Evaluating urban traffic congestion using OpenStreetMap and sparse public transport data

Andrei Bejan, University of Cambridge

Details about the talk:
Public transport location data recorded over long periods of time as
well as ``live'' feeds containing such data are increasingly
becoming available. Over the last decade, the bus fleet in the UK
public transport sector has become equipped with Real Time
Information (RTI) systems. For example, it was reported in the
recent UK Public Transport Technology Survey [Kn08],
based on data received from local authorities in England, Wales,
Scotland and Northern Ireland in 2008, that 56 Local Authorities
(compared with 37 in 2007) had operational RTI systems covering
approximately 185 towns and cities each with a population over
10,000.

Congestion in urban areas causes financial loss to business and increased use of energy compared with freeflowing traffic. Providing citizens with accurate information on traffic conditions can encourage journeys at times of low congestion and uptake of public transport. Installing the measurement infrastructure in a city to provide this information is expensive and potentially invades privacy. Increasingly, public transport vehicles are equipped with sensors to provide realtime arrival time estimates, but these data are sparse. Our work within the TIME project [TIME,Be10,Be11] shows how these data can be used to

(i) estimate journey times experienced by road users generally, and

(ii) evaluate congestion levels in urban areas and thus provide better information for road users and traffic managers.

In this talk I will describe the bus data and show how their specifics may be used in conjunction with OpenStreetMap data, in order to process and anayse the data, and reconstruct useful information on transport traffic flow dynamics. We have found that OpenStreetMap data and some of its features (not only the topological description of the road network, but its more detailed infrastructure and meta characteristics as well) have a great potential in analysing public transport data in a meaningful way. This, however, naturally relies heavily on the accuracy, coverage and level of detalisation of corresponding features of the map, that is on the state of the map. I will discuss these issues using real data in the context of the city of Cambridge, UK.

This talk is based on a joint work together with Jean Bacon, Alastair Beresford, David Evans, Adrian Friday, and Richard Gibbens within the TIME project [TIME]. We acknowledge the support of the UK EPSRC grant EP/C547632/1 and wish to thank Stagecoach, ACIS, and Cambridgeshire County Council for the bus location data used in this study.


[Kn08] Knoop, L., Eames, T. Public transport technology in the
United Kingdom: Annual survey 2008. RTIG Ltd, Tech. Rep., Feb. 2009.
Available online: http://www.rtig.org.uk/web/docs/PDF/Downloads/RTIG-PR010-D001-1.0-Survey-2008.pdf

[Be10] Bejan, A. et al. Statistical Modelling and Analysis of Sparse Bus Probe Data in Urban Areas.
Proceedings of the 13th IEEE Conference on Intelligent Transportation Systems, Madeira Island, Portugal, September 2010, pp. 1256-1263.

[Be11] Bejan, A., Gibbens, R. Evaluation of Velocity Fields via Sparse Bus Probe Data in Urban Areas.
(Submitted to the 14th IEEE Conference on Intelligent Transportation Systems, Washington DC, USA, October 5-7, 2011.)

[TIME] TIME (Transport Information Monitoring Environment) Project: http://www.cl.cam.ac.uk/research/time/

Details about the speaker:
Dr Andrei Bejan is currently a Postdoctoral Research Associate in Statistics in the TIME (Transport Information
Monitoring Environment) project at the University of Cambridge Computer Laboratory (www.cl.cam.ac.uk/research/time)
His interests within this project include: statistical modelling, data mining and analysis of large scale (archived and real-time) urban transport traffic data, exploring
how sensor networks and distributed systems can be used to improve transport traffic
management and control, and studying dynamics of velocity fields on urban road networks. The project is an umbrella framework for research, application
development and deployment of transport monitoring systems in the city of Cambridge.

Dr Bejan has a PhD in Statistics (Heriot-Watt, Edinburgh, 2010) and Operational Research (State University of Moldova, 2007). His research interests include modelling and analysis of priority queueing systems with switchover times as
well as inference and optimal experimental design for network and random graph interaction models (transportation
networks, plant epidemiology and disease spread models, social networks). Other interests include
spectral theory of sample covariance matrices and Tracy–Widom distributions. He has significant experience in applied statistical modelling with large scale data sets and use of spatial
and mapping data. He actively participates in an interdisciplinary research. He is an experienced and keen user of the mathematical packages MATLAB, R and S-Plus. He is a member of Edinburgh Mathematical Society, Isaac Newton Institute for Mathematical Sciences, European Mathematical Society, and IEEE.

Theme: research