Developing data collection methods to inform the quantitative design of cycle infrastructure

Abstract

Increased share of urban travel by bicycle is widely desired as a cost-effective and environmentally-beneficial means of travel, and one which has the potential to reduce road congestion and improve health outcomes. Recent rapid cycling growth in cities such as London has served to highlight the lack of robust empirically-backed quantitative literature to inform the practitioner, and the consequential barrier to the delivery of enabling infrastructure of the scale required to meet that demand. Even simple measures vary by orders of magnitude in the literature and some depend on intuitively na¨ıve assumptions, so a simulation (based on the Social Force Model) was defined and implemented to test the key underpinning (non-interaction) assumption of the Highway Capacity Manual’s quantitative definition of cycle level of service. The simulations indicate that an assumption of non-interaction between cyclists results in an outcome intrinsically at odds with fundamental traffic flow theory. Both the literature and simulation process serve to highlight the lack of existing appropriate empirical data and behavioural understanding. Furthermore, collecting such data is difficult, expensive and not easily scalable using current methods. Consequently, a methodology for the collection of key cyclist parameters from generic video data was created, and can be applied to bespoke video surveys and existing CCTV capture, across a variety of modes, and at a fraction of the cost of human operators. In addition, a bicycle simulator is developed which can test cyclist behaviour in a replicable manner and in a range of circumstances. The design and construction process is detailed, and a proof-of-concept, validated against real data, is presented. Subject to some minor improvements identified, the simulator can now be used more widely for the collection of behavioural data. These methodologies provide new and practical capabilities for the collection and application of cyclist data, and a greater understanding of cycle behaviour.

More information

Identifiers

ORCID for Christopher John Osowski: orcid.org/0000-0002-1077-6983

ORCID for Benedict Waterson: orcid.org/0000-0001-9817-7119

Catalogue record

Export record