 2013). Thus, it relies on a well-functioning and efficient public transit system and an urban layout encouraging walking and cycling. Its design therefore supports low energy consumption and reduction of pollution (Dempsey 2010). These characteristics of compact cities are relevant and crucial when it comes to dealing with the issue of strong congestion in and around the city of Salzburg. Therefore, the conditions of the local environment as well as human scale factors are emphasized – and deal as parameters for our analysis, such as density of population, stores, points of interest or working opportunities. In its original form, the concept of compact cities is based on the idea of proximity, meaning that the daily needs of residents and workers are available within walking or short cycling distance (Haase, Bauer et al. 2013).

Some years ago, the focus on proximity has been gradually replaced by emphasizing accessibility. This leaves room for a new interpretation of the compact cities concept where travel time has become more important over distance (Zonneveld 2005). Therefore, the revised concept promotes the integration of transport. Thus, the concept is closely related to mobility – or as HAASE expresses it: “There is no compact city concept without emphasis on public transport or, more specifically, a reduced need for transportation by car” (Haase, Bauer et al. 2013). This paradigmatic shift is reflected in our analysis: the idea of proximity is taken up in the investigation of densities of the selected parameters (population, POI, etc.) within service areas of 300 meters around each stop location; whereas the idea of accessibility is reflected in our data about the service stop locations themselves with regard to frequency of service and number of lines operating at each stop respectively.

However, AGUILERA and MIGNOT (MIGNOT AND AGUILÉRA 2004) state that the commuting behavior and patterns are changing, since subcenters for employment are growing in addition to the city center. Identifying the stop locations most suitable for the implementation of mobility hubs within our analysis, the station Europark matches this observation: The site, being located at the community boarder to the neighboring community of Wals-Siezenheim, combines a shopping mall, a furnishing house and the national headquarter of a commercial chain. There, one finds an agglomeration of workplaces, summing up to about 1600 employees. It can thus be seen as a subcenter for working and apart from traffic volume caused by commuters going to work, this area is also frequently affected by congestion due to strong motorized customers traffic. This makes the location an interesting spot for closer investigation and thinking, how the place can be adapted as a mobility hub to promote sustainable means of transportation.

The accessibility of present as well as potential service stops for public transportation are identified. The analysis is done with a certain focus on the potential of multifunctional, "smart" service stops. In the prospective future, service stops do not only serve as waiting devices for passengers but provide additional functionalities: for strong bikes on safe bikestands, for receiving real-time information about waiting times, availability of bike/carsharing vehicles or for storing/receiving goods thanks to a smart locker located nearby.

Therefore, the identification of potentials (optimization/extension) for public transport in combination with cycling infrastructure and smart lockers to rise attractiveness of public transport can play a significant role to meet these challenges that affect everyday life in the central region. The temporal accessibility of all-day (0-24h) available smart lockers for storing/exchanging goods not only complement service offers of the Salzburg AG and local businesses but may reduce time and travel expenses for designated users. Accordingly, the inhabitants as (potential) passengers and users of the smart lockers could be seen as final beneficiaries of this project.

The analysis involves the definition of relevant parameters regarding accessibility, data acquisition, multi-criteria analysis in GIS and communicate the achieved results. The findings primarily offer a basis for planning and argumentation for the Salzburg AG and the city administration. Inhabitants as (potential) passengers serve as final beneficiaries. The dissemination of the results is done with a map/poster as well as a story map transparently stating the workflow. The resulting proposals regarding accessibility with public transport and on the other hand improve the handling of transport service for the Salzburg AG. Regarding data protection, the used datasets of population and passenger numbers for the respective lines are aggregated to a degree that no conclusions on single persons can be made.

2 Motivation and State-of-the-Art

A review on existing literature about mobility hubs was conducted to inform about: characteristics of mobility hubs; approaches and methods that were used in other sources about potential analysis; and for getting references which parameters and indices to use related to the analysis.

The usage of the term “mobility hub” still is rather new and has geographical focal areas where it occurs: Thus, it is already used in documents provided by planning authorities and planning departments of universities in the US (especially west coast) and Canada, as well as India, while the expression is almost absent in the scientific literature such as papers or articles. Therefore, the present article

2.1 Definition and characteristics of a “Mobility Hub”

There is no consistent definition about what a mobility hub is and what it is characterized by. For our project, we stick with the definition provided by the City of Toronto, which sees mobility hubs as “places of connectivity where different modes of transportation – from walking to riding transit – come together seamlessly and where there is intensive concentration of working, living, shopping and/or playing” (Engel-Yan and Camino 2011).

The following graphic shows criteria for mobility hubs: The graphic is an interface between a) giving the observer an idea, what characteristics of a mobility hub are and what it stands for and b) it leads towards the parameter to be used in the analysis.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1: Characteristics of a Mobility Hub (source: own figure)

The different parameters mentioned in the circle include the following, which are adapted from ENGEL-YAN (Engel-Yan and Camino 2011):

- Multimodal transportation: fostering seamless transfer and integration of different travel modes at one location.
- Residential and employment density: in order to be suitable as a mobility hub, a certain amount of residents as well as workers shall be found around the respective stop.
- Pedestrian priority: for pedestrians, a secure and attractive environment is promoting the actual willingness to walk to destinations.
- Embedded technology at the hubs: this refers to real-time travel information as well as eg. to the provision of WIFI-access for customers at the stops.
- Economic vitality and competitiveness: another factor for a successful mobility hub is that there are different stores, services, … available close to the location
- Further points of interest: which for example include educational institutions, touristic attractions, hotels, museums or galleries …

2.2 “Mobility Hubs” in the present literature

Mobility hubs propose to bring several modes to travel together in the same place and are seen as having potential to address notable deficiencies in the current transportation system (Anderson, Blanchard et al. 2015).

The city of Greater Toronto defines mobility hubs as “key transit stations [that] become mobility hubs, where transportation modes, including rapid transit, local transit, specialized transit, cycling and accessible pedestrian networks come together seamlessly. […] They offer amenities to travelers such as traveler information […] and services” (Engel-Yan and Camino 2011). Within the transportation planning literature, HENRY and MARSH define mobility hubs as “location where multiple transport modes converge to enhance connectivity to and from each other and provide first- and last-mile accessibility to destinations from transit facilities”(Henry and Marsh 2008). So they are guaranteeing a seamless confluence of several transportation modes at one location; furthermore, it provides additional value to passengers who benefit from the possibilities of this seamlessness as well as reduced trip costs, as people are provided with a choice of travel mode (Henry and Marsh 2008).

Generally, mobility hubs tend to be located close to key transportation infrastructure, which possess high service frequency and several lines operating, as well as large employment centers and diversity in land-use, eg. through stores, educational, artistic or leisure institutions (Anderson, Blanchard et al. 2015).

2.3 Methodological approaches in the literature

ANDERSON [et.al.] from the University of California conducted a mobility hub suitability analysis for the city of Oakland in 2015. Their approach also included a multi-criteria framework which used data from various sources to for indices-construction. In addition to the quantitative analysis, a qualitative evaluation supplemented the approach to generate a set of proposed mobility hub locations. The qualitative evaluation also included an assessment, which specific modes would be fostered in each mobility hub. The authors also admit that their mixed-methods approach employs a certain degree of subjectivity when it gets to the selection of indices as well as the selection of final locations (Anderson, Blanchard et al. 2015). Also our potential analysis for mobility hubs within the city of Salzburg and its surrounding municipalities follows this approach to combine quantitative analysis (several data sources were used for finding indices and creating a suitability index) with qualitative sources considerations (on-site-inspection and survey, final selection of potential mobility hubs).

Apart from population data, employment centers and transportation infrastructure, land use diversity is a frequently mentioned parameter in the existing literature. RATTI for example included the following variables into her analysis: income, immigrants, unemployment rate, household size, population density, job accessibility and density, service frequency, walk score as well as transit mode shares as dependent variable (Ratti 2017). Here it has to be stated that some of these data are not or not inexpensively obtainable on a small-scale (block or household-level) in Austria. So, some of these variables such as income (not available), immigration (not considered as reasonable and not available on the small-scale) were not used, while others as job or population density or service frequency were considered. Within the “Mobility Hub Profiles Methodology” guidelines, METROLINX explains the inclusion of points of interest in mobility hub potential analysis: It aims at differentiating according to different to their importance into “more regional destinations that deserve highlighting, such as stadiums, museums, regional malls” (Metrolinx 2015) and secondary localized destinations, such as libraries, community centers or high schools. Also in our analysis, a differentiation according to importance and range of attraction was conducted. The methodology that was suggested in these guidelines includes to take the following variables for the analysis: destinations (corresponding our points of interest), population density/growth and age, income, job density, household composition, vehicle ownership as well as kilometers of bikeways and accessibility to stations.

A principle that is obvious and well known to geographers through TOBLERS first law of geography is that the distance to as well as the density of destinations are appropriate indicators for the trip generation potential. In our case for mobility hubs, this potential also applies to further services that are available at mobility hubs, eg. bike sharing or deposition of goods in smart-lockers (Ewing and Cervero 2010). Within our analysis, this is done via the creation of Service area and the inspection, which population density, density of work places, the density of the cycling network or points of interest is found within each service area respectively. The service areas are designed for a 300 meters buffer, which is considered as being accepted by the majority of the population as walking distance to a bus service stop (Frey 2015). According to CHATMAN [et.al.], the construction of service areas for walking distances as transit catchment areas is a widely used concept in transit planning literature (Chatman, Cervero et al. 2014). Also ENGEL-YANS mobility hub guidelines for Toronto define catchment zones for mobility hubs, with 250 meters being the “primary zone”, corresponding 1.5 minutes walking time.

Generally spoken, the methodological red thread for our analysis was inspired by ANDERSONS “Mobility Hub Suitability Analysis” as well as METROLINX “Mobility Hub Profiles Methodology”.

2.4 Selection of mobility hubs in the literature

About the index construction, ANDERSON states that they represent an aggregation of multiple individual variables (Anderson, Blanchard et al. 2015). This approach is reflected in our creation of a final “suitability index” which returns the appropriateness of each service stop to deal as mobility hub. Furthermore, Anderson admits that for the final selection of mobility hubs, quantitative approaches might not be enough and therefore favor a qualitative evaluation of the outcomes. We follow this idea in our approach, leaving the final suggestions for mobility hub locations to qualitative assessment.

With regard to this selection, transportation connectivity of the respective stops and the parameter densities that can be found in the surrounding service areas can be thought of in two opposite ways: 1) that areas which already possess high transportation connectivity and characteristics of mobility hubs may be good candidates for an update to a mobility hub; 2) areas with low characteristics might benefit more from an increased investment.

In our case, we decide for the first way of thinking, as it is more realistic the responsible authorities decide to update in already existing good prerequisites than taking the risk of investing into a random stop to highlight it.

3 Methodological Approach

3.1 Methodology to identify Mobility Hubs

The analysis involves four steps: the definition of relevant parameters regarding accessibility, data acquisition, multi-criteria analysis in GIS and communication of the achieved results.

The different parameters that were fed into the analysis are based on the abovementioned literature and were adapted to the availability of data sources for Salzburg. According to literature and Philipp Blüthl, who works as assistant of the traffic department at Salzburg AG, population is one of the most important factors for setting up new service stops no matter if it will be considered as a mobility hub. Job locations are also a factor that increases passenger frequency on a certain bus stop. To cover the field of accessibility and attractiveness of existing bus stops various factors were chosen as relevant: the number of serving lines which also gives information on possibilities of interchangements and therefore connectivity, the time interval between busses and the current passenger volume if available. In addition to that also important Points of interests nearby can have an influence on the establishment of a service stop and could increase the attractiveness.

Apart from the classical factors that are considered for a service stop also issues that emphasize a mobility hub were integrated: The shops were treated separately from the POIs as they are interesting for deploying goods after opening hours in a smart locker. This analysis takes also account of biking lanes as they serve as connection point to public transport for the last meters to a destination. The core of this analysis is to find out how good each service stop performs regarding these parameters. The workflow (Figure 2) involves the establishment of service areas to find out what is in a reachable walking distance of each service stop. The service stops should further be compared concerning their suitability and the best ones should be highlighted. The final step will be to pick one of the service stops out and deal detailed with this identified mobility hub regarding planning.

Abbildung in dieser Leseprobe nicht enthalten

Figure 2: Workflow (source: own figure)

3.1.1 Data Acquisition

The following table should provide an overview of all the datasets that were used for the suitability analysis. Each of them is described in its initial stage before preparation and is used to fulfil the previously stated parameters.

Table 1 Data and Sources

Abbildung in dieser Leseprobe nicht enthalten

3.2 Data preparation

Subsequently, the data obtained from Austrian sources had to be integrated with the acquired German data. This step of data preparation required certain attention, as the data’s reference frame was of different type and needed some transformational steps. Also, the scope of the different datasets is different and had to be harmonised.

3.2.1 Area of interest

In order to generate the area of interest, which contains the city of Salzburg and its neighbouring municipalities, two datasets had to be joined. The City of Freilassing was extracted from the administrative boundary’s dataset of Bavaria. Further it had to be projected to the coordinate system of the 8 selected municipalities in Austria. After merging them, the commuting data of stadtregionen.at was assigned to the matching municipalities. They already give an indication about the traffic links between Salzburg and its neighbouring municipalities on a typical weekday. The next step involves clipping of all spatial datasets to the focus area. Especially for big datasets like the Points of interest which cover the whole country of Austria this is necessary to save time and computing power.

3.2.2 Population and employee data

To overcome the lack of population data for Freilassing, the building dataset was used to approximate the population. The OSM buildings were categorised according to their purpose. Industrial, public and commercial buildings were sorted out as they are supposed to not contain any residents. The remaining buildings which are represented as points were assigned with the mean household size from 2015: 2,03 persons per household (Bayerisches Landesamt für Statistik 2016). This may be quite accurate for one-family houses but when it comes to buildings with flats the actual population count may be way higher. For the purpose of finding suitable mobility hubs, however, this is a sufficient level of detail as it already gives a good overview on the distribution of the population.

The population for the Austrian municipalities is available as point feature. Since population is distributed over an area and not concentrated on a certain point the population data had to be converted into areal units. This should avoid the effect that service areas are dependent on reaching exactly that point to get a high population value. For the decision of the polygon type there are only three shapes that can repeat the same form repeatedly as a grid: squares, hexagons and triangles. (Figure 3) In this case hexagons were used because they match the curves of the service areas best. Furthermore, points within a hexagon are more similar to their centroid compared to triangles and squares since they better resemble a circle. (ESRI 2018)

Abbildung in dieser Leseprobe nicht enthalten

Figure 3: Square grid, hexagon grid and triangle grid (source: ESRI 2018)

The population data of Salzburg was integrated into the established hexagons with regards to only match one point with one hexagon. Only in the border areas where the points are not regularly distributed, more points serve as input for one hexagon. The ideal size for the hexagons to reach this requirement was 10.000 square meters in this case. In addition to that also the buildings of Freilassing were integrated and multiplied with the average household size. The hexagon grid does not only contain the population data for all municipalities but also the employee number except for Freilassing. Finally, only the hexagons with population values or employee values over zero were kept.

3.2.3 Points of interest

The Points of interest dataset of OpenStreetMap offers over 100 different categories reaching from a like alpine huts to w like wayside shrines. For this analysis only important travel destinations were taken into consideration and exported from the initial dataset. The following list of categories were identified as relevant:

- Shopping Centres
- Airport and Railway stations
- Tourist Hotspots
- Shops
- Leisure Time Facilities
- Public Amenities
- Doctors
- Public Pools
- Education
- Hotels and Student Dorms
- Gastronomy

Some of them are composed of several OpenStreetMap categories, for instance doctors, veterinarians and dentists were simplified into the category doctors. Moreover, the categories were enriched by the Points of Interest dataset for possible smart locker locations in Salzburg to cover this aspect of a mobility hub. Also, the railway stations and the airport were added from the transportation dataset were integrated because they serve as important connection point to supra-regional and international transportation.

To model the difference in importance of the points of interests they were all assigned with a weighting from 0,05 to 1,0 according to their relevance. For example, a shopping centre (value: 1) attracts much more people than a snack bar (value: 0,1). Some categories even have differences in weighting within their categories: The faculty of cultural and social sciences have a higher weighting as the faculty of catholic theology because of a higher number of students and therefore more possible passengers for public transport.

3.3 Potential Analysis for Mobility Hubs

3.3.1 Creation of Service Areas around service stop locations

As aforementioned, the creation of service areas, serving the theoretical background of proximity relating to the compact cities (Haase, 2013) is the ground for further analysis steps. Kilchenmann, Schwarz-von Raumer (2013) state a distance between 300 and 500 metres to be acceptable for potential users of bus stops. Figure 4 (Räppel 1984) shows the probability of the usage of a service stop on the y-axis and the according distance on the xaxis. The acceptance which is represented by a gaussian distribution decreases rather fast, after 300 metres it is below 60 percent and after one kilometre it already goes down to zero.

One of the main reasons for this phenomenon is human perception. Sensed walking time does not correspond to the actual measured walking time. A study has shown that 300 metres distance are perceived as 2 minutes walking time whereas the double distance is already sensed as 10 minutes. (Kilchenmann, Schwarz-von Raumer, 2013: 113) To sum up, perceived walking time increases exponential and not linear in comparison to walking distance. For this analysis a distance of 300 meters was chosen as this has still a high acceptance over 50 percent.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4: Acceptance of distance to a service stop as gaussian distribution (source: Räppel 1984)

In comparison to a buffer with an Euclidian distance of 300 metres a service area takes the street network into consideration and is also able to deal with walking time. Pedestrians are only able to move along the given street network and so they are faster in directions with a dense street network. So even if a service stop is in only 200 metres air-line distance that does not mean that it is reachable for people by walking within 200 metres.

As a basis for the underlying network the GIP Street network and walking ways dataset was used. Before converting this into a network dataset with nodes and edges the motorways were sorted out because people are not allowed to walk on them. ArcGIS Pro only offers the possibility to establish service areas that are based on their own previously defined network dataset, so in this case ArcMap and the Network Analyst extension were used instead. As a next step the bus stops were set as input for the facilities to calculate 300 metre distances originating from them. As there are always two bus stops, one for each street side, the service areas were merged according to their names. In some cases, like Justizgebäude or Gaswerkgasse there are more than two bust stops that sum up to one service area.

3.3.2 Integration of data into service areas

To get information about the suitability for each factor, the parameters that where spatially available were integrated into the generated service areas. (figure 5) This involves population, employees, biking lanes and points of interest. The integration was done by the summarize within tool in ArcGIS Pro which combines spatial join and summary statistics. All the features that are within a service area are identified and the amount is calculated: For instance, biking lanes in metres. The points of interests were then further assigned with their weighting value to get an overall point of interest’s value:

Abbildung in dieser Leseprobe nicht enthalten

The population and employee data were given as polygon feature. To keep it simple it is assumed that population is evenly distributed in a hexagon. The tool calculates the share of the hexagon that is covered by the service area and applies this proportion to the population value. For example, if a service area overlays a hexagon with a population of 100 to 50 percent and a second hexagon with 300 inhabitants to 20 percent the population in this service area would be 110 people. The same applies to the employees which were calculated equally.

[...]