Response of drivers to variable message signs in dynamic lane assignment application using artificial neural networks and other traditional techniques

Dynamic lane assignment (DLA) is one-way of improving the efficiency of traffic operation at signalized intersections by dynamically changing the number of lanes assigned for a given turning movement depending on the instantaneous demand. For the successful implementation of DLA, the drivers should be aware of the lane assignment well before approaching the intersection through variable message signs (VMS). VMS is one of the widely used intelligent transportation systems (ITS) in urban areas, which can significantly support the implementation of DLA by providing drivers with real-time information on the existing lane group configuration while approaching a signalized intersection. Saudi Arabia has very diverse drivers' population with a large percentage of chauffeurs working for households, industries, as well as taxi drivers with a considerable variation in their educational background and driving habits. This paper investigates the factors affecting the comprehension of VMS when used in conjunction with DLA to identify drivers who need extra attention in the licensing stage if DLA is applied. Statistical analysis and artificial neural network (ANN) techniques were used to assess age, education, occupation, and driver's experience as possible predictors of VMS comprehension.