2019-11-18, Nazemi, Mohsen, van Eggermond, Michael, Erath, Alexander, Joos, Michael, Schaffner, Dorothea, Axhausen, Kay W.

2018, Nazemi, Mohsen, van Eggermond, Michael, Erath, Alexander, Schaffner, Dorothea, Joos, Michael, Axhausen, Kay W.

Survey methods for bicycle research have been evolved in time, ranging from conventional tel- ephone surveys, paper-based and web-based surveys, to in-depth Virtual Reality (VR) ones, all aiming to provide insights about diversity among behaviours (McNeil et al., 2015; Tilahun et al., 2007; Xu et al., 2017). The underlying criteria in all of the aforementioned methods is that re- spondents need to understand what they are valuing or they will make potentially wrong as- sumptions based on different experiences and frames of reference. VR allows for highly-detailed observations, accurate behaviour measurements, and systematic environmental manipulations under controlled laboratory circumstances. It therefore has the potential to be a valuable re- search tool to carry out behavioural experiments to study cyclists’ perception of safety and com- fort. In order to fully understand VR as a valid environmental representation, it is essential to examine to what extent not only user cognition and behaviour, but also if users’ experiences are analogous in real and virtual environments (Kuliga et al., 2015). The aim of this research is to find out how accurately participants perceive speed and distance of the passing vehicles and find out solutions to calibrate virtual environments, particularly for bicycle research. To this end, a VR experiment was devised that allows respondents to cycle through a variety of streets, and in which the passing speed and distance of passing vehicles - two are crucial variables influencing cyclists’ perception of environment - is varied.

2016, van Eggermond, Michael, Erath, Alexander

In thispaper, we connect two notions of accessibility that are more often than not considered separately: pedestrian accessibility and transit accessibility. We move away from the notion of zonal accessibility and measure fine-grained accessibility using door-to-door travel times. Two pedestrian networks are compared to a baseline scenario considering Euclidean distances for a large metropolitan area in which each individual building is considered as an activity opportunity. It is shown that pedestrian accessibility to jobs differs when pedestrian distances are approximated with different networks that are more representative of reality. Stop-to-stop public transport travel times are extracted from an agent-based simulation of public transport smart card data. The effect of less-than-optimal connections from transit to the pedestrian network, a local measurement, can be seen when calculating the accessibility to all destinations in the city. We suggest moving away from Euclidean-based distance analyses. Limitations can be found in the data available; the connection of buildings to the network becomes important, as does the inclusion of pedestrian crossings. For an inclusive accessibility measure, it will be necessary to calculate generalized costs for pedestrians and generate different pedestrian networks that reflect the limitations of different user groups.