ROMI-3.1 Least-Cost Lumber Grade Mix Solver Using Open Source Statistical Software

Abstract

The least-cost lumber grade mix solution has been a topic of interest to both industry and academia for many years due to its potential to help wood processing operations reduce costs. A least-cost lumber grade mix solver is a rough mill decision support system that describes the lumber grade or grade mix needed to minimize raw material or total production cost (raw materials plus processing cost). Because raw material costs in typical rough mills comprise 40 to 70 percent of total rough mill manufacturing expenses, the least-cost lumber grade mix problem, as it is referred to, is important. An existing second-order polynomial least-cost lumber grade mix model integrated into the US Department of Agriculture (USDA) Forest Service's rough mill simulator, ROMI-3.0, which uses SAS 8.2 for statistical calculations, was used for the research described in this article. For this existing model, the USDA Forest Service purchased a SAS server license to allow free use of the software to least-cost lumber grade mix users via the Internet. Several issues around this rather involved setup necessitated the search for an alternative, local solution for the statistical computations. The open source statistical package R 2.7.2 was tested to see if it is an equivalent replacement for SAS 8.2. Comparisons of the SAS-based and a newly developed R-based least-cost lumber grade mix solver indicate no statistically significant difference between the two decision support systems. Therefore, the new R-based least-cost lumber grade mix solver was incorporated into ROMI-3.0. Thus, rough mill operators now have a new version of ROMI-3.0 with the integrated least-cost lumber grade mix solver at their disposal that does not require their computers to communicate with an outside server.

Dynamic safety stock in co-production demand-driven wood remanufacturing mills: A case study

“…Moreover, Nicholls and Roos (2006) and Bumgardner and Bowe (2002) studied species selection in the secondary wood manufacturing. Much research has been conducted to determine yields of least-cost grade mixes through software products which have been developed over time (Harding and Steele, 1997;Buck et al, 2010).…”