With increasing public concern about the environment, liveability and sustainability have become important issues in minimal fuel consumption estimation for transportation systems. Microscopic fuel planning and emission models use vehicle speed and acceleration as inputs and are suitable for predicting the amount of fuel at the link level. However, the lack of microscopic traffic data limits the application of these models. A method is provided for acquiring microscopic information from macroscopic traffic data. The main approach is to reconstruct the state and vehicle group trajectories with an Expectation Maximization algorithm with nice convergence properties and then to apply Dijkstra's algorithm in order to find a transport route with minimal fuel consumption. Validation of the method shows that the estimated fuel consumption reflects the real fuel amount and hence, the route with minimal fuel consumption determined by Dijkstra's algorithm is actually suitable for optimal transport planning. (C) 2016 Elsevier Ltd. All rights reserved.