The energy consumption of electrical vehicles is not easy to estimate, as there are many possible variables, that might influence it.
It is especially difficult to account for features, that cannot be represented on single road segments, such as intersections and turn direction.
In this thesis we perform an analysis of the influence of some such cross-segment features on the energy consumption of electrical vehicles.
We introduce the concept of supersegments, which combines sequential road segments and road elements into a single unit, which contain more information than the sum of its parts.
Afterwards we propose a method to transform a road network in a way that better facilitates the use of supersegments, and use a machine learning model to evaluate the impact the inclusion of supersegments has on energy estimation accuracy.
We conclude that the inclusion of supersegments in a road network has a significant positive impact on energy estimation.