Understanding How Big Data and Crowd Movements Will Shape the Cities of Tomorrow
Andrew Leeson, AECOM, Pablo Alvarez, AECOM, Samya Ghosh, AECOM
This paper highlights the increasing importance of planning for pedestrian movements in large urban realms, and the need to drive forward new research and development on Big Data applied to crowd movements.
Given the complexity of crowd movement behaviour there is surely no more important or challenging ‘Big Data’ to capture in the coming years. Crowd modelling has developed a long way in the last 10-15 years thanks to developments in both hardware and software. But has data capture and understanding of pedestrian movements kept pace? The industry relies largely on research undertaken over 40 years ago by Fruin, yet many social and work practices have changed since then as well as demographics. The needs of the industry have also broadened, with many authorities and developers now looking at much larger areas requiring pedestrian modelling – for example ‘gateway’ rail stations no longer looked at in isolation, but as being conjoined with surrounding public thoroughfares and open spaces linking to other transport hubs, retail/commercial centres, large employment or residential zones or even leisure facilities such as sports stadia. Leading city authorities are already asking for modelling techniques to be developed for whole city centres to be able to guide development planning over the next 20-30 years. Safeguarding walking space in big cities is going to have a major impact on all other transport forms as well as building development.
But how are these models to be developed? Accurate data capture and understanding of individual and crowd movement patterns is crucial for localised safety issues as well as wider spatial planning needs. We aim to summarise where the industry stands at present, what is coming in the near future and where we are lacking and need to focus resources. An emphasis will be given to the limitations in existing data collection techniques, the limitations they pose in forming robust model development inputs, the recent initiatives in developing sophisticated and innovative data collection techniques and the automated ways of manipulating and utilising those data to form robust model inputs or validation benchmarks.
The paper will draw out an outline on where the research needs to be targeted: data capture, software development, understanding of behaviour patterns, or other areas?
Association for European Transport