This paper presents a vehicle energy management system that uses information about upcoming topography and speed limits along the planned route to schedule the speed and the gear shifts of a heavy diesel truck. The proposed control scheme divides the predictive control problem into two layers that operate with different update frequencies and prediction horizons. The focus in the paper is on the top layer that plans the vehicle speed in a convex optimization problem leaving the gear decision to be optimized in the lower layer in a dynamic program. The paper describes how predictive information of the movement pattern of surrounding vehicles can be incorporated into the convex optimization of the vehicle speed by using a moving time window constraint.