Investigation in Crowdshipping Enabled Urban and Last-mile Delivery Paradigms
thesisposted on 01.05.2020, 00:00 by Sudheer K Ballare
Urban logistics, especially the last mile delivery, is a major urban challenge due to the high density in cities and the use of diesel-powered freight vehicles, resulting in traffic congestion and other negative impacts such as air and noise pollution. With the rise of e-commerce and customer expectation of express delivery, the metropolitan areas are seeking innovative ways to better manage urban freight movement and create an efficient and environment friendly transportation system. Information and communication technology advancement is ushering in a new era of mobility, changing the way people and goods move. Traditionally, freight movement takes the so-called hub-and-spoke delivery model. While the hub-and-spoke model is proven efficient with the economies of scale, it is ill-suited for fulfilling the increasing demand for same-day and even one-hour delivery in urban areas with the inflexible structure and is restricted by the hub capacity. Using real-time information technology for speedy coordination, crowdshipping brings reduction in the transportation costs for the last mile. Crowdshipping provides a feasible alternative in the first and last mile deliveries, provides flexibility of delivery time windows and uses a variety of transport modes for the delivery, for e.g. personal automobiles, taxis, bicycles, cargo cycles, walking etc. In Chapter 3 of this dissertation, an empirical investigation is presented of an existing crowdsourcing delivery company with respect to the operational factors such as parcel size, delivery distance, demand frequency and distribution, the user characteristics including customer and driver profiles, and the pricing model. Both quantitative and qualitative analyses are performed to shed light on the market demand trending and growth opportunities in crowdsourcing deliveries. Another solution to reducing truck trips into busy urban centers is the use of Automated Parcel Station (APS). APS is an automated parcel collection (and sometimes dispatch as well) station located in public spaces. For the ease of access, these APS’ are located at shopping centers, transit stations, gas stations, etc. Parcel recipients travel on foot or by car to collect their parcels from an APS. In Chapter 4 of this dissertation, we propose, formulate, and evaluate a new and innovative delivery paradigm where the last-mile demand fulfilment is done through Microhubs with Crowdshipping (M+C). In this paradigm, an urban service area is divided into a number of smaller service zones (e.g., by zipcode). Within each zone, there is a microhub to temporarily store inbound and outbound parcels of small to medium size. These parcels are collected or distributed by automobile or bicycle crowdshippers between customers (shippers and end receivers) and the zonal microhub. Commercial trucks are dispatched periodically to visit the microhubs in the service area to transfer parcels to their respective destination microhubs. Though, an initiative involving microhubs and dedicated freight bikes has been field tested recently in the Citylab project in Europe for the first time, the performance of a microhubs and crowdshipping paradigm has not been analytically assessed before this dissertation. In Chapter 5 we show that the proposed M+C paradigm is a Many-to-Many Split Pickup and Delivery Problem (M-MSPDP) for truck routing between microhubs. We present a general formulation of M-MSPDP and a Maximum Split-Benefit with Tabu Search (MS-BTS) heuristic to solve the large-scale M-MSPDP. MS-BTS is evaluated with the exact solution methods and other existing heuristics. We further apply the MS-BTS heuristic to solve for two applications of the M-MSPDP: parcel pickup and delivery among parcel stations (i.e., M-MSPDP-FPD) and bike rebalancing in a bike-sharing system (i.e., M-MSPD-OC). The computation time of MS-BTS is considerably improved over the other methods while maintaining a comparable level of the solution quality. Lastly, Chapter 6 considers a futuristic delivery paradigm where all stages of the last-mile demand fulfillment are handled without any involvement of human factor, including for parcel loading/unloading, sorting and transportation between hubs and the customers. This Chapter presents a brief commentary on the impacts of such a proposed delivery paradigm.