在物流需求日益增加的同时，末端物流作为物流过程的最后环节面临着显著挑战。而机器人及自动化技术提供了一种新的末端物流解决方案，其在城市中的广泛应用还有助于缓解人口老龄化压力和提高城市治理和服务水平等。本研究分析了城市机器人的概念、特征，以及其面临的挑战，并聚焦即时配送类别，以大学校园作为应用场景，运用深度学习和仿真模拟方法，测度了即时配送需求并分析了末端物流无人化所带来的影响，最后提出相应的规划应对策略，以期为机器人应用下的城市空间规划、设计和管理提供科学参考。本研究开发了一种即时配送需求测度方法，结合YOLOv8、ByteTrack算法建立模型，实现对校园监控视频中外卖骑手的自动化识别和计数，并结合问卷调查得到合成的建筑物尺度、分钟级的订单数据作为模拟基础。在此基础上，研究提取了模拟中配送流程的关键参数以及多维度评价的绩效指标框架，并利用AnyLogic平台对末端物流无人化的多情景进行仿真模拟，全面分析了机器人在校园环境中的应用潜力及其综合影响。仿真结果显示，无人化物流系统在提高配送效率和降低能耗方面具有显著优势，尽管在服务水平方面表现略有不足。提高机器人或骑手速度、增加机器人规模和增加配送枢纽数量、优化配送枢纽空间布局等措施，可以显著提高配送效率，减少顾客等待时间，同时减少能源消耗。此外，为满足当前校园的即时配送需求，研究建议至少部署40台机器人，并提出了在校园内布局4-5个配送枢纽、每个枢纽15-20个机器人为较经济的规划方案，而13-15km/h是对机器人及骑手合理的管控配送速度。进一步，研究结合仿真模拟进行多方案比选，基于最佳方案进行了校园未来无人配送体系的规划，提出了一套面向校园环境的无人物流系统规划与设计指导原则，涉及配送分区、配送路线以及枢纽设计等关键方面。最后，研究结合文献和实践案例，构建机器人特征分析框架，并总结其在城市空间应用中面临的问题，从宏观管控规则、中观系统规划和微观空间设计三个层面提出相应的应对策略。
As the demand for logistics continues to grow, last mile delivery, as the final step in the logistics chain, faces significant challenges. Robotics and automation technologies provide a new solution for last mile logistics, which, when widely applied in urban environments, also help alleviate the aging pressure and enhance urban governance and service levels. This study analyzes the concept and characteristics of urban robots, the challenges they face. Focusing specifically on instant delivery within the context of university campuses, this study employs deep learning and simulation methods to measure instant delivery demand and analyzes the impacts of unmanned last mile logistics. Ultimately, corresponding planning response strategies are proposed, aiming to provide scientific references for urban spatial planning, design, and management under the application of robotics.This study develops a method for measuring the demand for instant delivery by integrating the YOLOv8 and ByteTrack algorithms to create a model that automates the identification and counting of delivery riders in campus surveillance videos. Combined with survey data, which provided synthesized building-scale and minute-level order information, this data forms the basis for simulation. The study then extracts key parameters of the delivery process and a multi-dimensional performance metrics framework. Using the AnyLogic platform, the study conducts simulations of various scenarios of unmanned last-mile logistics, comprehensively analyzing the potential applications and overall impacts of robots in the campus environment.The simulation results reveal that unmanned logistics systems offer significant advantages in enhancing delivery efficiency and reducing energy consumption, although its performance in terms of service level is slightly lacking. Measures such as increasing the speed of robots or riders, expanding the scale of robots, adding more delivery hubs, and optimizing the spatial layout of these hubs can significantly enhance delivery efficiency, reduce customer waiting times, and decrease energy consumption. Additionally, to meet the current campus's instant delivery demand, the study suggests deploying at least 40 robots and recommends establishing 4-5 delivery hubs across the campus, with each hub housing 15-20 robots as a cost-effective planning solution. A controlled delivery speed of 13-15 km/h is deemed appropriate for both robots and riders. Further, the study conducted multi-scenario comparisons through simulations and based on the best scenario, planned a future unmanned delivery system for campus. The study proposed a set of planning and design guidelines for an unmanned logistics system tailored to campus environments, addressing key aspects such as delivery zones, routes, and hub designs. Lastly, combining literature and practical cases, the study constructed a framework for analyzing robot characteristics and summarized the challenges they face in urban spaces, proposing corresponding strategies at macro-control rules, meso-system planning, and micro-spatial design levels.
