Timber Harvesting & Computer Simulation Modeling - Daniel V. Goulet
With the growth in mechanized timber harvesting operations came the question of how best to use men and machines in a complex interactive environment to produce a timely flow of wood to the mill at a reasonable cost. The failure of analytical techniques, i.e., linear programming, queueing theory, etc., to answer this question has led to a great interest and investment in computer simulation to address the problem.
Initially, harvesting simulation modeling focused on machine simulators. However, machines in a harvesting operation do not function singly, but in conjunction with others both in sequence and in parallel. Today, there is much emphasis on the total system, and the component interactions that effect costs and production. The movement to the systems concept has produced modeling problems that are non-trivial: How does one account for and validly represent the sequencing of operations having great variability?, What elements significantly contribute to the performance variables of production and cost?, Can the modeling constructs used be mathematically verified?, and more. Most of these problems are still unsolved. In addition, the models have grown in size, so that for many, a medium to large computer system is necessary for their execution.
The presentation will address the modeling problems that have developed in the rapid growth of timber harvesting simulation. Also, some directions for fertile research will be suggested, which if completed, would both put the strategy on a sound mathematical foundation, and make it more accessible to the user.