The aim of the project is to investigate and develop techniques and systems for automatic incident detection and traffic monitoring.
The main aim of the research was to investigate and develop techniques and systems for automatic incident detection and traffic monitoring.
The research focussed on:
strategies for urban and extraurban incident and congestion detection;
software for AID (automatic incident detection) using UTC (urban traffic control) sensors.
The following results have been obtained:
in the urban area, a program to detect incidents using information from UTC detectors has been written and the software installed in London;
in the extraurban area a study has been made of the factors affecting the functional requirements of a detection and monitoring system;
methods for the short to medium term prediction of traffic flow have been developed and tested on 3 principal roadways in France;
a study of the potential for detecting incidents using information from beacon based dynamic route guidance systems has been completed;
work on the use of a knowledge based system (KBS) has been completed
Methods currently being used for incident detection and traffic monitoring or potentially available, from a number of different sources:
- urban traffic control system detectors (urban only)
- highway/motorway traffic control sensors (extra urban only)
- route guidance systems
- image processing techniques
- in-vehicle equipment
- knowledge based systems
The project will examine the information available from the above sources and assess its usefulness for traffic monitoring and for the detection of congestion and incidents. Where appropriate, strategies for the detection of incidents will be developed. The strategies developed for congestion and incident detection and traffic monitoring will be compared taking into account their effectiveness in different network types/traffic levels, cost, etc.
Guidelines will be produced to show which systems or combination of systems should be used in particular circumstances.
Monitoring / incident detection systems in the urban and extra-urban areas will be studied separately. Recommendations on the value and feasibility of integrating urban and extra-urban systems and the most appropriate methods of monitoring at the interface between the urban/extra urban networks will be made.
Strategies for urban extra urban automatic incident detection, software for AID using UTC sensors.
GU24 0BL Woking