In this study, we address a pressing yet underexplored problem in humanitarian logistics (HLs), which is enhancing the safety and security of humanitarian personnel in conflict zones, through the introduction of non-routineness of trips in vehicle routing models. This research is critically justified, as evidenced by the alarming statistics from 2020, where over 100 road ambushes were reported, affecting over 200 humanitarian workers through kidnappings, injuries, and fatalities. Moving beyond traditional approaches that primarily focus on increasing convoy sizes or incorporating ambush probabilities, our research poses a pivotal question: How can existing knowledge from security journals and manuals be effectively leveraged to address this problem? In response, we have developed a groundbreaking vehicle routing model that ensures variability in routes and travel times, significantly reducing the predictability of humanitarian convoys and enhancing their security. Our approach also integrates additional security measures, including avoiding night travel and implementing convoy camping strategies, to further mitigate risks. A central innovation of our study is the creation of a first-of-its-kind index, designed to quantify the routineness of trips – a critical yet previously overlooked metric in HLs. The efficacy of our model is demonstrated through a comprehensive case study in South Sudan, a region afflicted by ongoing civil unrest, underscoring the real-world applicability and urgency of our research. Additionally, we provide an approximate closed-form solution for the aid allocation subproblem, optimizing for both fairness and effectiveness. This research marks a significant leap forward in logistics optimization, especially within HLs in conflict zones, addressing a vital gap in the literature and offering innovative perspectives and practical solutions to complex security problems faced by humanitarian logisticians.

中文翻译：

---

非常规出行的人道主义车辆路径问题

在这项研究中，我们解决了人道主义物流（HL）中一个紧迫但尚未得到充分探索的问题，该问题通过在车辆路线模型中引入非常规行程来增强冲突地区人道主义人员的安全和安保。这项研究是非常合理的，2020 年的令人震惊的统计数据证明了这一点，据报道发生了 100 多起道路伏击事件，导致 200 多名人道主义工作者被绑架、受伤和死亡。我们的研究超越了主要关注增加车队规模或纳入伏击概率的传统方法，提出了一个关键问题：如何有效利用安全期刊和手册中的现有知识来解决这个问题？为此，我们开发了一种突破性的车辆路线模型，可确保路线和行驶时间的可变性，从而显着降低人道主义车队的可预测性并增强其安全性。我们的方法还整合了额外的安全措施，包括避免夜间旅行和实施车队露营策略，以进一步降低风险。我们研究的一个核心创新是创建了首个同类指数，旨在量化出行的常规性——这是 HL 中一个关键但之前被忽视的指标。我们的模型的有效性通过南苏丹的全面案例研究得到了证明，南苏丹是一个遭受持续内乱的地区，强调了我们研究的现实世界适用性和紧迫性。此外，我们为援助分配子问题提供了近似封闭式解决方案，优化了公平性和有效性。这项研究标志着物流优化方面的重大飞跃，特别是在冲突地区的高级别组织内，填补了文献中的一个重要空白，并为人道主义后勤人员面临的复杂安全问题提供了创新视角和实用解决方案。