Optimizing and Scheduling of a Pooled Log Transport System

Abstract

The log truck scheduling problem (LTSP) under capacity constraints and time window constraints which is an NP-hard problem, involves the design of a set of best possible routes for a set of trucks starting and ending the day at a hub serving a set of loggers and mills. The objective is to minimize the unloaded miles traveled by trucks by pooling all the trucks owned by the trucking companies and optimizing the routes of the trucks. The other constraints include the speed limit, shift duration of the drivers and mandatory return of trucks to the hub at the end of the day. A deterministic simulator is developed using a C program which emulates any set of routes and gives total unloaded miles as an output besides other performance metrics. A simulated annealing algorithm is

used to generate an initial feasible solution and thereby improve the solution using various search operators functioning in the neighborhood. A sensitivity analysis is made to reduce the number of trucks needed to meet the demand for the day. For a small problem, the algorithm produces very close results to the optimum solution. A large practical problem with 68 trucks, 22 loggers and 13 mills is solved to find the best set of routes using the developed algorithm. Finally an economic analysis is made to find the impact of the location of loggers and mills on the solution. Computational results show that the number of unloaded miles is reduced by 28% and the number of trucks needed to serve the loads is reduced by 17%. The results are further verified by feeding the set of routes as input to the simulator.