- Email: [email protected]
Evaluation of Munich's Cycle Route Planner Data Analysis and Customer Survey
Available online at www.sciencedirect.com
ScienceDirect Transportation Research Procedia 19 (2016) 225 – 240
International Scientific Conference on Mobility and Transport Transforming Urban Mobility, mobil.TUM 2016, 6-7 June 2016, Munich, Germany
Evaluation of Munich’s Cycle Route Planner Data Analysis and Customer Survey Florian Paul a *, Klaus Bogenberger a, Bernhard Fink b a
University of the Federal Armed Forces, Werner-Heisenberg-Weg 39, 85579 Munich, Germany b Munich Transport and Tariff Association (MVV), Thierschstr. 2, 80538 Munich, Germany
Abstract A convenient combination of pre- and on trip route information for cyclists are online routing tools. The Munich Transport and Tariff Association in cooperation with the Department of Environment and Health recently developed the MVV Cycle Route Planner. This online route-planning tool can be used either on desktop computers or on smartphones and covers the greater Munich region based on Open Street Map Data. It enables both users and municipalities to add information of cycle routes and other new bicycle infrastructure to the route planner and improve the system continuously. It provides a navigation and map service especially for the requirements of cyclists and can be combined with the use of public transport. This paper concentrates on three different research approaches. In a data analysis more than 130.000 single requests from April until August 2015 were examined with focus on the spatial distribution of origins and destinations in Munich and the suburban region. As a result, the demand and frequency of the user requests give a strong lead where cycle flows are to be expected and further infrastructure, like cycle super highways or improved cycle routes, needs to be considered. A customer survey of users and non-users revealed their mobility patterns and the actual use of the cycle route planner for trips to work or for leisure. The survey focused on usage and assessment of the cycle route planner itself. It further analyzed which means of transport are used for frequent trips and what the restrictions are so far, not to use the bicycle. © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license © 2016 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of mobil.TUM 2016. Peer-review under responsibility of the organizing committee of mobil.TUM 2016. Keywords: Bicycle; E-Bikes; Route Planning; Cycle Highways; Customer Survey
* Corresponding author. Tel.: +49 89 60042529; fax: +49 89 60042501. E-mail address: [email protected]
2352-1465 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of mobil.TUM 2016. doi:10.1016/j.trpro.2016.12.083
226
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
1. Introduction Bicycle Route Planners are a very popular and efficient tool to promote cycling, not only for tourist destinations in rural areas, but also for everyday cycling in urban agglomerations. In combination with the increase of online mapping tools and geo-information systems based on crowdsourcing, everyone on personal computers and smartphones can use these route planners. Although the development of reliable and coherent cycling route networks is a big challenge and also depending on the special needs and demands of cyclists, there is a growing number of cycle route planners for many cities and regions. This paper is focusing on the MVV-Cycle Route Planner for the City of Munich, which has been developed by the Munich Transport and Tariff Association together with the Department of Environment and Health. It will give a brief overview about existing online-route planners and navigation tools for cyclists and then focus on the evaluation of Munich’s Cycle Route Planner. In Munich there is a current lack of cycle traffic data. Only seven automatic bicycle counting spots, spread over the whole city, are counting the number and direction of cyclists. There is no current data available about cycle traffic flows, the modal share or the satisfaction of cyclists with the conditions of cycle tracks, for example. The last large-scale study, called MIDMUC, was carried out in 2008 and showed a modal split of 13.6% for cycling in Munich (LHM 2010). There are several developments in the cycle network in the past, which indicate that this number did presumably grow in the last eight years. The evaluation of Munich’s Cycle Route Planner can help to deliver a current data input to quantify the bicycle traffic. It provides furthermore some evidence where the most important origins and destinations for cyclists in the city and the region are. Further research questions about recommendations to prioritize bicycle highways or enhance the existing bicycle infrastructure at certain locations will be discussed. 2. Literature Review So far, web-based cycle route planners are only subject of rare studies, due to the very recent implementation of these routing programs in the last years. Nevertheless there are several studies and research projects about bicycle route choice, using available datasets and surveys. HUNT and ABRAHAM (2006) for example presented results of a questionnaire on the influences on bicycle use. They figured out that cycling in mixed traffic on streets is more exhausting for cyclists than on separate bike lanes. Cycling becomes less exhausting in mixed traffic, as the level of experience increases. Concerning the cycling facilities at the destinations, they found out that secure bike parking is more important than showers. WILLIS, MANAUGH and EL-GNEIDY (2013) give a broad overview about travel behavior and mode choice concerning cycling. They summarize the literature on social and psychological factors, which have an influence on the choice for cycling and make suggestions to increase the bicycle mode share. A study from the United Kingdom is investigating a route choice analysis of urban cycling behaviors using OpenStreetMap (OSM). YEBOAH and ALVANIDES (2015) examined GPS track data of cyclists using OSM and show the importance of consistent cycling networks, in order to support higher destination accessibility and increase the route directness. Besides comfort and route directness the attribute safety plays a significant role for cyclists, considering different route choices. SINGLETON and LEWIS (2011) are combining bicycle accident information with bicycle route planning, using the example of London. They analyzed the spatial locations where bicycle accidents occurred and drew a comparison between the quickest route and an accident avoidance weighted route. For a certain number of evaluated trips, they show that routes avoiding areas of high accident volume did not increase the trip length significantly. Another study from London has investigated the decisions that cyclists make when deciding which route to take (GLEAVE 2012). The key considerations around route choice of the cyclists centered on choosing the safest routes, and avoiding traffic. Less experienced cyclists prefer routes with less traffic and a cycle lanes, avoiding also confusing junctions. Comparing distance-based routes and attribute-based routes BEHESHTITABAR ET AL. (2014) are presenting a bicycle route choice model that is based on a cost function. They tried to find out which parameters have higher influence on bicycle route choice, and how they were contributing. The resulting model is capable to predict the most probable path a regular commuter would take between two points located in a defined area. The model does not only find the shortest path for cyclists, but also prefers attributes for the most suitable route in terms of safety and comfort.
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
227
3. Characteristics of cycle route networks and requirements of different user groups Compared to classic car-orientated road networks, cycle route networks contain many different characteristics due to a various number of infrastructural differences in most European countries. Bicycle streets, one-way streets with contra-flow for cyclists, cycle paths, car-free greenways, traffic calmed streets and many more categories become often only visible to cyclists, if there are explicit and comprehensible sign posting, cycling maps or navigation systems. Depending on different cycling target groups, the requirements for route choice options vary largely. Families with children for example, look for routes that are free of car traffic, without heavy inclination, easy to follow and embedded in an attractive landscape. Commuters on the bicycle often prefer the most timesaving routes, which also are suitable for everyday use under different weather conditions. Sport cyclists, using a race bike, especially pay attention to a smooth and tarmacked road surface. Recreational cyclists are usually looking for attractive destinations with a convenient surrounding, like green spaces and attractive points of interests, like swimming lakes or beer gardens. Concerning the development of route planning for cyclists, signposting appeared at first to show important destinations and routes. Printed cycling maps and cycling travel guides also provided important information for cyclists over the last decades. With the appearance of online route planners and mobile applications in the last years, the requirements of the target groups could be satisfied even more specific (DIFU 2012). 4. Online Route Planners and Navigation Tools The emergence of digital geo-information tools on the internet had a strong impact on cycle navigation. Navigation software combined with location-based information platforms for special points of interest can help to promote cycling and encourage cyclists to coordinate their interactions and decisions. Online route planners are a useful example how cycling-related contents can be merged with location-based information (DIFU 2012). So far, navigation software and route planning programs were mainly developed for motorized traffic and public transport. The increasing number of cyclists in many European cities also raised the demand for high-quality cycle route planners, not just for tourism purposes, but also for everyday cycling. The route planners themselves differ depending on the source of the locations-based data associated with the map and different providers, like State Transport Departments, private companies or research projects. Depending on the main purpose for the trips, the route planners concentrate more or less on touristic cycling or non-leisure, everyday cycling. The level of interactivity can also play a significant role. Route selection modes, start and destination points, personal adjustments like speed level or type of cyclist, as well as data collection in form of GPS tracks, are crucial for the development of a cycle route planner. The integration and connection of cycling route planners to public transport networks and journey planner are a crucial improvement and can help to support an intermodal way of transport. According to the particular provider of the route planner, there are different priorities of mapping information. Whereas regions and federal countries in Germany focus more on a touristic use, municipalities and communities orientate more on mapping networks for every-day use for cyclists (DÖLGER 2013). One of the first bike route planners in Germany was created in 2011 by the public transport association of BadenWürttemberg (NVBW). The NVBW is in charge of the bicycle route planner for the Federal State of BadenWürttemberg, by order of the Ministry for Transport and Infrastructure. The route planner is focusing especially on intermodal route choice options in connection with public transport. The goal was to strengthen a sustainable and environment-friendly mobility and also to develop a new bicycle culture. The route planner is available also for smartphones and enables the user to plan their cycle routes in combination with all modes of public transport. The use of environmental-friendly means of transportation for intermodal trips should be as convenient as possible. The Regional Office for Environment, Measurement and Conservation is in charge of data administration and management. The local authorities are enabled to gather their information about improvements of bicycle infrastructure by using a web-based Geoinformation System (GIS). There are more than 40,000 kilometers of bicycle network currently included. The cooperation between cities and rural districts concerning the data administration is important for offering high quality and effective data of the bicycle network. Because of the close cooperation for data management between the different municipalities the bicycle route planner has a positive effect
228
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
on transport policy and touristic issues. The ongoing update of bicycle network data is the fundament for a successful performance and popularity among many cyclists in Baden-Württemberg (LENZ 2015). 5. Evaluation of Munich’s Cycle Route Planer The MVV-Bicycle Route Planner was introduced in April 2015 by the Munich Transport and Tariff Association together with the Department of Environment and Health. The route planner is integrated into the public transport journey planner and offers an individual bicycle navigation within the greater Munich region, far from the city boundaries. The bicycle network information and data set is based on Open Street Map (OSM). The mapping device is specialized on the needs of cyclists and crowdsourced, meaning that users are able to make updates which improve the system continuously. 5.1 Specifications of the MVV Cycle Route Planner There are many options for the users of the cycle route planner to get detailed information about the recommended routes. The following information is available: x x x x x x
Journey time and length of trip Roadbook (directions) Surface condition Elevation profile GPS-Tracks (download on mobile devices) Public Transport Journey planner
After entering an address or location for origin and destination the route planner always gives you three different proposals for route choice. The “Fastest Route” represents the quickest and not necessarily the shortest route. The “Green Route” leads the cyclist as much as possible along green spaces and park areas. The “Family Route” prefers explicitly cycle ways, traffic calmed streets and avoids as much as possible ascending slopes. Based on individual choices the user can also make different adjustments for each route. Besides the choice of a personalized average speed level, on which the trip duration is dependent, the user is able to decide whether he wants to avoid ascending slopes. He can further define the surface conditions by choosing the type of bicycle (race bike or mountain bike) and can specify if he is en route with a bicycle trailer for children. After having made those decisions, the user is able to decide between three different trip adjustments: x x
x
Bike: exclusive use of the bicycle Bike and Ride: by bike to a station and then using public transport Bike entrainment: using the bike to a station, bike carried with public transport and cycling to the destination
A route example with three different proposals for route choice is shown in Figure 1.
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
229
Figure 1: MVV Cycle Route Planner – Route Example
Choosing the option “Bike”, the focus is more on the different type of bike, e.g. race bike or mountain bike, which has an influence on the surface condition. Choosing the adjustment “Bike and Ride” the trip duration and comfort are most important. The user is enabled to define which route section before or after using public transport he wants to go by bike. So they can decide for a route to the nearest station or choose the maximum time by bike from origin or to the destination. The entrainment option is considering the off-period for bicycles in public transport and recommends which size of bike can be carried on the vehicles. Besides all the adjustments concerning a specific route recommendation, the route planner can also visualize different bike-related services and points of interest. There are a number of route recommendations for bike excursions out of the city. The bike and ride facilities at the stations and various places in the city are displayed as well as the locations for bike sharing bikes. Furthermore the whole public transport network with S-Bahn, U-Bahn, Tramway and Busses can be shown (MVV, LHM 2016). 5.2 Results of the data analysis of route calculations A data set with 414,000 route calculations in spring and summer 2015 was the basis for the data analysis of the cycle route planner. Corrected of inaccurate data and double draws, which constantly appear because for every single request there are always three different routes calculated, altogether 136,000 single requests and route calculations from April until August 2015 were examined. The dataset was analyzed with MS Excel and SPSS and visualized with ArcGIS. The different route choice proposals e.g. “Fastest Route” or “Green Route” were not assessed, but the coordinates of the origin and destination inputs. So the focus was on the spatial distribution of origins and destinations in Munich and the suburban region. Besides that also the different adjustments concerning travel speed or combination with public transport were evaluated. The vast majority of 79.1% of all origin inputs are within the municipal area of Munich. The suburban communities are clearly underrepresented with the municipality of Fürstenfeldbruck (0.8%) and Unterföhring (0.7%) on rank two and three. So far the cycle route planner was mainly advertised in the city of Munich and is possibly rather unknown in the suburban areas. Regarding the inputs for destinations, still the municipal area of Munich is far ahead with 72.6% of all route calculations. But there is a significant increase of requests in the municipality of Starnberg (2.1%) and Unterföhring (1.1%). Starnberg represents a popular leisure destination for cyclists, because it’s also situated next to the Starnberger See.
230
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
Unterföhring is an important place of business with a large number of companies, which employ many commuters living in the city of Munich. Giving a closer look on the results of origin- and destination inputs within the city of Munich (see Figure 2), the dense populated and larger urban districts in the western (Nymphenburg-Neuhausen, Pasing-Obermenzing) and central-northern part (Maxvorstadt, Schwabing-Freiman) of the city represent the highest number for route calculations. The share of origins and destinations for each district does not differ much, but the city center has a higher demand for destinations than origins. Many users seem to simply calculate just a route to the center of the city.
ShareofOriginͲ andDestinationInputinMunich ShareinPercentage
12,0%
10,9% 10,0%
10,0%
8,4%
7,5%
8,0%
5,9%
6,2%6,5%
6,8% 5,9%
5,7%
6,0%
4,7%4,4%
4,7% 4,0%
5,1% 4,5%
4,4%4,5%
4,0%
4,3% 3,7%
2,0% 0,0%
CityDistricts OriginInput
DestinationInput
Figure 2: Share of OD-Input in Munich
Analysing just the share of input locations for origins and destinations in the suburban region of Munich, the dense populated municipalities in the north and south, next to the city boundary, as well as the suburban business places are high-ranked. The community of Starnberg is once more ahead, comparing the most requested destinations for cyclists in the suburban region (see Figure 3).
ShareofOriginͲ andDestinationInputintheSuburbanRegion 8,0% ShareinPercentage
6,3%
6,0% 4,1%
4,0% 2,0%
3,2%
3,0% 2,6%
2,8% 1,8%
2,4%2,2%
2,3% 1,7%
2,0% 1,6%
2,0%
2,1% 1,9%
2,4% 1,8%
1,8% 0,8%
0,0%
SuburbanCommunities OriginInput Figure 3: Share of OD-Input in Suburban Region of Munich
DestinationInput
1,7% 0,7%
231
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
Concentrating on the different trip adjustments there is a very clearly result that 98.6% of all route calculations are done for exclusive use of the bike. Only 0.8% consider the option bike entrainment and just 0.6% the combination of Bike and Ride. The route planner is currently predominately used for cycle trips without intermodal connection of public transport. The possibility to change the speed level in the settings is marginally used. The speed level is only changed for 5% of the calculated trips from 14 km/h to a higher level of speed. 1.4% of the calculated trips are on 20km/h, which might be an indicator that there is a small group of users, going frequently faster for example by the use of E-Bikes. The majority of trip calculations are taking place during the week on working days with 78%. 22% of the requests are generated on weekends. The purpose of these requests will be later answered in the paper. At the moment the large part of requests are made on personal computers with 88.3%. 11.7% are done on mobile devices. Since April 2015 more than 500,000 routes were calculated with the cycle route planner. Especially in the beginning, when the route planner was introduced and got a lot of attention by media and press, the requests increased to more than 100,000 per month. In the summer months the interest of users was still on a high level and decreased in September and autumn to 30,000 to 40,000 requests per month (see Figure 4). This course of the year can be seen also in other route planners, e.g. VVS Cycle Route Planner, due to bad weather conditions and less cyclists in the winter period.
CycleRoutePlannerͲ RouteCalculationspermonth 140000
NumberofCalculations
120000
114,407
100000
109,173 89,983
94,301
93,112
80000 60000
46,739 38,614
40000
32,640
20000 0 April2015
May2015
June2015
July2015
August2015
September2015
October2015
November2015
Figure 4: Route Calculations per month in 2015
Regarding the spatial distribution of user requests and route calculations within the cycle route planner, the statistical picture of the data analysis becomes visualized in Figure 5 and 6. They are showing the spatial distribution of route calculations for origins and destinations based on city districts. The amount of calculated trips is in the range of four-digit-numbers and illustrate a high request for trip calculations in the city centre and towards directions to west and north. Both origins and destination are situated in almost identical city districts. Besides the balance points of spatial distributions of the route calculations, the maps are also visualizing six potential cycle highways, starting in the city centre and running in all cardinal directions. The cycle highway-network is based on a study which was recently published and shows the capability of cycle highways in the metropolitan region of Munich. Out of 15 possible cycle highways with a length between ten and 20 kilometers, six routes with a high estimated demand were chosen, for an individual feasibility study. The general goal of the cycle highways is to attract a significant number of commuters to frequently use their bike or E-Bike to go to work, for instance. In the long-term car traffic congestion on the commuter routes could decrease through a shift from motorized commuter traffic to the bicycle or E-Bike. Especially the routes to the north-east and west are indicating a high demand for possible cycle traffic (PV München 2015).
232
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
Figure 5: Spatial Distribution of Origins in the City of Munich
Figure 6: Spatial Distribution of Destinations in the City of Munich
The particular advantage of the MVV-Cycle route planner is the incorporation of the suburban and rural areas around the city. As visualized in Figure 7 and 8 there is also a significant number of requests for trips in the rural parts of the greater region, although the amount of calculated routes is clearly smaller and in the range of three-digitnumbers. Again the difference between the amount of requests both of origins and destinations is marginal, although the destination points in the suburban region around the city boundary is slightly higher than the origin points. There is a slight concentration of request in the north-east in the direction of the International Airport and the business centres in the fringe area. Also the south-west shows a higher demand for origins and destinations, where popular excursion places next to recreational areas and lakes are situated. In summary the results indicate a concentric decrease of calculated trips with further distance from of the boundaries and expand far out of Munich into all directions also so the Austrian border in the south.
Figure 7: Spatial Distribution of Origins in the Region
Figure 8: Spatial Distribution of Destinations in the Region
A closer look on the top 50 coordinates for origins and destinations in Munich and the region, as shown in Figure 9 is emphasizing the previous results. Important central places and transport hubs in Munich are primarily chosen for route calculations of the users. Those places are complemented by high-frequented public transport locations and terminal stops in the region. This is a strong indicator for leisure trips by bike into the recreational areas in the south with large forests and lakes. There is also a concentration of locations north of the city centre, along the
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
underground line U6 to the University Campus in Garching. Beyond that the airport and the densely populated cities of Erding and Freising are also in the List of the top 50.
Figure 9: Top 50 Coordinates for Origins and Destinations in Munich and the Region
5.3 Results of a Customer Survey Unfortunately the results of the data analysis give no information about the very important question, if the cycle route planner and the calculated trips are real cycling trips. There is no hard evidence in the data set for what purpose the route calculations are retrieved and how often a cycling trip in reality follows. How big is the likelihood that a calculated route is actually cycled afterwards? How reliable is the information and recommendation for special routes in general? How do the users assess the different functions of the route planner and how strong is the impact of route recommendations on the actual choice of the cyclists. Those questions were the main reason to carry out a survey among users and non-users of the cycle route planner. The survey was designed on the internet-platform “umfrageonline.com” and conducted during a period of four weeks from beginning of November until December 2015. The questionnaire was finally completed with 294 participants at the age between 18 and 78 years. 10,049 single answers were given to 22 questions about the cycle route planner itself, mobility patterns and sociodemographic characteristics. The survey was publically announced in the “Rathaus Umschau”, the city hall news. There were several press reports in domestic newspapers and various references on Munich Cycling Websites (ADFC München, ADFC Bayern) as well as Facebook announcements. The questionnaire consisted of six main chapters: x x x x x x
Usage and assessment of the cycle route planer (only for users) Evaluation of route choice and route proposals (only for users) Usage of the route planner for real trips by bike General use of transport modes and typical mobility patterns Commuter-patterns Sociodemographic characteristics of the participants
The first question was designed to filter the participants who already had experience in using the route planner, from the non-users. The proportion was nearly equal with 52% of participants who already used the route planer and 48% of non-users. The following ten questions were only answered by participants, who already were experienced using the route planner. Figure 10 gives an answer to the purpose of using the cycle route planner, which was one of the central questions and couldn’t be clarified by the data analysis. The results show that 70.5% are using the route planner to plan a leisure trip by bike. 56.1% of the participants use it to get an advice for their cycling trip to work.
233
234
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
These two categories are predominately. The other notable results are the categories “without a precise intention” (28.8%) or “for orientation on a digital map” (23.5%).
ForwhichreasondidyouusetheCycleRoutePlanner?
N=132
(multipleanswers)
AdviceforLeisureTrips
70,5%
AdviceforTripstoWork
56,1% 28,8%
WithoutaPreciseIntention
23,5%
ForOrientationonaDigitalMap
13,6%
ToCompareDistanceandJourney TimewithMotorizedTraffic
3,8%
0%
Others
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Figure 10: Purpose of using the Cycle Route Planner
Another essential question, which could not be explained from the previous data analysis, was the question about the likelihood of real cycle trips calculated in the route planner. There is also a clear result to that question in the survey: with an overall likelihood of 60.4% the participants are cycling a route, which was requested in the route planner (see Figure 11). So more than every second route calculation is representative for an actual cycle trip. It is still unclear how often and when the trip is done. Whatisthelikelihoodforreallycyclingaroute,whichwascalculatedintheroute planner?
N=142
30
NumberofMentions
25
=60.42%
20 15 10 5 0 0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95% 100%
PercentageLikelihood
Figure 11: Likelihood of cycling a calculated route
Two questions were dealing with the reliability and quality of the selected routes. Both results show a very consistent picture. In the first question the participants were asked to assess the impact of the proposed routes on their actual choice of bicycle route (Figure 12). The results show that almost 70% of the participants say that the
235
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
calculated routes in the route planner have a strong (59.2%) or rather very strong (9.2%) impact on their choice of route. Only 19.2% say the influence is rather low and both 6.2% say it’s low or it has no impact on their actual choice of route. The picture is completed by the second question about the deviation of the proposed routes from the preference for the individual choice of the participants (Figure 13). Here 20.6% say the deviation is low. Still almost 40% state that the deviation is rather low and only 34.4% agree that the deviation is strong. Just a very small minority of 3.1% assess the deviation very strong. Both results illustrate that the cycle route planner does have an additional benefit for the users and the calculated routes are suitable for a majority of the participants of this survey.
Howmuchdotheproposedroutes influenceyouractualchoiceofroute
Howstrongisthedeviationofthe proposedroutesfromyourown preferenceforanindividualroute?
N=130
6,2%
N=131
2,3%
9,2%
3,1%
6,2%
20,6% 19,2%
34,4%
59,2% 39,7%
verystrong
strong
ratherlittle
Figure 12: Impact on choice of route
little
notatall
verystrong
strong
ratherlittle
little
notatall
Figure 13: Deviation of proposed routes
Another question was dealing with the publicity of the cycle route planner. Nearly one third of the participants paid attention to this new offer by using the official MVV Website with the public transport journey planner. 20% were attracted over the official website of the City of Munich and another 20% learned about it by recommendation of friends and colleagues. The rest was attracted by online newsletter, bicycle websites and press releases. There is also a smartphone app existing, which gives the user the possibility to use the cycle route planner as navigation device. The app is so far rather unknown and only in use by 20% of the participants. Out of the 20% only 5% use it on the bike for direct route navigation. The general assessment of the cycle route planner was evaluated fairly good, as shown in Figure 14. Five categories were evaluated. The visual design was rated best, followed by the information output and the overall usability. The lower rated categories route-choice and input options are still on a good rating score.
236
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
HowdoyouliketheCycleRoutePlanner?
N=131
1
140 120 2,08
NumberofMentions
2,31
2,29
2,25
2
80 3 60 40
AverageRating
1,94
100
4
20 0
5 VisualDesign
notasessable
RouteChoice
verypoor
InputOptions
poor
middle
InformationOutput
good
Useability
verygood
Average
Figure 14: Assessment of the Cycle Route Planner
To conclude the first section of the survey, the users of the route planner were on the one hand asked in an open question what they like or dislike using the cycle route planner in particular. 60% of the given answers in total were positive, 30% negative an around 10% mixed. A lot of positive comments were on the visual design and the digital base map, which has an additional benefit compared to Google Maps. The arrangement of different windows on the map and the zoom factor was criticised. On the other hand the participants were asked for particular suggestions for improvement. The visualization of all three route options at one time was mentioned several times just as well as the wish to display traffic lights, construction sites on the routes or dangerous spots. A more clearly hint on the surface quality was also mentioned as well as the possibility to give ideas for improvements on the website directly. The second part of the questionnaire focused on mobility patterns and the individual use of transport and was asked to all the participants. A large majority of them are in possession of a bicycle (96.2%) and 65% have a car available. The fact that 38% have a permanent subscription for public transport and 20.5% are members of Car Sharing companies indicate that the users of the cycle route planner share a very multi-modal transport behaviour. The bicycle is also the mean of transport, which is used most frequently in everyday life, followed by public transport and the private car. E-Bikes, Bike Sharing and motorcycles play only an insignificant role (Figure 15).
237
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
Howoftenareyouusingthefollowingtransportmodes?
N=263
1
300
250
1,8
2,48
150
3
3,08
AverageRating
NumberofMentions
2 200
3,61
100
4 50
4,35
4,63 4,68
4,67
4,75
0
5
CarasDriver
Never
Caras Passenger
Motorcycle
Lessthan1xamonth
Bicycle
EͲBike
Upto4xamonth
Public Transport
CarSharing BikeSharing
Severaltimesaweek
Daily
Others
Average
Figure 15: Usage of Transportation
This issue was followed by the question, what mode of transport the participants do usually use to go to work (Figure 16). Here again the bicycle (1.72 points on a scale up to 8) dominates the field of different transport possibilities. Closely following is public transport (2.24 points) also with a high number of given answers. Using the private car and walking to work or the place of education are ranked on the following places. There is a small number of participants who are using the combination of Bike and Ride to go to work. Most seldom means of transport, which are used to go to work are E-Bikes and the combination of car and public transport, Park and Ride.
238
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
WhichModeofTransportdoyouusuallyusetogotowork?
N=245 1
250 1,72
2
2,24
NumberofMentions
2,76
3
150
3,47 4
100
4,66
4,69
4,82
50
5
AverageRanking
200
6
6,12 0
7
Car notused
Motorcycle
Bicycle
8=mostseldom
7
EͲBike 6
Public Transport 5
4
Walking 3
2
Park&Ride
Bike&Ride
1=mostfrequently
Average
Figure 16: Mode of Transport to go to work
To estimate the capability of the cycle route planner for daily trips by bike to work, the participants were asked to specify their distance to work or their place of education. The results illustrate that shortly one third have a distance between zero and five kilometers to go to work. Another third of the participants (29.9%) have to cover a distance of five to ten kilometers to work. Summing up two thirds of all participants need to travel distances between zero and ten kilometers to work, which is rather an opportunity to cover the trips by bicycle. Only one third have longer travel distances in between 10 and more than 20 kilometers. This result is reflected in another question asked, considering the period of time, the participants need to go to work. Here also one third are having a trip duration between zero and 20 minutes (32.2%) to travel to work. Another third (29.3%) has a trip length between 20 and 30 minutes, corresponding with the distance of five to ten kilometers. 20.2% need to go to work between 30 and 40 minutes. The rest needs longer than 40 minutes (18.2%). To conclude the second part of the questionnaire about mobility patterns and the individual use of transportation, the participants were asked to give their approval to different response categories, concerning the question if they would use the bicycle more often under different circumstances (Figure 17). The approval rating reached the highest level for categories representing a better bicycle infrastructure. Bicycle highways for faster connections and an improvement of the existing cycling infrastructure with winter services or lighting were rated best. Also the safety for cyclists is an important issue as well as the weather conditions was ranked behind. The availability of an E-Bike or a private car does not have a significant impact on using the bicycle more often.
239
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
Wouldyouusethebicyclemorefrequentlyforyourdailytrips,if...?
N=244
250
1
200 1,5 1,7
1,66 1,83 1,96
2
2,19
100
2,28 2,39
AverageRanking
NumberofMentions
150
2,5 50 2,7 2,87 0
3 ...acyclehighway ...thequalityof ...thebicycle ...thebicycle wouldexist? cyclepathswould couldbe couldbe beimprovede.g. bettere/secure better/secure winterservice, storedathome? storedatwork? lighting,green wave?
Notassessable
No,certainlynot
...theriskfor accidentswould beverylow?
No,rathernot
...theweather ...thereisaEͲBike ...thereisnocar ...thereisnot wouldhaveno available? available? publictransport influence? stationcloseͲby?
Yes,eventually
Yes,certainly
Average
Figure 17: Reasons for more frequently use of the bike
The third and last part of the survey was dealing with sociodemographic characteristics of the participants. The average age of all participants was 44.2 years, which represents almost exactly the average age of the population in Germany (44.3 years). As stated at the beginning the youngest participant was 18 and the oldest 78 years old. 62.7% of the participants were male, 37.3% female. Two thirds of them are in full-time employment, 10.2% retired persons and only 3.3% students. A high majority of 77.4% holds a university degree. In a closing statement the attendees of the questionnaire were asked in an open question to give feedback and make suggestions. Altogether 74 persons gave individual answers, which were later categorized. Most of the feedback was concerning the desire for improvements of cycling facilities in general and the cycling infrastructure in particular. A few comments were dealing with the survey itself and some of the participants passed criticism on the question about reasons for more frequently use of the bike. The question was initially designed to question only the persons, who are not regularly using their bike, but after some changes before the survey was launched, this question was posed to all persons and caused some irritation of the already frequent bike users. Another category with several feedback was the explicit appreciation of the cycle route planner. Furthermore there were many comments on miscellaneous issues, e.g. to increase the awareness level for the cycle route planner, to improve the App for mobile phones, or to integrate traffic lights and construction sites.
240
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
6. Conclusion The cycle route planner is a valuable tool for all citizens to encourage cycling not only for leisure, but also in everyday life. The lack of route planning information for cyclists in an urban area is served with this special offer and enables the users also to combine cycling with public transport. The results described in this paper illustrate that there is a high demand for this kind of information and also a certain appreciation for the provided information, especially with focus on the particular needs of cyclists. There is still room for small improvements, but the main system is working on a solid base. In summary the main results of the data analysis and customer survey can be described as followed: x x x x x x
Most of the calculated routes concern the urban area of Munich The route planner is predominately used for leisure trips by bike and to plan trips to work 60% of the calculated routes are real trips on the bicycle The route planner is rated overall very well The route planner is mainly used for exclusive use of the bicycle without combination of public transport The route planner is primarily used by people who already often use the bike
Acknowledgements The authors are very grateful to the Munich Transport and Tariff Association and the Department of Environment and Health for providing the data of the Cycle Route Planner and supporting the ongoing research in this project.
References BEHESHTITABAR, E., AGUILAR RÍOS, S., KÖNIG-HOLLERWÖGER, D., SVATÝ, Z., RYDERGREN, C. 2014: Route Choice Modelling for Bicycle Trips. International Journal for Traffic and Transport Engineering, 2014, 4(2): 194 – 209. DIFU (= Deutsches Institut für Urbanistik) 2012: Thiemann-Linden, J., van Boeckhout, S. Mapping – Routing – Navigation for cycling. Berlin. DÖLGER, R. 2013: VeRa - Radroutenplaner Deutschland. Verknüpfte Routenplaner und Informationsdienste für Fahrradfahrer. Folienvortrag auf dem 3. Nationalen Radverkehrskongress am 13. und 14. Mai 2013 in Münster. Münster. GLEAVE, S. G. 2012: Cycle route choice. Final survey and model report. London. HUNT, J.D., ABRAHAM, J.E., 2007: Influences on bicycle use. Transportation, vol. 34, no. 4, pp.453-470. LENZ, S. 2015: Der Radroutenplaner Baden-Württemberg. In: Die Gemeinde, Stuttgart: Gemeindetag Baden-Württemberg; 138 (2015); Nr. 4; S. 215-216. LHM (=Landeshauptstadt München) 2010: Mobilität in Deutschland (MiD). Alltagsverkehr in München, im Münchner Umland und im MVV-Verbundraum. München. MVV (=Münchner Verkehrs- und Tarifverbund), LHM (=Landeshauptstadt München) 2015: MVV-Radrouten-Planer. Flyer. PV MÜNCHEN (=Planungsverband Äußerer Wirtschaftsraum München) 2015: Radschnellverbindungen in München und Umland. München. SINGLETON, A., LEWIS, D. 2011: Including Accident Information in Automatic Bicycle Route Planning for Urban Areas. Urban Studies Research, Volume 2011, Article ID 362817. New York. WILLIS, D., MANAUGH, K., ELǦGENEIDY, A. 2013: Cycling Under Influence: Summarizing the influence of attitudes, habits, social environments and perceptions on cycling for transportation. Paper accepted for presentation at the 92nd Transportation Research Board Annual Meeting, Washington D.C., USA. YEBOAH, G., ALVANIDES, S. 2015: Route Choice Analysis of Urban Cycling Behaviors Using Open Street Map: Evidence from a British Urban Environment. In book: OpenStreetMap in GIScience: Experiences, Research, and Applications. Springer, pp.189-210.
ScienceDirect Transportation Research Procedia 19 (2016) 225 – 240
International Scientific Conference on Mobility and Transport Transforming Urban Mobility, mobil.TUM 2016, 6-7 June 2016, Munich, Germany
Evaluation of Munich’s Cycle Route Planner Data Analysis and Customer Survey Florian Paul a *, Klaus Bogenberger a, Bernhard Fink b a
University of the Federal Armed Forces, Werner-Heisenberg-Weg 39, 85579 Munich, Germany b Munich Transport and Tariff Association (MVV), Thierschstr. 2, 80538 Munich, Germany
Abstract A convenient combination of pre- and on trip route information for cyclists are online routing tools. The Munich Transport and Tariff Association in cooperation with the Department of Environment and Health recently developed the MVV Cycle Route Planner. This online route-planning tool can be used either on desktop computers or on smartphones and covers the greater Munich region based on Open Street Map Data. It enables both users and municipalities to add information of cycle routes and other new bicycle infrastructure to the route planner and improve the system continuously. It provides a navigation and map service especially for the requirements of cyclists and can be combined with the use of public transport. This paper concentrates on three different research approaches. In a data analysis more than 130.000 single requests from April until August 2015 were examined with focus on the spatial distribution of origins and destinations in Munich and the suburban region. As a result, the demand and frequency of the user requests give a strong lead where cycle flows are to be expected and further infrastructure, like cycle super highways or improved cycle routes, needs to be considered. A customer survey of users and non-users revealed their mobility patterns and the actual use of the cycle route planner for trips to work or for leisure. The survey focused on usage and assessment of the cycle route planner itself. It further analyzed which means of transport are used for frequent trips and what the restrictions are so far, not to use the bicycle. © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license © 2016 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of mobil.TUM 2016. Peer-review under responsibility of the organizing committee of mobil.TUM 2016. Keywords: Bicycle; E-Bikes; Route Planning; Cycle Highways; Customer Survey
* Corresponding author. Tel.: +49 89 60042529; fax: +49 89 60042501. E-mail address: [email protected]
2352-1465 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of mobil.TUM 2016. doi:10.1016/j.trpro.2016.12.083
226
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
1. Introduction Bicycle Route Planners are a very popular and efficient tool to promote cycling, not only for tourist destinations in rural areas, but also for everyday cycling in urban agglomerations. In combination with the increase of online mapping tools and geo-information systems based on crowdsourcing, everyone on personal computers and smartphones can use these route planners. Although the development of reliable and coherent cycling route networks is a big challenge and also depending on the special needs and demands of cyclists, there is a growing number of cycle route planners for many cities and regions. This paper is focusing on the MVV-Cycle Route Planner for the City of Munich, which has been developed by the Munich Transport and Tariff Association together with the Department of Environment and Health. It will give a brief overview about existing online-route planners and navigation tools for cyclists and then focus on the evaluation of Munich’s Cycle Route Planner. In Munich there is a current lack of cycle traffic data. Only seven automatic bicycle counting spots, spread over the whole city, are counting the number and direction of cyclists. There is no current data available about cycle traffic flows, the modal share or the satisfaction of cyclists with the conditions of cycle tracks, for example. The last large-scale study, called MIDMUC, was carried out in 2008 and showed a modal split of 13.6% for cycling in Munich (LHM 2010). There are several developments in the cycle network in the past, which indicate that this number did presumably grow in the last eight years. The evaluation of Munich’s Cycle Route Planner can help to deliver a current data input to quantify the bicycle traffic. It provides furthermore some evidence where the most important origins and destinations for cyclists in the city and the region are. Further research questions about recommendations to prioritize bicycle highways or enhance the existing bicycle infrastructure at certain locations will be discussed. 2. Literature Review So far, web-based cycle route planners are only subject of rare studies, due to the very recent implementation of these routing programs in the last years. Nevertheless there are several studies and research projects about bicycle route choice, using available datasets and surveys. HUNT and ABRAHAM (2006) for example presented results of a questionnaire on the influences on bicycle use. They figured out that cycling in mixed traffic on streets is more exhausting for cyclists than on separate bike lanes. Cycling becomes less exhausting in mixed traffic, as the level of experience increases. Concerning the cycling facilities at the destinations, they found out that secure bike parking is more important than showers. WILLIS, MANAUGH and EL-GNEIDY (2013) give a broad overview about travel behavior and mode choice concerning cycling. They summarize the literature on social and psychological factors, which have an influence on the choice for cycling and make suggestions to increase the bicycle mode share. A study from the United Kingdom is investigating a route choice analysis of urban cycling behaviors using OpenStreetMap (OSM). YEBOAH and ALVANIDES (2015) examined GPS track data of cyclists using OSM and show the importance of consistent cycling networks, in order to support higher destination accessibility and increase the route directness. Besides comfort and route directness the attribute safety plays a significant role for cyclists, considering different route choices. SINGLETON and LEWIS (2011) are combining bicycle accident information with bicycle route planning, using the example of London. They analyzed the spatial locations where bicycle accidents occurred and drew a comparison between the quickest route and an accident avoidance weighted route. For a certain number of evaluated trips, they show that routes avoiding areas of high accident volume did not increase the trip length significantly. Another study from London has investigated the decisions that cyclists make when deciding which route to take (GLEAVE 2012). The key considerations around route choice of the cyclists centered on choosing the safest routes, and avoiding traffic. Less experienced cyclists prefer routes with less traffic and a cycle lanes, avoiding also confusing junctions. Comparing distance-based routes and attribute-based routes BEHESHTITABAR ET AL. (2014) are presenting a bicycle route choice model that is based on a cost function. They tried to find out which parameters have higher influence on bicycle route choice, and how they were contributing. The resulting model is capable to predict the most probable path a regular commuter would take between two points located in a defined area. The model does not only find the shortest path for cyclists, but also prefers attributes for the most suitable route in terms of safety and comfort.
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
227
3. Characteristics of cycle route networks and requirements of different user groups Compared to classic car-orientated road networks, cycle route networks contain many different characteristics due to a various number of infrastructural differences in most European countries. Bicycle streets, one-way streets with contra-flow for cyclists, cycle paths, car-free greenways, traffic calmed streets and many more categories become often only visible to cyclists, if there are explicit and comprehensible sign posting, cycling maps or navigation systems. Depending on different cycling target groups, the requirements for route choice options vary largely. Families with children for example, look for routes that are free of car traffic, without heavy inclination, easy to follow and embedded in an attractive landscape. Commuters on the bicycle often prefer the most timesaving routes, which also are suitable for everyday use under different weather conditions. Sport cyclists, using a race bike, especially pay attention to a smooth and tarmacked road surface. Recreational cyclists are usually looking for attractive destinations with a convenient surrounding, like green spaces and attractive points of interests, like swimming lakes or beer gardens. Concerning the development of route planning for cyclists, signposting appeared at first to show important destinations and routes. Printed cycling maps and cycling travel guides also provided important information for cyclists over the last decades. With the appearance of online route planners and mobile applications in the last years, the requirements of the target groups could be satisfied even more specific (DIFU 2012). 4. Online Route Planners and Navigation Tools The emergence of digital geo-information tools on the internet had a strong impact on cycle navigation. Navigation software combined with location-based information platforms for special points of interest can help to promote cycling and encourage cyclists to coordinate their interactions and decisions. Online route planners are a useful example how cycling-related contents can be merged with location-based information (DIFU 2012). So far, navigation software and route planning programs were mainly developed for motorized traffic and public transport. The increasing number of cyclists in many European cities also raised the demand for high-quality cycle route planners, not just for tourism purposes, but also for everyday cycling. The route planners themselves differ depending on the source of the locations-based data associated with the map and different providers, like State Transport Departments, private companies or research projects. Depending on the main purpose for the trips, the route planners concentrate more or less on touristic cycling or non-leisure, everyday cycling. The level of interactivity can also play a significant role. Route selection modes, start and destination points, personal adjustments like speed level or type of cyclist, as well as data collection in form of GPS tracks, are crucial for the development of a cycle route planner. The integration and connection of cycling route planners to public transport networks and journey planner are a crucial improvement and can help to support an intermodal way of transport. According to the particular provider of the route planner, there are different priorities of mapping information. Whereas regions and federal countries in Germany focus more on a touristic use, municipalities and communities orientate more on mapping networks for every-day use for cyclists (DÖLGER 2013). One of the first bike route planners in Germany was created in 2011 by the public transport association of BadenWürttemberg (NVBW). The NVBW is in charge of the bicycle route planner for the Federal State of BadenWürttemberg, by order of the Ministry for Transport and Infrastructure. The route planner is focusing especially on intermodal route choice options in connection with public transport. The goal was to strengthen a sustainable and environment-friendly mobility and also to develop a new bicycle culture. The route planner is available also for smartphones and enables the user to plan their cycle routes in combination with all modes of public transport. The use of environmental-friendly means of transportation for intermodal trips should be as convenient as possible. The Regional Office for Environment, Measurement and Conservation is in charge of data administration and management. The local authorities are enabled to gather their information about improvements of bicycle infrastructure by using a web-based Geoinformation System (GIS). There are more than 40,000 kilometers of bicycle network currently included. The cooperation between cities and rural districts concerning the data administration is important for offering high quality and effective data of the bicycle network. Because of the close cooperation for data management between the different municipalities the bicycle route planner has a positive effect
228
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
on transport policy and touristic issues. The ongoing update of bicycle network data is the fundament for a successful performance and popularity among many cyclists in Baden-Württemberg (LENZ 2015). 5. Evaluation of Munich’s Cycle Route Planer The MVV-Bicycle Route Planner was introduced in April 2015 by the Munich Transport and Tariff Association together with the Department of Environment and Health. The route planner is integrated into the public transport journey planner and offers an individual bicycle navigation within the greater Munich region, far from the city boundaries. The bicycle network information and data set is based on Open Street Map (OSM). The mapping device is specialized on the needs of cyclists and crowdsourced, meaning that users are able to make updates which improve the system continuously. 5.1 Specifications of the MVV Cycle Route Planner There are many options for the users of the cycle route planner to get detailed information about the recommended routes. The following information is available: x x x x x x
Journey time and length of trip Roadbook (directions) Surface condition Elevation profile GPS-Tracks (download on mobile devices) Public Transport Journey planner
After entering an address or location for origin and destination the route planner always gives you three different proposals for route choice. The “Fastest Route” represents the quickest and not necessarily the shortest route. The “Green Route” leads the cyclist as much as possible along green spaces and park areas. The “Family Route” prefers explicitly cycle ways, traffic calmed streets and avoids as much as possible ascending slopes. Based on individual choices the user can also make different adjustments for each route. Besides the choice of a personalized average speed level, on which the trip duration is dependent, the user is able to decide whether he wants to avoid ascending slopes. He can further define the surface conditions by choosing the type of bicycle (race bike or mountain bike) and can specify if he is en route with a bicycle trailer for children. After having made those decisions, the user is able to decide between three different trip adjustments: x x
x
Bike: exclusive use of the bicycle Bike and Ride: by bike to a station and then using public transport Bike entrainment: using the bike to a station, bike carried with public transport and cycling to the destination
A route example with three different proposals for route choice is shown in Figure 1.
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
229
Figure 1: MVV Cycle Route Planner – Route Example
Choosing the option “Bike”, the focus is more on the different type of bike, e.g. race bike or mountain bike, which has an influence on the surface condition. Choosing the adjustment “Bike and Ride” the trip duration and comfort are most important. The user is enabled to define which route section before or after using public transport he wants to go by bike. So they can decide for a route to the nearest station or choose the maximum time by bike from origin or to the destination. The entrainment option is considering the off-period for bicycles in public transport and recommends which size of bike can be carried on the vehicles. Besides all the adjustments concerning a specific route recommendation, the route planner can also visualize different bike-related services and points of interest. There are a number of route recommendations for bike excursions out of the city. The bike and ride facilities at the stations and various places in the city are displayed as well as the locations for bike sharing bikes. Furthermore the whole public transport network with S-Bahn, U-Bahn, Tramway and Busses can be shown (MVV, LHM 2016). 5.2 Results of the data analysis of route calculations A data set with 414,000 route calculations in spring and summer 2015 was the basis for the data analysis of the cycle route planner. Corrected of inaccurate data and double draws, which constantly appear because for every single request there are always three different routes calculated, altogether 136,000 single requests and route calculations from April until August 2015 were examined. The dataset was analyzed with MS Excel and SPSS and visualized with ArcGIS. The different route choice proposals e.g. “Fastest Route” or “Green Route” were not assessed, but the coordinates of the origin and destination inputs. So the focus was on the spatial distribution of origins and destinations in Munich and the suburban region. Besides that also the different adjustments concerning travel speed or combination with public transport were evaluated. The vast majority of 79.1% of all origin inputs are within the municipal area of Munich. The suburban communities are clearly underrepresented with the municipality of Fürstenfeldbruck (0.8%) and Unterföhring (0.7%) on rank two and three. So far the cycle route planner was mainly advertised in the city of Munich and is possibly rather unknown in the suburban areas. Regarding the inputs for destinations, still the municipal area of Munich is far ahead with 72.6% of all route calculations. But there is a significant increase of requests in the municipality of Starnberg (2.1%) and Unterföhring (1.1%). Starnberg represents a popular leisure destination for cyclists, because it’s also situated next to the Starnberger See.
230
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
Unterföhring is an important place of business with a large number of companies, which employ many commuters living in the city of Munich. Giving a closer look on the results of origin- and destination inputs within the city of Munich (see Figure 2), the dense populated and larger urban districts in the western (Nymphenburg-Neuhausen, Pasing-Obermenzing) and central-northern part (Maxvorstadt, Schwabing-Freiman) of the city represent the highest number for route calculations. The share of origins and destinations for each district does not differ much, but the city center has a higher demand for destinations than origins. Many users seem to simply calculate just a route to the center of the city.
ShareofOriginͲ andDestinationInputinMunich ShareinPercentage
12,0%
10,9% 10,0%
10,0%
8,4%
7,5%
8,0%
5,9%
6,2%6,5%
6,8% 5,9%
5,7%
6,0%
4,7%4,4%
4,7% 4,0%
5,1% 4,5%
4,4%4,5%
4,0%
4,3% 3,7%
2,0% 0,0%
CityDistricts OriginInput
DestinationInput
Figure 2: Share of OD-Input in Munich
Analysing just the share of input locations for origins and destinations in the suburban region of Munich, the dense populated municipalities in the north and south, next to the city boundary, as well as the suburban business places are high-ranked. The community of Starnberg is once more ahead, comparing the most requested destinations for cyclists in the suburban region (see Figure 3).
ShareofOriginͲ andDestinationInputintheSuburbanRegion 8,0% ShareinPercentage
6,3%
6,0% 4,1%
4,0% 2,0%
3,2%
3,0% 2,6%
2,8% 1,8%
2,4%2,2%
2,3% 1,7%
2,0% 1,6%
2,0%
2,1% 1,9%
2,4% 1,8%
1,8% 0,8%
0,0%
SuburbanCommunities OriginInput Figure 3: Share of OD-Input in Suburban Region of Munich
DestinationInput
1,7% 0,7%
231
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
Concentrating on the different trip adjustments there is a very clearly result that 98.6% of all route calculations are done for exclusive use of the bike. Only 0.8% consider the option bike entrainment and just 0.6% the combination of Bike and Ride. The route planner is currently predominately used for cycle trips without intermodal connection of public transport. The possibility to change the speed level in the settings is marginally used. The speed level is only changed for 5% of the calculated trips from 14 km/h to a higher level of speed. 1.4% of the calculated trips are on 20km/h, which might be an indicator that there is a small group of users, going frequently faster for example by the use of E-Bikes. The majority of trip calculations are taking place during the week on working days with 78%. 22% of the requests are generated on weekends. The purpose of these requests will be later answered in the paper. At the moment the large part of requests are made on personal computers with 88.3%. 11.7% are done on mobile devices. Since April 2015 more than 500,000 routes were calculated with the cycle route planner. Especially in the beginning, when the route planner was introduced and got a lot of attention by media and press, the requests increased to more than 100,000 per month. In the summer months the interest of users was still on a high level and decreased in September and autumn to 30,000 to 40,000 requests per month (see Figure 4). This course of the year can be seen also in other route planners, e.g. VVS Cycle Route Planner, due to bad weather conditions and less cyclists in the winter period.
CycleRoutePlannerͲ RouteCalculationspermonth 140000
NumberofCalculations
120000
114,407
100000
109,173 89,983
94,301
93,112
80000 60000
46,739 38,614
40000
32,640
20000 0 April2015
May2015
June2015
July2015
August2015
September2015
October2015
November2015
Figure 4: Route Calculations per month in 2015
Regarding the spatial distribution of user requests and route calculations within the cycle route planner, the statistical picture of the data analysis becomes visualized in Figure 5 and 6. They are showing the spatial distribution of route calculations for origins and destinations based on city districts. The amount of calculated trips is in the range of four-digit-numbers and illustrate a high request for trip calculations in the city centre and towards directions to west and north. Both origins and destination are situated in almost identical city districts. Besides the balance points of spatial distributions of the route calculations, the maps are also visualizing six potential cycle highways, starting in the city centre and running in all cardinal directions. The cycle highway-network is based on a study which was recently published and shows the capability of cycle highways in the metropolitan region of Munich. Out of 15 possible cycle highways with a length between ten and 20 kilometers, six routes with a high estimated demand were chosen, for an individual feasibility study. The general goal of the cycle highways is to attract a significant number of commuters to frequently use their bike or E-Bike to go to work, for instance. In the long-term car traffic congestion on the commuter routes could decrease through a shift from motorized commuter traffic to the bicycle or E-Bike. Especially the routes to the north-east and west are indicating a high demand for possible cycle traffic (PV München 2015).
232
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
Figure 5: Spatial Distribution of Origins in the City of Munich
Figure 6: Spatial Distribution of Destinations in the City of Munich
The particular advantage of the MVV-Cycle route planner is the incorporation of the suburban and rural areas around the city. As visualized in Figure 7 and 8 there is also a significant number of requests for trips in the rural parts of the greater region, although the amount of calculated routes is clearly smaller and in the range of three-digitnumbers. Again the difference between the amount of requests both of origins and destinations is marginal, although the destination points in the suburban region around the city boundary is slightly higher than the origin points. There is a slight concentration of request in the north-east in the direction of the International Airport and the business centres in the fringe area. Also the south-west shows a higher demand for origins and destinations, where popular excursion places next to recreational areas and lakes are situated. In summary the results indicate a concentric decrease of calculated trips with further distance from of the boundaries and expand far out of Munich into all directions also so the Austrian border in the south.
Figure 7: Spatial Distribution of Origins in the Region
Figure 8: Spatial Distribution of Destinations in the Region
A closer look on the top 50 coordinates for origins and destinations in Munich and the region, as shown in Figure 9 is emphasizing the previous results. Important central places and transport hubs in Munich are primarily chosen for route calculations of the users. Those places are complemented by high-frequented public transport locations and terminal stops in the region. This is a strong indicator for leisure trips by bike into the recreational areas in the south with large forests and lakes. There is also a concentration of locations north of the city centre, along the
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
underground line U6 to the University Campus in Garching. Beyond that the airport and the densely populated cities of Erding and Freising are also in the List of the top 50.
Figure 9: Top 50 Coordinates for Origins and Destinations in Munich and the Region
5.3 Results of a Customer Survey Unfortunately the results of the data analysis give no information about the very important question, if the cycle route planner and the calculated trips are real cycling trips. There is no hard evidence in the data set for what purpose the route calculations are retrieved and how often a cycling trip in reality follows. How big is the likelihood that a calculated route is actually cycled afterwards? How reliable is the information and recommendation for special routes in general? How do the users assess the different functions of the route planner and how strong is the impact of route recommendations on the actual choice of the cyclists. Those questions were the main reason to carry out a survey among users and non-users of the cycle route planner. The survey was designed on the internet-platform “umfrageonline.com” and conducted during a period of four weeks from beginning of November until December 2015. The questionnaire was finally completed with 294 participants at the age between 18 and 78 years. 10,049 single answers were given to 22 questions about the cycle route planner itself, mobility patterns and sociodemographic characteristics. The survey was publically announced in the “Rathaus Umschau”, the city hall news. There were several press reports in domestic newspapers and various references on Munich Cycling Websites (ADFC München, ADFC Bayern) as well as Facebook announcements. The questionnaire consisted of six main chapters: x x x x x x
Usage and assessment of the cycle route planer (only for users) Evaluation of route choice and route proposals (only for users) Usage of the route planner for real trips by bike General use of transport modes and typical mobility patterns Commuter-patterns Sociodemographic characteristics of the participants
The first question was designed to filter the participants who already had experience in using the route planner, from the non-users. The proportion was nearly equal with 52% of participants who already used the route planer and 48% of non-users. The following ten questions were only answered by participants, who already were experienced using the route planner. Figure 10 gives an answer to the purpose of using the cycle route planner, which was one of the central questions and couldn’t be clarified by the data analysis. The results show that 70.5% are using the route planner to plan a leisure trip by bike. 56.1% of the participants use it to get an advice for their cycling trip to work.
233
234
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
These two categories are predominately. The other notable results are the categories “without a precise intention” (28.8%) or “for orientation on a digital map” (23.5%).
ForwhichreasondidyouusetheCycleRoutePlanner?
N=132
(multipleanswers)
AdviceforLeisureTrips
70,5%
AdviceforTripstoWork
56,1% 28,8%
WithoutaPreciseIntention
23,5%
ForOrientationonaDigitalMap
13,6%
ToCompareDistanceandJourney TimewithMotorizedTraffic
3,8%
0%
Others
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Figure 10: Purpose of using the Cycle Route Planner
Another essential question, which could not be explained from the previous data analysis, was the question about the likelihood of real cycle trips calculated in the route planner. There is also a clear result to that question in the survey: with an overall likelihood of 60.4% the participants are cycling a route, which was requested in the route planner (see Figure 11). So more than every second route calculation is representative for an actual cycle trip. It is still unclear how often and when the trip is done. Whatisthelikelihoodforreallycyclingaroute,whichwascalculatedintheroute planner?
N=142
30
NumberofMentions
25
=60.42%
20 15 10 5 0 0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95% 100%
PercentageLikelihood
Figure 11: Likelihood of cycling a calculated route
Two questions were dealing with the reliability and quality of the selected routes. Both results show a very consistent picture. In the first question the participants were asked to assess the impact of the proposed routes on their actual choice of bicycle route (Figure 12). The results show that almost 70% of the participants say that the
235
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
calculated routes in the route planner have a strong (59.2%) or rather very strong (9.2%) impact on their choice of route. Only 19.2% say the influence is rather low and both 6.2% say it’s low or it has no impact on their actual choice of route. The picture is completed by the second question about the deviation of the proposed routes from the preference for the individual choice of the participants (Figure 13). Here 20.6% say the deviation is low. Still almost 40% state that the deviation is rather low and only 34.4% agree that the deviation is strong. Just a very small minority of 3.1% assess the deviation very strong. Both results illustrate that the cycle route planner does have an additional benefit for the users and the calculated routes are suitable for a majority of the participants of this survey.
Howmuchdotheproposedroutes influenceyouractualchoiceofroute
Howstrongisthedeviationofthe proposedroutesfromyourown preferenceforanindividualroute?
N=130
6,2%
N=131
2,3%
9,2%
3,1%
6,2%
20,6% 19,2%
34,4%
59,2% 39,7%
verystrong
strong
ratherlittle
Figure 12: Impact on choice of route
little
notatall
verystrong
strong
ratherlittle
little
notatall
Figure 13: Deviation of proposed routes
Another question was dealing with the publicity of the cycle route planner. Nearly one third of the participants paid attention to this new offer by using the official MVV Website with the public transport journey planner. 20% were attracted over the official website of the City of Munich and another 20% learned about it by recommendation of friends and colleagues. The rest was attracted by online newsletter, bicycle websites and press releases. There is also a smartphone app existing, which gives the user the possibility to use the cycle route planner as navigation device. The app is so far rather unknown and only in use by 20% of the participants. Out of the 20% only 5% use it on the bike for direct route navigation. The general assessment of the cycle route planner was evaluated fairly good, as shown in Figure 14. Five categories were evaluated. The visual design was rated best, followed by the information output and the overall usability. The lower rated categories route-choice and input options are still on a good rating score.
236
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
HowdoyouliketheCycleRoutePlanner?
N=131
1
140 120 2,08
NumberofMentions
2,31
2,29
2,25
2
80 3 60 40
AverageRating
1,94
100
4
20 0
5 VisualDesign
notasessable
RouteChoice
verypoor
InputOptions
poor
middle
InformationOutput
good
Useability
verygood
Average
Figure 14: Assessment of the Cycle Route Planner
To conclude the first section of the survey, the users of the route planner were on the one hand asked in an open question what they like or dislike using the cycle route planner in particular. 60% of the given answers in total were positive, 30% negative an around 10% mixed. A lot of positive comments were on the visual design and the digital base map, which has an additional benefit compared to Google Maps. The arrangement of different windows on the map and the zoom factor was criticised. On the other hand the participants were asked for particular suggestions for improvement. The visualization of all three route options at one time was mentioned several times just as well as the wish to display traffic lights, construction sites on the routes or dangerous spots. A more clearly hint on the surface quality was also mentioned as well as the possibility to give ideas for improvements on the website directly. The second part of the questionnaire focused on mobility patterns and the individual use of transport and was asked to all the participants. A large majority of them are in possession of a bicycle (96.2%) and 65% have a car available. The fact that 38% have a permanent subscription for public transport and 20.5% are members of Car Sharing companies indicate that the users of the cycle route planner share a very multi-modal transport behaviour. The bicycle is also the mean of transport, which is used most frequently in everyday life, followed by public transport and the private car. E-Bikes, Bike Sharing and motorcycles play only an insignificant role (Figure 15).
237
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
Howoftenareyouusingthefollowingtransportmodes?
N=263
1
300
250
1,8
2,48
150
3
3,08
AverageRating
NumberofMentions
2 200
3,61
100
4 50
4,35
4,63 4,68
4,67
4,75
0
5
CarasDriver
Never
Caras Passenger
Motorcycle
Lessthan1xamonth
Bicycle
EͲBike
Upto4xamonth
Public Transport
CarSharing BikeSharing
Severaltimesaweek
Daily
Others
Average
Figure 15: Usage of Transportation
This issue was followed by the question, what mode of transport the participants do usually use to go to work (Figure 16). Here again the bicycle (1.72 points on a scale up to 8) dominates the field of different transport possibilities. Closely following is public transport (2.24 points) also with a high number of given answers. Using the private car and walking to work or the place of education are ranked on the following places. There is a small number of participants who are using the combination of Bike and Ride to go to work. Most seldom means of transport, which are used to go to work are E-Bikes and the combination of car and public transport, Park and Ride.
238
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
WhichModeofTransportdoyouusuallyusetogotowork?
N=245 1
250 1,72
2
2,24
NumberofMentions
2,76
3
150
3,47 4
100
4,66
4,69
4,82
50
5
AverageRanking
200
6
6,12 0
7
Car notused
Motorcycle
Bicycle
8=mostseldom
7
EͲBike 6
Public Transport 5
4
Walking 3
2
Park&Ride
Bike&Ride
1=mostfrequently
Average
Figure 16: Mode of Transport to go to work
To estimate the capability of the cycle route planner for daily trips by bike to work, the participants were asked to specify their distance to work or their place of education. The results illustrate that shortly one third have a distance between zero and five kilometers to go to work. Another third of the participants (29.9%) have to cover a distance of five to ten kilometers to work. Summing up two thirds of all participants need to travel distances between zero and ten kilometers to work, which is rather an opportunity to cover the trips by bicycle. Only one third have longer travel distances in between 10 and more than 20 kilometers. This result is reflected in another question asked, considering the period of time, the participants need to go to work. Here also one third are having a trip duration between zero and 20 minutes (32.2%) to travel to work. Another third (29.3%) has a trip length between 20 and 30 minutes, corresponding with the distance of five to ten kilometers. 20.2% need to go to work between 30 and 40 minutes. The rest needs longer than 40 minutes (18.2%). To conclude the second part of the questionnaire about mobility patterns and the individual use of transportation, the participants were asked to give their approval to different response categories, concerning the question if they would use the bicycle more often under different circumstances (Figure 17). The approval rating reached the highest level for categories representing a better bicycle infrastructure. Bicycle highways for faster connections and an improvement of the existing cycling infrastructure with winter services or lighting were rated best. Also the safety for cyclists is an important issue as well as the weather conditions was ranked behind. The availability of an E-Bike or a private car does not have a significant impact on using the bicycle more often.
239
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
Wouldyouusethebicyclemorefrequentlyforyourdailytrips,if...?
N=244
250
1
200 1,5 1,7
1,66 1,83 1,96
2
2,19
100
2,28 2,39
AverageRanking
NumberofMentions
150
2,5 50 2,7 2,87 0
3 ...acyclehighway ...thequalityof ...thebicycle ...thebicycle wouldexist? cyclepathswould couldbe couldbe beimprovede.g. bettere/secure better/secure winterservice, storedathome? storedatwork? lighting,green wave?
Notassessable
No,certainlynot
...theriskfor accidentswould beverylow?
No,rathernot
...theweather ...thereisaEͲBike ...thereisnocar ...thereisnot wouldhaveno available? available? publictransport influence? stationcloseͲby?
Yes,eventually
Yes,certainly
Average
Figure 17: Reasons for more frequently use of the bike
The third and last part of the survey was dealing with sociodemographic characteristics of the participants. The average age of all participants was 44.2 years, which represents almost exactly the average age of the population in Germany (44.3 years). As stated at the beginning the youngest participant was 18 and the oldest 78 years old. 62.7% of the participants were male, 37.3% female. Two thirds of them are in full-time employment, 10.2% retired persons and only 3.3% students. A high majority of 77.4% holds a university degree. In a closing statement the attendees of the questionnaire were asked in an open question to give feedback and make suggestions. Altogether 74 persons gave individual answers, which were later categorized. Most of the feedback was concerning the desire for improvements of cycling facilities in general and the cycling infrastructure in particular. A few comments were dealing with the survey itself and some of the participants passed criticism on the question about reasons for more frequently use of the bike. The question was initially designed to question only the persons, who are not regularly using their bike, but after some changes before the survey was launched, this question was posed to all persons and caused some irritation of the already frequent bike users. Another category with several feedback was the explicit appreciation of the cycle route planner. Furthermore there were many comments on miscellaneous issues, e.g. to increase the awareness level for the cycle route planner, to improve the App for mobile phones, or to integrate traffic lights and construction sites.
240
Florian Paul et al. / Transportation Research Procedia 19 (2016) 225 – 240
6. Conclusion The cycle route planner is a valuable tool for all citizens to encourage cycling not only for leisure, but also in everyday life. The lack of route planning information for cyclists in an urban area is served with this special offer and enables the users also to combine cycling with public transport. The results described in this paper illustrate that there is a high demand for this kind of information and also a certain appreciation for the provided information, especially with focus on the particular needs of cyclists. There is still room for small improvements, but the main system is working on a solid base. In summary the main results of the data analysis and customer survey can be described as followed: x x x x x x
Most of the calculated routes concern the urban area of Munich The route planner is predominately used for leisure trips by bike and to plan trips to work 60% of the calculated routes are real trips on the bicycle The route planner is rated overall very well The route planner is mainly used for exclusive use of the bicycle without combination of public transport The route planner is primarily used by people who already often use the bike
Acknowledgements The authors are very grateful to the Munich Transport and Tariff Association and the Department of Environment and Health for providing the data of the Cycle Route Planner and supporting the ongoing research in this project.
References BEHESHTITABAR, E., AGUILAR RÍOS, S., KÖNIG-HOLLERWÖGER, D., SVATÝ, Z., RYDERGREN, C. 2014: Route Choice Modelling for Bicycle Trips. International Journal for Traffic and Transport Engineering, 2014, 4(2): 194 – 209. DIFU (= Deutsches Institut für Urbanistik) 2012: Thiemann-Linden, J., van Boeckhout, S. Mapping – Routing – Navigation for cycling. Berlin. DÖLGER, R. 2013: VeRa - Radroutenplaner Deutschland. Verknüpfte Routenplaner und Informationsdienste für Fahrradfahrer. Folienvortrag auf dem 3. Nationalen Radverkehrskongress am 13. und 14. Mai 2013 in Münster. Münster. GLEAVE, S. G. 2012: Cycle route choice. Final survey and model report. London. HUNT, J.D., ABRAHAM, J.E., 2007: Influences on bicycle use. Transportation, vol. 34, no. 4, pp.453-470. LENZ, S. 2015: Der Radroutenplaner Baden-Württemberg. In: Die Gemeinde, Stuttgart: Gemeindetag Baden-Württemberg; 138 (2015); Nr. 4; S. 215-216. LHM (=Landeshauptstadt München) 2010: Mobilität in Deutschland (MiD). Alltagsverkehr in München, im Münchner Umland und im MVV-Verbundraum. München. MVV (=Münchner Verkehrs- und Tarifverbund), LHM (=Landeshauptstadt München) 2015: MVV-Radrouten-Planer. Flyer. PV MÜNCHEN (=Planungsverband Äußerer Wirtschaftsraum München) 2015: Radschnellverbindungen in München und Umland. München. SINGLETON, A., LEWIS, D. 2011: Including Accident Information in Automatic Bicycle Route Planning for Urban Areas. Urban Studies Research, Volume 2011, Article ID 362817. New York. WILLIS, D., MANAUGH, K., ELǦGENEIDY, A. 2013: Cycling Under Influence: Summarizing the influence of attitudes, habits, social environments and perceptions on cycling for transportation. Paper accepted for presentation at the 92nd Transportation Research Board Annual Meeting, Washington D.C., USA. YEBOAH, G., ALVANIDES, S. 2015: Route Choice Analysis of Urban Cycling Behaviors Using Open Street Map: Evidence from a British Urban Environment. In book: OpenStreetMap in GIScience: Experiences, Research, and Applications. Springer, pp.189-210.