In the transportation and distribution sector, the task of dispatching and controlling vehicles in real time is becoming increasingly complex. The task requires complex perceptual and cognitive abilities that are often applied under a high level of stress. At the same time, the task has never been so important for the economic viability and efficiency of many transportation companies, including freight transportation, courier services, emergency vehicle dispatching, transportation of handicapped persons, and many others. In this context, the development of an expert consultant for an advanced computer-aided dispatching system is essential. This paper proposes learning by examples as a means to uncover the human expertise in this domain. We investigate and compare two techniques based on linear programming and neural network models. Experimental results are reported on real data from a courier service company.
Financed by the National Centre for Research and Development under grant No. SP/I/1/77065/10 by the strategic scientific research and experimental development program:
SYNAT - “Interdisciplinary System for Interactive Scientific and Scientific-Technical Information”.