Cambridge Mobile Urban Sensing


CamMobSens is the Cambridge University pollution monitoring initiative consisting of groups from the Chemistry Department, Computing Laboratory and Cambridge eScience Centre. It was part of the MESSAGE project, a collaboration between Cambridge University, Imperial College London, Leeds University, Newcastle University and Southampton University. MESSAGE finished on the 30th September 2009, though CamMobSens conducted a large scale deployment, lasting three months, in the greater Cambridge area in the Spring/Summer of 2010. Work has now started on a NERC funded project to deploy an improved version of these devices, incorporating a novel particulates/aerosol sensor, at ~60 locations around Heathrow airport.

We use a variety of sensors, from small hand-held units which are carried by pedestrians, to slightly larger units which are fixed to lamposts. As an example of the data gathered, pollution levels are proxied by the height of the plots in the figures below.

carbon monoxide (CO)
nitrogen monoxide (NO)

Team members:

Prof. Jean Bacon, Department of Computer Science
Dr. Alastair Beresford, Department of Computer Science         
Dr. Mark Calleja, University Computing Service
Mark Hayes, Cambridge eScience Centre
Prof. Rod Jones, Department of Chemistry
Prof. Peter Landshoff, Department of Applied Mathematics
Dr. Iq Mead, Department of Chemistry
Michael Simmons, Centre for Scientific Computing

Here's a short, speeded up movie (1.9MB, MOV format) of someone with a nitrogen dioxide (NO2) sensor walking around Cambridge. The data is transmitted from the sensors via Bluetooth to Nokia mobile phones, for which we write Python applications under Symbian, and then via GPRS to a PostgreSQL database. The data can then be extracted immediately online and displayed either as a time series or in Google Earth (as in the pictures and movie above). Newer sensor units even have onboard GPRS functionality (dispensing with the need for a phone) via Telit's GM862 product family, which are also programmed in Python. Thanks deserve to go to our partners Nokia for providing the mobile phones and to O2 for providing GPRS access.

A transcript and podcast of a BBC radio programme (The Naked Scientists) that covered this work can be found here, whereas a podcast for Radio 4's Click On programme can be found here (starts at about 7.5 minutes into the latter podcast). It's also been covered (somewhat inaccurately) in the British newspaper The Guardian.

For general project information contact Mark Hayes (
For all computing/telecommunications information contact Dr Mark Calleja
For all chemistry/sensor information contact Prof Rod Jones