This paper is a synthesis of the eight year experience of the ZELT team on the subject of on-site testing of urban traffic control strategies and systems. A review is made of the various problems raised by this kind of evaluation and of the original solutions developed to solve these problems: constitution of experimental plans, definition of a reference, definition of criteria of effectiveness and of the means to measure those criteria, elimination of side effects such as demand variation effects, practical and economical issues are the main topics addressed. (A)
Abstract