I am not the expert in vehicle routing. But, I do learn a lesson from Prof. Sam Thangiah’s talk in FICT UTAR today.
The rough meaning is something like this:
Best solution appears in publication. But in real world, only the robust solution does matter.
For example, Dijkstra’s algorithm works perfectly in finding the shortest path from point A to point B. However, it seems not to be feasible when it is applied to calculating the pick-up points for buses in a city. Heuristic approach is much appreciated instead.