SUM: mobility status in Living Labs before interventions

SUM Mobility Survey

The objective of this survey is to investigate the travel preferences of the citizens of each Living Lab (LL). These travel preferences are collected before and after the implementation of transformative mobility measures per LL. The survey form was developed in English at the first stage and then it was translated to 8 different languages: German, Greek, Hebrew, French, Dutch, Polish, Portuguese and Norwegian. All datasets can be combined into a unique set of data that is used by the SUM project.

Survey design methods and techniques

The survey was designed as a revealed preferences survey form, where the objective is to record the travel behavior of LL citizens before and after the implementation of transformative measures. The first wave of the survey is distributed from mid-September to mid-November 2023.

One of the primary goals of this survey is to achieve a representative sample that reflects various age groups, genders, and employment statuses. It is of importance that we maintain this balanced distribution, especially when we plan the subsequent survey distribution process after the transformative measures are implemented. To increase the reach and flexibility of the survey distribution, multiple methods have been devised. The survey is distributed online for those who prefer digital interactions, additionally, for larger events, in-person survey distribution is available using tablets.

First wave (Sep-Nov 2023)

Modal split

Modal split refers to the share of trips per transport mode. As of mid-November 2023, a total of 147 respondents in Munich, had provided their insights by completing the SUM survey. A relatively high diversity of transport modes was observed, with cars accounting only for 18.5% of trips (Figure 2). A significant portion of travel involves shared mobility modes such as car-sharing, ride-hailing, and micro-mobility, which together constitute 11.4% of trips. This is the highest percentage compared to other LLs. Taxis also play a major role, responsible for 17.8% of travel. Public transport, including trains, metro, and buses, has a substantial share of 32.7%. Active modes like walking, bicycles, and e-scooters make up less than 7% of the total. Figure shows the modal split of Munich. It should be noted that the total number of trips that were described in this city was 297, which means approximately 2 trips per respondent.

In Athens, Penteli, 148 participants showed a strong preference for private cars, with 54.0% of the described trips made by this mode. The total number of trips included in this set was 298. Shared mobility, including car-sharing and ride-hailing, is almost negligible in Athens. Yet, conventional taxis account for 8.7% of trips, and public transport (trains, metro, and buses) makes up 22.9%. Walking is the main mode for 7% of short-distance trips, and motorcycles are used for 5%, while bicycle usage is very low. Figure 3 presents the transport mode distribution of Athens Penteli.

Geneva presents a stark contrast, with only 8.6% of trips made by car, likely due to the survey targeting public transport commuters. In this LL, 461 respondents had filled out the survey until mid-November 2023. Overall, 1022 trips were recorded. Shared mobility modes such as car-sharing, ride-hailing, and micro-mobility have small but notable shares of around 3%. Public transport dominates in this set of travel diaries, with a 40.1% share. Bicycles are significantly popular, being the primary mode for 31.7% of trips, and walking is preferred by 15.2% of respondents. In Figure 4, the modal split of Geneva is shown.

Rotterdam transportation scene is also diverse, with 49.4% of trips made by car. In this city, 488 respondents filled out the survey form; yet the diaries set includes only 434 trips. This means that some respondents choose to not give their present mobility patterns. Shared mobility modes have a combined share of approximately 4%, and conventional taxis are used for only 1.6% of trips. Public transport (trains, metro, and buses) accounts for 15.9%, and walking is the main mode for about 7% of trips. Notably, bicycle usage is the highest in this city, at around 21.9%. Figure 5 shows the modal split of Rotterdam.

In Krakow, cars are the dominant mode of transport, used for 50.7% of trips. Shared mobility modes have a minimal presence, with shares below 1%. Public transport, including trains to outside areas and buses within the city, accounts for 39.3% of total. Walking is chosen for 3.4% of trips, while bicycles and e-scooters have shares of 4% and 1.1%, respectively. The total number of trips included in the dataset was 349. In total, 306 respondents participated in the survey. Figure 6 shows the modal split of Krakow.

In Coimbra,135 participants showed a strong preference for private cars, with 67.3% of the described trips made by this mode. The total number of trips included in this set was 275. Trips with shared mobility modes including car-sharing and ride-hailing were not reported in this set of diaries. Yet, public transport (trains, metro, and buses) makes up 29.8%. Walking is the main mode for 1.1% of short-distance trips, and bicycles are used for 1.5% of trips. Figure 7 presents the transport mode distribution of Coimbra.

In Larnaca, cars are more than the dominant mode of transport, used for 96.9% of trips. Shared mobility modes have a minimal presence, with shares below 0.5%. Public transport including only buses within the city accounts for 2.5% of total trips. Walking is chosen for 2.5% of trips. In total 55 respondents participated in the survey. Figure 8 shows the modal split of Larnaca.

In Fredrikstad, the use of cars is particularly high at 63.2%. Unique to this city is the use of ferries, accounting for 7.7% of travel. Shared mobility modes are not widely integrated, except for car-sharing, which has a 2.6% share. Bicycles are used for 8.4% of trips, and buses account for 9.0%. The total number of trips included in this dataset was 155. In Figure 9, the modal split of Fredrikstad is shown.

Departure time

The forthcoming analysis unfolds through a series of Figures that intricately map out the temporal distribution of trips across various transport modes, delineated for each LL. These graphical representations serve a dual purpose: firstly, they lay bare the predominant transport modes that anchor the mobility framework of each LL, and secondly, they highlight the temporal windows when these modes are most and least utilized. These temporal windows refer to the peak and non-peak hours observed in each city.

The mobility patterns of Munich show a distinct morning peak starting at 8:00, with the afternoon peak spreading between 15:00 and 17:00. During these times, taxis, metro or train, and cars are the most frequently used modes of transport. Notably, metro usage spikes at 20:00, and bicycles become a popular alternative in these evening hours. Figure 10 shows the relationship between start trip times and transport modes in Munich.

Athens Penteli exhibits a different mobility pattern, heavily reliant on cars with high peaks observed from 9 to 11 in the morning and again from 17 to 18 in the evening. Car usage remains high even at 20:00, indicating a consistent preference for this mode of transport throughout the day. Public transport, especially the metro, emerges as a significant alternative, particularly during the morning and midday hours. Figure 11 shows the start time distribution of 298 trips reported in Athens Penteli.

In Geneva, the morning peak hour is sharply defined at 8:00, but a high proportion of trips also start at 11:00, mainly using public transport. Bicycle usage follows a similar time distribution, while walking trips increase during non-peak hours, reflecting a more spread-out pattern of mobility throughout the day. Figure 12 shows the relationship between start trip times and transport modes in Geneva.

Rotterdam peak hours mirror those of many cities, with morning traffic peaking at 8:00 and the afternoon peak at 17:00. Bicycle trips in Rotterdam align closely with the patterns of car and train usage, indicating a competitive situation between these three modes of transport. Figure 13 presents this temporal distribution.

Based on the survey data, Krakow travel behavior is characterized by an early start, with a significant number of trips reported between 5:00 and 8:00 in the morning, followed by a lower number in the afternoon peak. The city shows a particular trend where respondents often describe just one trip, leading to these distinctive patterns. Buses and cars are the dominant modes of transport for daily activities. Figure 14 shows the start time distribution of 349 trips reported in Krakow.

Coimbra peak hours follow the same patterns of those of many cities, with morning traffic peaking at 8:00 and the afternoon peak at around 17:00 to 18:00. Public transport trips in Krakow have the same time distribution of car usage, indicating a competitive situation between these two only transport modes in this city. Figure 15 presents this temporal distribution.

Larnaca peak hours mirror those of many cities, with morning traffic peaking at 8:00 and the afternoon peak at 17:00. Cars seem to be the only solution. Figure 16 presents this temporal distribution.

In Fredrikstad, mobility begins notably early, with significant activity starting at 5:00 in the morning. Cars dominate the peak hours, particularly at 17:00 and 20:00, showcasing a higher late-day activity compared to other cities. This early start and extended evening usage reflect Fredrikstad unique transportation trends. Figure 16 shows the relationship between start trip times and transport modes in Fredrikstad.

Trip purpose

This data analysis takes into account a third perspective by categorizing trip data based on the purpose behind each journey, with a focus again on the transport modes. The key purposes identified, include commuting to work, returning home, attending educational institutions, shopping, recreation, health-related visits, accessing other services, and other activities.

In Munich, the dataset predominantly contains recreation trips using a variety of transport modes, with cars, trains, or metro, and taxis being the most popular choices. Interestingly, only 39 work-related trips were reported. Trains are frequently used for shopping, work, and educational activities, whereas taxis do not share the same popularity in these categories as they do for recreational purposes. The previously mentioned insights are presented in the matrix of Figure 18.

Athens Penteli shows a different and more expected pattern with 67 work-related trips predominantly made using private cars. Cars are also a popular choice for recreational activities, accounting for 30 trips. This indicates that people use their cars for both commuting to work and leisure activities. Public transport and taxis have emerged as the main alternatives for work commutes. Interestingly, walking is a popular mode for both recreation and educational activities within the municipality of Penteli. The heatmap of Figure 19 gives the relationship between transport mode and trip purpose in Athens.

Geneva survey data primarily reflects 445 trips from home to work, with bicycles and walking being the primary modes of transport, indicating a uniform trend. These modes, seen as alternatives to cars, are used for multiple purposes, which is significant considering the respondents are primarily public transport commuters. The previously mentioned insights are presented in the matrix of Figure 20.

In Rotterdam, bicycles are commonly used for commuting to work and for recreational or shopping activities. The use of private bicycles for returning home completes the trip chain. The metro is primarily used for work commutes, while cars, due to their flexibility, are used for various activities including work, shopping, and others. Interestingly, Rotterdam reported the highest number of trips (10 trips) made for health-related reasons. The heatmap of Figure 21 gives the relationship between transport mode and trip purpose in Rotterdam.

Krakow reported mobility patterns are dominated by work commutes, primarily using buses and cars. The lower share of afternoon trips results in fewer trips being recorded for activities other than work, such as recreation, making it challenging to draw concrete conclusions considering this perspective. The previously mentioned insights are presented in the matrix of Figure 22.

Coimbra survey data primarily reflects 131 trips from home to work, with car and bus being the primary modes of transport. Moreover, these transport modes are highly used by the respondents both for education and recreation activities. The previously mentioned insights are presented in the matrix of Figure 23.

Larnaca reported mobility patterns are dominated by work commutes, primarily using cars (204 trips). The lower share of afternoon trips results in fewer trips being recorded for activities other than work. The previously mentioned insights are presented in the matrix of Figure 24.

Fredrikstad exhibits similar work commute patterns, with cars, buses, bicycles, and uniquely, ferries, being the main modes of transportation. The use of ferries for commuting to work highlights the distinct nature of Fredrikstad transport landscape, differentiating it from the other cities. The heatmap of Figure 25 gives the relationship between transport mode and trip purpose in Fredrikstad.

Transport system evaluation

Perceived safety

The interplay between perceived safety, in-vehicle travel time, and actual travel choices is a critical aspect of urban mobility. Safety concerns, particularly related to crash occurrence, can significantly deter individuals from using certain transport modes, like active modes. This perceived risk creates a tangible barrier to accessibility, often leading to a preference for 'safer' modes of transport, regardless of their speed or convenience.

In Munich, a notable 15.1% of respondents consider cars very unsafe, despite the ability to reach destinations within a 0-10 minute interval. The city has the lowest perceived safety rates across all transport modes, with safe ratings of 3 or higher not exceeding 25% for any mode. Walking is perceived as both the unsafest and slower mode, with an 18.1% safety rating at level 1 and estimated travel times of around 30-40 minutes. Cycling as a transport mode, however, sees most of the safety ratings concentrated at level 2, indicating a moderate level of safety concerns. Figure 34 presents the previously described insights.

Contrastingly, Athens Penteli displays a high variance in perceived safety ratings across cars, taxis, and walking, indicating a high heterogeneity among individuals. This diversity in opinions is less pronounced in public transport, where more than 70% rate safety at level 3 or higher. However, public transport is not perceived as particularly fast, with travel times estimated between 30 to 45 minutes. Motorcycles and bicycles are seen as fast but unsafe, with cycling perceived as less safe than motorcycles. Despite safety concerns, motorcycles are recognized for their speed and ability to quickly reach destinations compared to other private modes in Athens. The heatmaps of Figure 35 give the relationships between in-vehicle peak travel time and perceived safety per transport mode.

In the interest of brevity, and in order to collect the highest possible number of responses over a short time horizon, Geneva Living Lab decided not to ask questions relative to safety.

Rotterdam presents an interesting case with generally short in-vehicle travel times of around 10 minutes. Public transport, followed by cars and taxis, is perceived as the safest mode of transport. Motorcycles are viewed as the least safe, while bicycles receive higher safety ratings, with levels 4 and 5 being the most common, accounting for 50% of responses. Walking is also considered safe and interestingly not viewed as a very slow mode even during peak hours. Figure 36 presents the previously described insights.

Krakow shows high variance in safety and time perceptions across all transport modes, indicating diverse preferences among its residents. Cars, taxis, and walking generally receive higher safety ratings (3 or higher). Public transport is considered very safe but with an extended travel time of nearly 20 minutes. Motorcycles, while quick, receive more negative safety ratings. Some contradicting evaluations are presented in cycling, with safety perceptions ranging between levels 2 and 5, but travel times are mostly concentrated within the 20–30 minute range, compared to longer times for public transport. The heatmaps of Figure 37 give the relationships between in-vehicle peak travel time and perceived safety per transport mode.

Coimbra also reports a considerably high variance in perceived safety ratings across all transport modes. This diversity in opinions is also pronounced in public transport, where the safety rate ranges from 1: very unsafe to 7: very safe. However, public transport is not perceived as particularly fast, with travel times estimated between 30 to 45 minutes. Motorcycles and bicycles are seen as fast but unsafe. Yet, there is a high standard deviation in safety ratings, while around 30% of participants consider these modes as safe giving values from 3 to 6. Walking is very unsafe according to Coimbra citizens. The heatmaps in Figure 38 give the relationships between in-vehicle peak travel time and perceived safety per transport mode.

The survey results depicted in the graphs for Larnaca transport modes reveal varied perceptions of safety and travel time at peak hours. Car users rarely experience over 30 minutes of delay, with safety concerns being minimal. Taxi passengers generally feel safe, although a significant number expect longer peak hour in-vehicle travel times. In public transport, respondents note moderate safety and some peak-time delays, while motorcycle riders report a quite wide range of travel time differences yet feel unsafe overall. Survey participants expressed substantial safety concerns for bicycle using, contrasting with pedestrians who largely feel very safe in this city. Figure 39 presents the previously described insights.

Fredrikstad also exhibits high variance in safety and time perceptions, except for motorcycles, which are consistently viewed as the least safe. Cycling follows in terms of safety concerns. A significant concentration of responses, 16.1%, falls within the category of very unsafe and more than 90 minutes of walking time, suggesting that walking is not a favored option in Fredrikstad. The heatmaps of Figure 40 give the relationships between in-vehicle peak travel time and perceived safety per transport mode.

General satisfaction with mobility services

When it comes to citizens’ satisfaction with the provided mobility services, Athens shows a mean satisfaction score on the lower end, barely under 2.7, hinting at a desire for improvement in transport services. Coimbra's satisfaction score lowers further, with a mean of around 2.2, reinforcing the need for enhanced transport solutions. In contrast, Fredrikstad and Krakow exhibit higher satisfaction levels, with means above 3, suggesting a more favorable assessment. Larnaca, again with limited data, sits at a mean of 2.6. Munich and Rotterdam, similar to their policy acceptance scores, report high satisfaction means, approximately 4.4 and 4.5 respectively, indicating that residents are quite content with their transportation systems. The next histograms (Figures 41- 47) give the distribution of satisfaction scores.