An artificial neural systems framework for delivery truck scheduling

The authors examine the feasibility of developing neural nets to solve specific problems in a general business environment. The specific application addressed here is that of scheduling delivery trucks from a regional distribution center. Artificial neural systems have proved very efficient at recognizing patterns. By considering current delivery routes as distribution patterns, neural nets can be constructed to recognize and select the pattern that best matches a distribution load. It is concluded that a neural net scheduling technique, while not an optimization tool, can result in improved solutions over time due to its learning capabilities