Advanced data-driven approaches to design and plan robust and sustainable network for the forest biorefinery value chains taking into account uncertainty

当今社会的主要挑战之一在于管理资源的可持续使用以及将其转化和分配为既具有低环境影响又具有高积极社会和经济影响的商品和服务。供应链网络的设计是复杂的，涉及战略和敏感的决策。在森林工业中，战略决策包括森林经营战略、道路建设、生产能力、技术投资、产品和市场开发、生产战略和库存定位。 加拿大林业寻求通过新的生物产品和新技术来改变其经济，以使其产品组合多样化，并进入高端产品的新市场。由于新市场、新的先进材料、相关技术、新的供应以及与现有价值链的整合的不确定性，这种转型增加了额外的风险。不确定性也是林业管理的一个主要特点，因为它涉及不确定的生物和自然过程、虫害、大片森林领土和长期规划范围内的决定。基于运筹学的决策支持系统（DSS）已被引入到森林价值链管理中，目前已被政府和工业界所使用。大多数基于确定性模型，不考虑不确定性。这一缺点限制了使用，降低了决策者对这些DSS的信任，以及应用DSS提供的决策的能力。 森林部门价值链使用大量的处理和商业信息系统数据。这些数据来自工业价值链中的许多来源以及外部传感器和系统，并且与不同程度的不确定性相关。这一信息对所有规划工作都至关重要，但其质量和不确定性程度不同，难以对所取得的成果进行评估。目前缺乏包括不确定性和风险建模在内的方法和工具，可用于回答具有挑战性的问题，以支持利益攸关方逐步将林业过渡到更具竞争力和可持续性。由于各种原因，以往的随机规划、鲁棒优化、数据驱动的自适应鲁棒优化等方法都没有用于处理不确定性和风险问题。在这项发现补助金中，我们的目标是开发先进的数据驱动方法，以设计和规划一个强大和可持续的森林生物炼制价值链网络，同时考虑到不确定性。该计划将带来新的解决方案，以帮助各利益相关者解决经济，环境和社会可持续价值链的设计和规划，同时考虑各种不确定性来源。

One of the key challenges in today's society resides in managing a sustainable usage of the resources as well as of their transformation and distribution into goods and services with both low environmental impact and high positive social and economic impacts. Design of the supply chain network is complex and deals with strategic and sensitive decisions. In the forest industry, strategic decisions include forest management strategy, roads construction, production capacity, technology investments, product and market development, production strategy and inventory location. The Canadian forest industry seeks to transform its economy through novel bioproducts and new technologies to diversify its products portfolio and entering into new markets with high end-value products. This transformation adds additional risks due to the uncertainty of the new markets, new advanced materials, associated technologies, new supply, and integration with existing value chains. Uncertainty is also a major feature in forestry management since it deals with uncertain biological and natural processes, infestations, large forest territories and decisions over long-term planning horizon. Decision supports systems (DSSs) based on operations research has been introduced to manage value chain in the forest and now are used by governments and industry. Most of are based on deterministic models and do not consider uncertainty. This drawback limits the usage and reduces the trust of decision makers on these DSSs and ability to apply the decisions provided by the DSSs. Forest sector value chains use huge volumes of processing and business information systems data. The data comes from many sources in the industrial value chain and external sensors and systems, and are associated with varying degrees of uncertainty. This information is critical for all planning but has different level of quality and uncertainty making it problematic to evaluate the results obtained. There is a lack of approaches and tools which include uncertainty and risk modeling that can be used to answer the challenging questions to support the stakeholders in the transition of the forest industry to more competitive and sustainable over time. In the past, stochastic programming, robust (RO), data-driven adaptive robust optimization (ARO) have not been used to deal with uncertainty and risk due to many reasons. In this Discovery grant, we aim to develop advanced data-driven approaches to design and plan a robust and sustainable network for forest biorefinery value chains taking into account uncertainty. This program will lead to new solutions to assist the various stakeholders address the design and planning of value chains that are economically, environmentally and socially sustainable while considering various sources of uncertainty.