Meal delivery has become increasingly popular in past years and of great importance in past months during the COVID-19 pandemic. Sustaining such services depends on maintaining provider profitability and reduced cost to consumers while continuing to support autonomy and independence for customers, restaurants, and delivery drivers (here crowdsourced drivers). This paper investigates the possible enactment of curbside regulations in the U.S. that limit the number of drivers simultaneously waiting at restaurants to pick up meals for delivery on both public safety and delivery efficiency. Curbside regulations would aim to increase safety by enabling social distancing between delivery personnel at pickup locations and have a secondary benefit of improving local traffic flows, which are sometimes impeded in busier, urban locations. Curbside space limits are studied in relation to their impacts on consumer-related performance measures: freshness of the food on delivery and click-to-door time. This investigation is enabled through a proposed hybrid discrete-event and time-advanced simulation platform that replicates meal delivery service calls and pickup and delivery operations across a region built on data from a leading meal delivery company. Embedded within the simulation is an integer program that optimally assigns orders to drivers in a dynamically changing environment. Order assignments are constrained by imposed curbside capacity limits at the restaurants, and potential efficiencies and curbside violation reductions from bundling orders are assessed. Results of analyses from numerical experiments provide insights to state and local communities in designing curbside restrictions that reduce curbside crowding yet enable delivery companies to retain their profitability.

在过去的几年中，送餐变得越来越流行，在过去的几个月中，在COVID-19大流行期间，送餐变得非常重要。维持此类服务取决于维持提供商的盈利能力并降低消费者的成本，同时继续支持客户，饭店和送货驱动程序（此处为众包驱动程序）的自主性和独立性。本文研究了美国可能颁布的路边法规，该法规限制了在餐厅等车的乘客同时接送饭菜的数量，这既对公共安全又有利于提高效率。路边法规旨在通过使取货地点的送货人员之间的社会距离扩大来提高安全性，并具有改善当地交通流量的次要好处，有时在繁忙的城市地点会受到阻碍。研究了路边空间的限制与它们对与消费者有关的性能指标的影响有关：食物在交付时的新鲜度和上门点击时间。通过建议的混合离散事件和时间先进的仿真平台可以进行此调查，该平台可以根据领先的送餐公司提供的数据在整个区域内复制送餐服务呼叫以及取餐和送餐操作。模拟中嵌入的是一个整数程序，该程序可以在动态变化的环境中为驱动程序最佳地分配顺序。订单分配受餐厅施加的路边容量限制的约束，并评估了潜在效率和捆绑订单减少的路边违规行为。