Everyday Cycling in Urban Environments: Understanding Behaviors and Constraints in Space-Time - Computational Approaches for Urban Environments - page 183

lanes by improving the completeness of cycle network while connecting key hubs
and trip generators with the aim of improving everyday cycling (LTP3 2011 , p. 160).
The last but one part of this background section examines earlier research on
understanding cycling behaviors within the context of route choosing and trip shares
using street network. Very few published cycling studies in Britain implement
the revealed preference (RP) approach for understanding cyclists' route choice
preferences. This research contributes to fill this emerging gap in the UK while
acknowledging and identifying few published cyclists' route choice studies from
Ottawa, Guelph, and Toronto (Aultman-Hall 1996 ); Minneapolis (Harvey et al.
2008 ); Zurich (Menghini et al. 2009 ); Texas (Sener et al. 2009 ); San Francisco
(Hood et al. 2011 ); Montreal, Quebec (Larsen et al. 2011 ); Portland (Broach et al.
2012 ); Auckland (Ehrgott et al. 2012 ); and ongoing work in Denmark ( www.
bikeability.dk ) . Almost all of these studies have some form of stated preference
(SP) component as part of the research design, with the exception of Zurich where
only GPS secondary data without additional stated preferences of the sample was
used for the research. Moreover, Duncan and Mummery ( 2007 ) in comparing GIS
measures with data from GPS conclude that the use of GPS in active transport
research is encouraged, enabling further work to be undertaken especially in cycling.
These aforementioned studies have used a variety of techniques in the quest of
understanding cycling behaviors which some are not suitable for analyzing detailed
quantitative cycling data. For example, Larsen et al. ( 2011 ) study concludes that
the grid-cell method is not appropriate for detailed analysis of cyclists' actual
route choice preferences. Additionally, they emphasized the importance of cycling
infrastructure and the fact that methods assisting objective revelation of priority
areas are essential to provide the evidence needed as input to effective use of finite
resources allocated to the building and improvement of cycling infrastructure. The
proposed technique in this study is therefore to be considered as an addition to
existing basket of techniques for understanding route choice in cycling research.
The reader may refer to study by Prato ( 2009 ) reviewing alternative solutions
in determining preferences of various travelers with the aim of increasing route
heterogeneity but in the context of general route choice modeling.
This study also has potential to provide some empirical bases for modelers in
cycling research to reasonably validate models based on understanding cycling
behaviors. The ability of ABM to capture emergent phenomena providing a natural
description of a system and its flexibility are some of its strengths (Bonabeau 2002 ),
which could be of benefit in using it to understand cycling behavior either at the
city or route levels. The uses of such simulations are numerous: (1) to inquire
better understanding of some aspect of the real social world (Axelrod 1997a , b ),
(2) to enable prediction or forecasting (Gilbert and Troitzsch 1999 , p. 4-5), (3)
to fabricate new tools that substitute weakness in human capabilities (Gilbert and
Troitzsch 1999 , p. 5), (4) for training, (5) for entertainment, and (6) for potential
in facilitating discovery and formalization. Due to some of its numerous uses,
suggestive call has been placed for building of community of social scientists who
promote simulation as a field (Axelrod 2003 ). Examples of ABM applicable areas
have been categorized but in a business context, namely, (1) flows which comprise