D-ISN: TRACK 1: Supply Chain Analysis to Thwart Illegal Logging: Machine Learning-based Monitoring and Strategic Network Inspection

这项非法供应网络破坏行动（D-ISN）奖将通过调查破坏非法供应网络的战略，特别是非法采伐，为国家安全做出贡献。非法采伐是指在木材供应链的任何环节违反当地或国际法，包括在出口节点非法采伐或过度采伐，在中间节点误报原料或加工产品数量，以及在进口节点对木材的来源或品种进行错误分类等非法活动。尽管有一些打击非法采伐的现行法律，但这些活动在地方和全球范围内继续蓬勃发展。非法采伐被认为是破坏主要森林生态系统的主要原因，而森林生态系统在固碳和支持生物多样性方面发挥着重要作用。打击非法采伐的法律和机制的效力受到限制，因为人们对采伐木材的地点、数量和种类及其在全球木材产品贸易中的去向缺乏了解。为了应对这一系统性挑战，该项目将开发量化模型和方法来监测采伐速度，并进行战略检查，以发现全球供应链中的非法采伐或交易木材。该项目将展示如何利用现有的监测和检查技术：(1)量化全球供应链中非法木材的流动；(2)提高木材合法贸易的法律遵从性和执行力。该项目将培训研究生从事业务研究和环境可持续性的工作，并将开发有助于进一步推进美国可持续发展政策的方法。该项目开发了时空网络模型和分析工具，以识别和阻止非法木材在全球供应链中的流动。该项目将调查：(i)根据遥感数据、植物特征和木材物种生态，估计受保护和特许林区的非法木材采伐率；（ii）供应链网络分析，以评估非法木材的流动并估计非法贸易量；战略网络检查，以改进对欺诈行为的侦查，并在关键网络地点提高木材鉴定的法医能力。该项目将建立在机器学习、遥感、网络优化和博弈论的进步基础上。该项目的数据集和模型将对多个社区有用：负责跟踪非法木材流动的机构，寻求推进自然资源供应链分析和认证的科学方法的研究人员，以及对使用机器学习评估人为变化对森林动态的影响感兴趣的生态学家。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This Disrupting Operations of Illicit Supply Networks (D-ISN) award will contribute to the Nation's security by investigating strategies for disrupting illicit supply networks, with specific focus on illegal logging. Illegal logging refers to breaking local or international laws at any point along the supply chain of timber, including illegal activities such as illegal harvesting or over-logging at export nodes, misreporting of raw or processed product volume at intermediate nodes, and misclassification of origin or species of timber at import nodes. Despite several existing laws to combat illegal logging, these activities continue to thrive at both local and global scale. Illegal logging has been implicated as a leading cause of the destruction of major forest ecosystems that play a significant role in carbon sequestration and in supporting biodiversity. The effectiveness of laws and mechanisms for combating illegal logging is limited by poor understanding of where, how much, and what type of timber is harvested and where it goes in the global trade of wood products. To address this systemic challenge, the project will develop quantitative models and methods to monitor the rates of harvest and conduct strategic inspection to detect illegally harvested or traded timber in the global supply chain. The project will demonstrate how current monitoring and inspection technology can be leveraged to: (1) quantify the flow of illegal timber in the global supply chain; and (2) improve the compliance with and enforcement of laws for legal trade of timber. The project will train graduate students to work at the interface of operations research and environment sustainability and will develop methods that can contribute to further advance the sustainable development policies in the United States. The project develops spatiotemporal network models and analysis tools to identify and thwart the flow of illegal timber throughout the global supply chain. The project will investigate: (i) estimation of illegal timber harvest rates in both protected and concession forest regions based on remote sensing data, botanical characteristics, and ecology of timber species; (ii) supply chain network analysis to evaluate the movement of illegal timber and estimate illegal trade volume; (iii) strategic network inspection to improve the detection of fraud and increase the forensic capacity for wood identification at critical network locations. The project will build on advances in machine learning, remote sensing, network optimization, and game theory. The datasets and models from this project will be useful to multiple communities: agencies responsible for tracking movement of illegal timber, researchers seeking to advance scientific methods for analysis and certification of natural resources supply chains, and ecologists interested in using machine learning to assess the impact of human-induced changes on forest dynamics.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.