UK scientists are working on a £3.5m, three-year study into creating huge networks of wireless sensors to monitor air pollution.
The Pervasive Mobile Environmental Sensor Grids (PMESG) project will use wireless sensors mounted on cars, street furniture and even people to obtain accurate data on pollution in towns and cities.
The universities of Cambridge, Southampton, Leeds, Newcastle and Imperial College London are participating in the project, which is backed by the Engineering and Physical Sciences Research Council, the Department for Transport and Transport for London.
The main objective of PMESG is to research how a dynamic sensor network capable of measuring gases across large areas would work, says principal investigator Professor Philip Blythe of Newcastle University.
‘We are planning to use cheap wireless devices with sensor payloads to measure pollutants such as nitrogen oxide and carbon dioxide to give us totally pervasive monitoring,’ he said.
‘We will be able to measure how pollution builds up in real-time and how it then disperses.’
That information could be used to trigger demand for management systems to control and mitigate the traffic problems in an area and reduce the pollution, according to Blythe.
‘You could even add an automatic pricing element based purely on the conditions at any given time,’ he said.
The project will also look at the implications of taking data from the sensors to a central point and then processing it.
The wireless devices can sense in real-time, making measurements every half a second and creating a lot of data, says Blythe.
‘Dealing with the unknown quantities of wireless data that will be generated is the largest challenge in this project,’ he said.
Back-end systems are definitely a key issue for the workability of any pervasive system, says Butler Group analyst Mark Blowers.
‘Data centre storage capacity is growing rapidly, but whether it can keep pace with this sort of thing is another story,’ he said.
‘The amount of data involved could be absolutely huge. The solution lies in having business intelligence-type systems to extract trends and relevant data.’
What do you think? Email us at feedback@computing.co.uk
Further Reading:
reader comments