Home Reports Automotive Engineering Content

Be a jammy dodger – new sat nav to beat traffic jams

next article

02.07.2008

Researchers are developing a new in-car navigation system which informs motorists about traffic jams ahead and advises the driver of the best route for their journey before they reach the congestion.

The ‘Congestion Avoidance Dynamic Routing Engine’ (CADRE) uses Artificial Intelligence (AI) to interpret live traffic information shared between vehicles fitted with a special GPS.

The project is part of a consortium, consisting of the University of Portsmouth, ComSine, Smartcom Software, the Transport Research Laboratory, ViaMichelin and Hampshire County Council.

CADRE can sense traffic slowing down and building up into jams and works by ‘monitoring’ other vehicles on the road, informing motorists 5 to 10 miles away of a situation as it’s happening and recommends steps to avoid it while they can.

The AI software is built around ‘fuzzy logic’ which mimics human reasoning. The capability comes from the University of Portsmouth’s Institute of Industrial Research (IIR) which specialises in using artificial intelligence techniques for industrial applications.

Dr David Brown, Head of the IIR, said: “The system interprets live data from current traffic conditions so the motorist receives up-to-the-minute advice and can make an informed choice. It’s designed to take the pain out of that agonising decision about whether to try an alternative route which could be equally congested.”

The system takes into account of how traffic speeds vary by day of the week and time of day and even on individual roads. It means that journey times are predicted more accurately and better routes are calculated that take account of the typical traffic conditions for the time of travel.

CADRE learns constantly from motorists’ incoming data and will adapt itself to long term and short term speed predictions and ever changing circumstances to continuously update and improve its knowledge.

“At present routing can be carried out for minimum time or distance, but this can easily be extended to other criteria such as minimum cost or minimum CO2 emissions,” said Dr Brown.

Over 2,100 journeys were made on routes around Hampshire including the M27, M3, A3 and the notorious M25 to provide data for analysis. Additional speed data came from the British Highways Agency data to form the background knowledge needed within the AI system.

Richard Walker is from the Transport Research Laboratory, formerly the research arm of the Department of Transport, who advised researchers on data sources and the best methods of testing. He said: “The transport system in the UK is one of the key drivers of the economy and with more and more cars on the roads; a system like CADRE would be a valuable tool in keeping traffic moving.”

Future plans for the system would extend it to ferries, trains and even planes allowing travellers to examine different departure times to estimate the best time and route to travel. CADRE could be in the shops in as little as 18 months.

The University of Portsmouth’s technical partners, ComSine and Smartcom Software, both specialise in developing software and systems for the mobile communications, navigation and geographic information markets.

The project was sponsored by the South East of England Development Agency (SEEDA) as part of its role in the Government’s Innovation Platform in Intelligent Transport Systems.

Lisa Egan | Source: alphagalileo
Further information: www.port.ac.uk

next article