Heuristics for a real-world mail delivery problem

Authors:
Elisabeth Gussmagg-Pfliegl;Fabien Tricoire;Karl F. Doerner;Richard F. Hartl;Stefan Irnich
Affiliations:
University of Vienna;University of Vienna;University of Vienna and Johannes Kepler University Linz;University of Vienna;Johannes Gutenberg University Mainz
Venue:
EvoApplications'11 Proceedings of the 2011 international conference on Applications of evolutionary computation - Volume Part II
Year:
2011

Citing 4
Cited 0

The vehicle routing problem

The vehicle routing problem
Arc routing in a node routing environment

Computers and Operations Research
Combination of Metaheuristic and Exact Algorithms for Solving Set Covering-Type Optimization Problems

INFORMS Journal on Computing
Survey: matheuristics for rich vehicle routing problems

HM'10 Proceedings of the 7th international conference on Hybrid metaheuristics

Quantified Score

Hi-index	0.00

Visualization

Abstract

We are solving a mail delivery problem by combining exact and heuristic methods. The problem is a tactical routing problem as routes for all postpersons have to be planned in advance for a period of several months. As for many other routing problems, the task is to construct a set of feasible routes serving each customer exactly once at minimum cost. Four different modes (car, moped, bicycle, and walking) are available, but not all customers are accessible by all modes. Thus, the problem is characterized by three interdependent decisions: the clustering of customers into districts, the choice of a mode for each district, and the routing of the postperson through its district. We present a two-phase solution approach that we have implemented and tested on real world instances. Results show that the approach can compete with solutions currently employed and is able to improve them by up to 9.5%.