Analysis Of Atis Effect On Mode And Route Choices

The objectives of this paper are to present models of travelers\u27 mode choice, drivers\u27 diversion from normal route, and drivers\u27 adherence to pre-trip advised route under different levels of information using data from the same experiment. A travel simulator was used as a dynamic data collection tool. The simulator uses realistic network, two modes of travel, real historical volumes, and different weather conditions. In this study, sixty-five subjects have completed ten trial days each. Generalized Estimating Equations (GEEs) with binomial probit link function were used to account for correlation between choices made by the same subject in the three models. The modeling results show that the travel time and familiarity with the device that provides the information had significant effects in the three models. It is shown that developing in-vehicle information devices may lead to less transit usage in some cases, this flags a drawback of this technology. As the level of information increases (pre-trip to pre-trip and en-route; advice-free information to information with advice), drivers become more likely to divert from normal route. Number of traffic signals on the normal route and advised route affect diversion from normal route and compliance with pre-trip advised route