Skip to main content

Artificial Intelligence Based Systems for Traffic Control


The overall objective is to develop a Knowledge Base System (KBS) for traffic control. The project will focus on the study of urban traffic control techniques (fixed and real-time).
The overall objective of the project is to delineatetechniques for the application of artificial intelligence (AI) to traffic control problems through the development of an expert system, in order to overcome many shortcomings of existing operator controlled systems. Traffic control practices were reviewed to give an informed base from which standard parameters and indicators for traffic control were defined. To cope with the real time demands of the problem, intelligent data processing and performance monitoring software was developed. The core of the expert system, the knowledge base, was developed to provide strategies for congestion and traffic control. Work on specifying standard interfaces between the expert system and traffic control systems was also undertaken. The expert system has been developed and tested using offline modelling techniques, finally resulting in the CLAIRE expert system prototype.
Aa industrial product was made available on 09/01/92

The research aimed at identifying the functions and components of a prototype expert system. Studies to identify differences in signal control parameter definition and urban traffic policy of European Cities were carried out. Identifying the differences was fundamental to ensuring that the system was compatible with, and would operate independently of, current urban traffic control systems. Through studies of congestion monitoring and measures of performance, methods to diagnose and evaluate congestion tress, were derived. The offline modelling capability of the system was used to define remedial control strategies for congestion and was based on signal optimization and route assignment.

The prototype expert system developed has the following functions:
it performs congestion management independently of the type of urban traffic control systemsimplemented;
through a flexible programming facility, it can meet the constraints of the various control philosophies exercise by different traffic authorities (ie, area policy, green wave, etc);
it has intelligent online monitoring capability to build up a historic picture of events;
it is capable of recognizing congestion patterns building up in a network;
it uses diagnostic methods online to interpret the different congestion problems and selects a remedial control strategy from a library of such plans;
relies on offline modelling to derive the library of remedial plans.

The system has been used to process congestion information from the demand responsive control system using a link with Leicestershire County Council's Traffic Management Computer. The preliminary assessment of the system is encouraging.
It is expected that the KBS will be compatible with most of existing UTC in Europe. The project will lead to proposals for demonstrations in several European cities.
Sub-objectives for the KBS are as follows:
- to review traffic control practices;
- to define standard parameters and indicators for traffic control;
- to develop intelligent data processing and performance monitoring software;
- to develop a knowledge base on congestion and traffic control strategies;
- to specify a standard for the interface between the KBS and UTC;
- to specify, design and develop to Export System prototype (database, test of inference engine, interfaces,...);
- to develop off-line modelling capability and traffic signal optimization for the command.
Main Deliverables:
Knowledge Based System Prototypes for UTC


Institut National de Recherche sur les Transports et leur Sécurité (INRETS)
109 Avenue Salvador Allende
69675 Bron

Participants (6)

Castle Rock Consultants Ltd
United Kingdom
Heathcoat Building Highfields Science Park
NG7 2QJ Nottingham
United Kingdom
Woodhouse Lane
LS2 9JT Leeds
University of Nottingham
United Kingdom
University Park
NG7 2RD Nottingham