Pedestrian Crossing Behaviour at Signalised Crossings
S Ahuja, X Hao, Mott MacDonald, UK; M Bell, Imperial College London, UK; S Phull, Department for Transport, UK
This paper aims to explain pedestrian behaviour around traffic signals and gives guidelines to promote pedestrian safety and design of traffic control systems. A signal control taking into account pedestrian route choice and behaviour is proposed
Town centres are often characterised by large pedestrian flows. Historically, traffic signal control has sought to optimise vehicle flows, sometimes with priority for public transport, while pedestrian flows have been fitted around the vehicle flows. A typical example of this is the staggered pedestrian crossing, which splits the pedestrian crossing movement into two movements allowing the green time lost to vehicles to be minimised at the expense of delay incurred by pedestrians who (if they comply) wait on a traffic island. This raises two issues:
1.Significant numbers of pedestrians fail to comply with the detour and delay involved in a staggered pedestrian crossing, leading to unsafe crossing behaviour.
2.Where the primary function of a junction is to allow pedestrians to cross, local authorities may well wish to shift priority from vehicles to pedestrians.
This paper aims to explain pedestrian behaviour around traffic signals, and gives guidelines to promote pedestrian safety and safer design of traffic control systems. The study also investigates pedestrian perceptions under different crossing situations and their route choice around traffic signals. Finally, this work feeds into the development of a traffic signal control mechanism which takes into account pedestrian route choice and their behaviour in determining optimum settings.
To identify junctions where priority should be shifted to pedestrians it is necessary to construct a pedestrian network, establish the origins and destinations of pedestrian flows within this network, and predict pedestrian route choice. The usage of particular pedestrian crossing facilities could then be estimated. This includes development of a conceptual model before optimising parameters. However, there is at present an inadequate understanding of pedestrian route choice behaviour, and in particular how attractive or otherwise different types of pedestrian crossing are perceived. This in turn depends on how different types of crossing are used, and in particular how compliant pedestrians are to the intended crossing routes and signals.
In previous studies, pedestrian crossing behaviours (route choice and compliance) were basically explained by observed pedestrian data on site. This study (sponsored by the UK Department for Transport, DfT) looks into pedestrian safety and non compliance under current road infrastructure and signal control settings.
Recently the DfT has promoted the installation of ?Puffin? pedestrian crossings which aim to decrease ambiguity for pedestrians and vehicles at conflicts and promote greater pedestrian priority. This research looks at how people respond to different types of signalised pedestrian crossings such as pelican, puffin and toucan which take varying degree of pedestrian priority into account.
The issue of pedestrian behaviour is of extra importance as non-compliance by pedestrians at ill-designed signals has significant implications for safety. In this study we have coupled our observations from CCTV camera recordings of sites in West Midlands and London (a total of 24 locations) with face to face interviews of pedestrians. In addition the pedestrians have been subjected to a stated preference survey aimed to understand pedestrians? crossing behaviours when some crossing conditions would change.
The responses from the SP survey were analysed to obtain their trade offs in pedestrian behaviour under binary choice situations. In the questionnaire, attributes (variables), which influence pedestrian crossing behaviour is introduced, including status of pedestrian signal, traffic volume, characteristics of pedestrians and infrastructure of pedestrian crossing. All these attributes are used in the modelling analysis to measure pedestrians? crossing behaviour (e.g., to what extent an attribute would influence pedestrians? crossing behaviour and route choices at and around signalised intersections). A multinomial logit (MNL) model has been developed to analyse respondents? discrete choice behaviours.
The analysis of data (to be completed by April 2008) will be expected to explain the following aspects: effects of junction design on compliance of crossing; importance of factors influencing compliance of crossing; pedestrian route choice decision at signals (presented by aggregated choice percentages); signalised junction design that takes into account compliant route choices. Preliminary results indicate that pedestrians are more compliant at traffic signals if they are accompanied by children or if their mobility is impaired or if they have heavy luggage. In most of the other cases pedestrians tend to seek gaps at traffic lights and will ignore traffic signal status.
This important research will lead to development of tools that can be used to design signal settings that account for pedestrian behaviour. In addition the methodology includes contribution of pedestrians to the delay benefit calculations in appraisal of traffic signals and guidance on how to model and treat pedestrians at traffic signals.
Furthermore, the research project will provide evidences and guidance with the DfT, UK, to improve pedestrian crossing conditions at signalised intersections.
Association for European Transport