CAREER: Integrative Resource Optimization Framework for Large-scale Drone Delivery Systems

该学院早期职业发展计划（CAREER）拨款将通过促进安全，高效和公平地使用低空空域进行无人机交付操作，为国家繁荣和经济福利的发展做出贡献。无人机送货有可能将最后一英里的送货时间从几小时缩短到几分钟，并消除数十亿次不安全、污染和浪费的个人旅行。然而，安全性和可扩展性方面的挑战目前阻碍了大规模的商业部署。该奖项支持对如何管理各种资源以实现有效的城市规模无人机交付服务的基本理解的研究。 设计和运营方面的挑战包括最佳的站点和集合点位置、容量规划、车队管理以及在电池、交通和天气限制下的动态车辆路线。该研究将支持空中交通工具物流，车载智能和空中交通管理的整合。该教育计划旨在生成新的课程，培养新的STEM人才，并在航空，机器人和运筹学的交叉点上激励新的企业家。 该教育计划还将为本科生创造研究机会，特别是为STEM中代表性不足的少数民族创造机会。该项目将介绍一个通用的，基于优化的框架，用于模拟复杂的决策过程中的空中物流业务，以及新的算法，用于解决在这样一个框架中建模的问题。三个安全关键资源，包括地面站，电池和空域，将被全面研究和整体建模：地面站的位置和配置由一套电池电量预测模型提供信息，反过来，确定无人机的可行路径，从而确定空域走廊的最佳动态分配。该研究将开发（1）可分解的数学程序来解决鲁棒网络设计和连续定位问题，（2）将联合收割机进化计算与凸优化相结合的新算法来解决非线性组合问题，（3）基于马尔可夫决策过程的创新扩展的随机多智能体系统建模的新框架，以及（4）改进的飞行中电池电源管理算法。该研究方法将理论发展与实际实施相结合，并在很大程度上受到商业应用潜力的影响。该奖项反映了NSF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 Faculty Early Career Development Program (CAREER) grant will contribute to the advancement of national prosperity and economic welfare by promoting safe, efficient and equitable use of low-altitude airspace for drone delivery operations. Drone delivery has the potential to reduce last-mile delivery time from hours to minutes and eliminate billions of individual trips that are otherwise unsafe, polluting and wasteful. However, safety and scalability challenges exist that currently impede commercial deployment at scale. This award supports research toward a fundamental understanding of how diverse resources can be managed to enable efficient city-scale drone delivery services. The design and operational challenges include optimal location of depots and rally points, capacity planning, fleet management, and dynamic vehicle routing under battery, traffic, and weather constraints. The research will support integration of aerial vehicle logistics, vehicle onboard intelligence, and air traffic management. The education program aims to generate new curricula, train new STEM talent and inspire new entrepreneurs in a new field at the intersection of aviation, robotics and operations research. The education program will also create undergraduate research opportunities, particularly for under-represented minorities in STEM. This project will introduce a general, optimization-based framework for modeling complex decision-making processes in aerial logistics operations, as well as new algorithms for solving problems modeled in such a framework. Three safety-critical resources, including ground stations, batteries, and the airspace, will be studied comprehensively and modeled holistically: the location and configuration of ground stations are informed by a suite of battery power prediction models, and in turn, determine the feasible paths of drones and hence the optimal dynamic allocation of airspace corridors. The research will develop (1) decomposable mathematical programs to tackle robust network design and continuous location problems, (2) new algorithms that combine evolutionary computing with convex optimization for solving nonlinear combinatorial problems, (3) a new framework for modeling stochastic multi-agent systems based on innovative extensions of Markov decision processes, and (4) improved algorithms for in-flight battery power management. The research approach integrates theoretical developments with practical implementation and is heavily informed by commercial application potential. Software apps will be developed and flight trials will be conducted to deliver practical relevance and impact.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.