ASTRID: Automatic SCOOT traffic information database

Hounsell, N. and McDonald, M. (1990) ASTRID: Automatic SCOOT traffic information database (Contractors Report, 235) Wokingham, GB. TRL 60pp.

Abstract

The primary purpose of the SCOOT (Split Cycle Optimisation Offset Technique) traffic responsive Urban Traffic Control (UTC) system is the optimisation of signal timings in a network to give minimum overall delay. However, in the process of optimisation, SCOOT provides a wealth of traffic information which is potentially useful for a variety of purposes. This study describes an investigation of the data available, its usefulness and potential applications and the development, production and demonstration of a micro-computer based database system for processing, storage and analysis of the data.

The traffic data available from SCOOT carried out by Southampton University for the Traffic Operations Division of the Transport and Road Research Laboratory (TRRL) can be obtained at a variety of levels of detail. Geographically this may vary from data for an individual link to aggregated data for a whole region, with potential time aggregation from one second to five minutes or more. More precise data requirements were determined from responses to a questionnaire sent to a sample of Local Authorities, Regional Offices and Policy Divisions of the Department of Transport. The database system operates on an IBM PC (or compatible) microcomputer linked to the SCOOT computer. The system is largely menu-driven and incorporates a number of user-defined options both at installation and for subsequent data processing and analysis.

Some key parameters that the user can define at installation are described and guidance on the selection of appropriate input values are included for most parameters, together with default values. Data processing involves receipt by the microcomputer of the 'raw' SCOOT messages, on-line stripping of this data to its minimum size (eg removal of repeated data descriptors) and data aggregation to its user-defined aggregation level. Regional data is scanned for missing data (ie due to faulty detectors on some links) and data is then corrected or discarded as appropriate.

The data is then: (i) added to previous 'profile' data for that time interval/location using exponential smoothing, allowing a current set of data values to be maintained. These files do not grow through time; (ii) aggregated further into 'trend files' giving average data values for peak, daytime off-peak, etc. Aggregation is only accepted if sufficient data has been received (as defined at installation). These files grow through time; (iii) discarded or stored off-line in binary format to minimise storage requirements.

All data being processed is compared with corresponding profile 'limits'; data lying within these limits is accepted, while outlying data, indicative of abnormal traffic conditions (eg due to an incident) is diverted to an 'outliers' file for separate analysis. A variety of data analysis options are included in the database system, based on dBASE and its supporting graphical software dGE. The analysis options provided include: tabulations of data

This record has no associated files available for download.

Contributors

Download statistics