The way we walk

Mobile tools for wayfinding combined with Location Based Services (LBS) can provide pedestrians with practical information concerning optimal routes and useful facilities in their vicinity. However, what is considered as “optimal” and “useful” largely varies between different kinds of individuals. Inappropriate information may hinder effective information extraction for a person seeking specific navigational and environmental information; a successful mobile spatial information service should therefore be based on a profound understanding of pedestrian spatio-temporal behaviour. A current study applies an “across-method” triangulation approach for studying human spatial behaviour, combining localisation and tracking techniques as well as inquiries concerning intentions, lifestyle attributes and socio-demographic characteristics in order to define a pedestrian typology of mobility styles.

Introduction

The rapid development in the field of mobile information and communication technologies as well as the increasing amount of ubiquitously available information offer a wide range of possibilities to supply mobile users with location based information. In recent years, ubiquitous computing technologies have made it possible for individuals to gain ubiquitous access to information services. Facilitated access to online resources and increasing availability of location related information now give fresh impetus to the development of mobile navigation tools for pedestrians. However, the increasing availability of various kinds of information also leads to a higher risk of information overload. A successful mobile spatial information service for pedestrians must therefore be able to provide useful instructions with respect to the individual’s requirements and the specific context, and avoid redundant information.

A currently ongoing project examines the basic requirements for the development of mobile wayfinding tools based upon ubiquitous cartography. The aim is to provide pedestrians with efficient and practical information using a combination of active and passive systems in a smart environment. The wayfinding process is to be enhanced with additional, location based information and various representation forms. We aim at the description of specific types of pedestrian route choice behaviour and interest foci, based on observed motion behaviour as well as lifestyle related attributes. Results are not only crucial for personalising navigational and environmental information, but can also be used for determining motion parameters in pedestrian simulation models. This article outlines the applied methodology for monitoring and classifying pedestrian spatio-temporal behaviour as well as initial results from the first of two consecutive empirical phases (for a more detailed description of this work see [1]).

Investigating Pedestrian Spatio-temporal Behaviour

Several studies prove that there are differences in the way pedestrians choose a path to a particular desired destination: Pedestrians often prefer routes offering different qualities than simple shortness, e.g. the “most beautiful”, the “most convenient” or the “safest” [2-4]. Golledge [5] found that “fewest turns” (simplest path) appeared to be one of the most relevant factors influencing route planning strategies. Other factors mentioned in his results are e.g. “least time”, “most scenic/aesthetic”, or “different from previous (novelty)”. All these findings support the assumption that the choice of a specific route and the actual walking behaviour depend on a variety of influence factors, like the task a person wants to perform, the availability of facilities in the environment, or individual preferences based on personal attitudes and lifestyles

For the current project we decided to use a combination of qualitativeinterpretative and quantitative-statistical methods following the concept of “across-method” triangulation [12] in order to minimise the specific limitations each applied method is restricted to.

Approaching a Typology of Pedestrian Behaviour

The current approach contains of two phases of empirical data collection including observation and inquiry techniques. Figure 1 shows the design of the current study, including two phases of empirical data collection followed by hypothesis testing and the development of a model of pedestrian mobility types.

The first heuristic phase is to hypothesise and identify basic types of pedestrian route choice behaviour and movement patterns based on motion data collected by technology based tracking techniques. In the following deductive phase the initial typology is tested using a combination of localisation technologies (outdoor: GPS; indoor: Bluetooth) and detailed semi-standardised interviews. Results of both empirical phases are subsequently related to each other in order to identify specific spatial behavioural styles for the provisional categories.

a) Heuristic Phase: Tracking

The first phase of the empirical study has been conducted in Vienna in 2007. As investigation fields a shopping centre for the indoor observations and two major shopping streets for monitoring outdoor behaviour have been chosen in order to avoid the occurrence of behavioural differences caused by different context situations. In this phase the main empirical method used for data collection consisted of unobtrusive observation (also known as “shadowing” or “tracking”).

Observation procedures started with a random selection of unaccompanied walking pedestrians and following the individual as long as possible while mapping her or his path on a digital map, recording the specific point in time and the coordinates within the map for each point of the path. After completing an observation (e.g. when the subject left the observation area or the observer lost sight of the subject), additional notes concerning visible attributes of the subjects were taken (gender, age, appearance). Figure 2 shows an example of a typical trajectory including detected stops and velocities.

The analysis procedures performed on the collected data included qualitativeinterpretative methods and multivariate analysis methods (factor analysis, cluster analysis). The following section describes initial results of the analysis of indoor and outdoor tracking datasets.

b) Experimental Results

In total, 111 individuals with a balanced gender and age ratio have been observed (outdoor: 57; indoor: 54). The collected trajectories have been analysed according to the velocity computed between each marked point in the observed path; additionally locations and durations of stops within the trajectories have been detected. Subsequently, histograms of each trajectory have been compiled, showing the proportial amount of time an individual walked at a velocity within a specific speed interval. A comparison of all histograms indicates differences in spatiotemporal behaviour in indoor and outdoor observations. Figure 3 shows that subjects observed in the indoor environment spend significantly more time standing e.g. in front or inside a shop (speed interval 1: 0-0.1 m/s) and walk in general at lower speed than subjects observed in the outdoor area.

Figure 3: Histograms of all outdoor (a) and indoor (b) observations

Clustering analysis performed on both datasets form outdoor and indoor observations produced rather diverse results. While the analysis of the outdoor data showed eight discriminative clusters, the analysis of indoor observations resulted in only three homogeneous clusters of motion behaviour. This difference might be explained by the greater diversity of potential goals pedestrians might be pursuing in the outdoor environment: On a shopping street people passing by are not necessarily there for shopping (but e.g. on their way to or from work), whereas a person entering a shopping centre usually plans to buy something.

As an example the results of the indoor analysis are now explained in more detail, as the context situation (shopping) can be assumed to be more homogeneous than outdoors. The three clusters of motion behaviour can be interpreted as “swift shoppers”, “convenient shoppers”, and “passionate shoppers”:

Swift shoppers

This group consists of 60% male and 40% female participants who are relatively young compared to the other groups. They walk at comparably high speed (on average 1.2 m/s) and stop rarely and for a very short time (7 seconds on average, up to a maximum duration of 1 minute).

Convenient shoppers

Almost two thirds of this group are male shoppers (64%). The average age lies between 35 and 40 years and is higher than in the comparison groups. They stop more frequently (on average 1.4 times per observation) and hence show a lower average speed (0.6 m/s). Stops last approximately 2.5 minutes (up to 8 minutes).

Passionate shoppers

Two thirds of this group are females (67%), aged around 30 to 35 years. They stop quite frequently (about 3.6 times per observation) and for a comparatively long time (4.7 minutes on average, maximum 17 minutes). This results in a very low average speed of 0.2 m/s.

Conclusion

Preliminary results of the first empirical phase indicate that a number of homogenous behaviour patterns can be observed, especially in consistent context situations. Further investigations using a non-disguised form of observation combined with detailed interviews include and currently test basic findings of the first analyses.

Further empirical analyses of more data during the currently ongoing second empirical phase as well as a careful examination of the results in different context situations during the final stage of the study are expected to lead to a comprehensive interpretation of pedestrian spatio-temporal behaviour. This can on the one hand be used in future mobile navigation services to provide customised route suggestions and location based information, and on the other hand also serve as a basis for determining parameters for pedestrian simulation models.

Acknowledgements

This work is part of the “UCPNavi” project, a cooperation project between the Vienna University of Technology and arsenal research, Vienna. The project is supported by the Austrian Funds for Scientific Research (FWF). The author would like to thank M. Ray (arsenal research) for developing the shadowing tool and N. Brändle (arsenal research) for his help and advice concerning data analysis. The digital map used in Figure 2 has been provided by Stadt Wien – ViennaGIS (www.wien.gv.at/viennagis/).

References

[1] Millonig, A. and Gartner, G. (2007). On Defining Pedestrian Typologies for Customised Mobile Information Services. In Proceedings of the 4th International Symposium on Location Based Services & TeleCartography.

[2] Helbing, D., Molnár, P., Farkas, I.J., Bolay, K. (2001). Self-organizing pedestrian movement. Environment and Planning B: Planning and Design 2001 28, pp. 361-383.

[3] Thomas, C. (2003). Zu Fuß einkaufen. Project report.

[4] Millonig A., Schechtner K. (2007). Decision Loads and Route Qualities for Pedestrians – Key Requirements for the Design of Pedestrian Navigation Services. In: Waldau, N., Gattermann, P., Knoflacher, H., Schreckenberg, M. (eds.): Pedestrian and Evacuation Dynamics 2005. Springer Berlin Heidelberg, pp. 109-118.

[5] Golledge, R. G. (1995): Defining the criteria used in path selection, Technical Report UCTC No. 78, University of California Transportation Center.

[6] Hill, M. (1984). Stalking the Urban Pedestrian: A Comparison of Questionnaire and Tracking Methodologies for Behavioral Mapping in Large- Scale Environments. Environment and Behavior 16, pp. 539-550.

[7] Shoval, N., Isaacson, M. (2007). Tracking Tourists in the Digital Age. Annals of Tourism Research, 34 (1), pp. 141–159.

[8] Spek, S.C. van der (2007). Legible City – Walkable City – Liveable City: Observation of Walking Patterns in City Centres. Introductory paper, Urbanism On Track – Expert meeting on the application in urban design and planning of GPS-based and other tracking-based research, Delft, The Netherlands.

[9] Svetsuk, A. (2007). Experiments in urban mobility analysis in Rome using mobile phone data. Position paper, Urbanism On Track – Expert meeting on the application in urban design and planning of GPS-based and other trackingbased research, Delft, The Netherlands.

[10] Daamen, W., Hoogendoorn, S.P. (2003). Research on pedestrian traffic flows in the Netherlands, Proceedings Walk 21 IV. Portland, Oregon, United States: Walk 21 conference, pp. 101-117.

[11] O’Connor, A., Zerger, A., Itami, R. (2005). Geo-Temporal Tracking and Analysis of Tourist Movement. Mathematics and Computers in Simulation 69, pp. 135-150.

[12] Jakob, A. (2001). Möglichkeiten und Grenzen der Triangulation quantitativer und qualitativer Daten am Beispiel der (Re-) Konstruktion einer Typologie erwerbsbiographischer Sicherheitskonzepte. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (On-line Journal), 2(1).