Improving the decision-making process by considering supply uncertainty – a case study in the forest value chain
研究了如何利用数据质量评估结果构建概率分布，通过随机规划方法优化森林供应链中的运输和生产计划，降低不确定性带来的成本，案例显示运输成本从15.90美元/立方米降至15.32美元/立方米。
Planning decisions are generally subject to some level of uncertainty. In forestry, data describing the resources available have a major impact on operations performance and productivity. This paper aims to present a method to improve decision-making in the forest supply chain by taking supply uncertainty into account using the results of data quality assessments. The case study describes the operations planning process of a Canadian forest products company dealing with an uncertain volume of wood supply. Three approaches to constructing probability distributions based on data quality are tested. Each approach offers a different level of precision: (1) a frequency distribution of accuracy, (2) a normal distribution based on average accuracy, and (3) a normal distribution based on data quality classification. Using stochastic programming to plan transport and production shows that lower costs can be achieved with a general characterisation of the data accuracy. Not considering uncertainty when planning operations leads to a significant replanning transportation cost. Using classes of data quality to include uncertainty in operations planning contributes to reducing the transportation cost from $15.90/m3 down to $15.32/m3 representing 3.6%.