No abstract

No abstract

No abstract

Due to the rapid development of online retailers, there is a great demand for package express shipping services, which causes traffic congestion, resource consumption, and environmental pollution (e.g., carbon emission). However, there is still a large amount of under-utilized capacity in the public transportation systems during off-peak hours. In this paper, we investigate the same-day package distribution using crowdsourced public transportation systems (CPTSs). Specifically, given a number of packages and the timetable of available CPTSs trips, we optimize the schemes of delivering the packages using the underutilized capacity of the CPTS trips, without impacting the quality of passenger experience. To estimate the amount of under-utilized capacity of each trip across any two adjacent stations, we propose the passenger transit model based on the history data. To assign the under-utilized capacity of each trip to the package deliveries, we develop the minimum limitation delivery (MLD) method, which only utilizes the minimum amount of under-utilized capacity of the whole trip to deliver packages. However, the available capacity is not fully utilized at most stations by MLD. Therefore, we further propose the adaptive limitation delivery (ALD) method, which loads as many packages as possible, until the volume of loaded packages reaches the available capacity in theory. The experimental results and theoretical analysis show that both MLD and ALD could distribute packages efficiently. Moreover, given a set of packages, scheduling of ALD only consumes about 67% time compared to the scheduling of MLD, with a little higher risk of impacting passengers.INDEX TERMS Package distributions, crowdsourced, public transportation systems, quality of passenger experience.

The dock-less shared bike systems provide a convenient transportation mode for users to find, ride, or return a bike anywhere via GPS-based smartphone apps, with the bike position turmoil arises as side effects. To solve this problem, the geofence technology has been explored and then equipped in the ride-sharing service. However, the inadequate utilization and unreasonable distribution of the geographical resource impact the effectiveness of the geofence sites seriously. In this paper, we propose a collaborative geofence site selection (CGSS), which first pick up the hotspots based on a density-based and collaborationinspired method, and then allocate the geofence sites in the top-ranking hotspots. The CGSS aims to optimize the distribution location and the occupied area for each geofence in the city, so as to maximize the satisfactory degree of customers for both coverage ratio and capacity, with the total supplied geographical area for building geofence sites. The experimental results show that the CGSS method distributes geofence sites with a highly satisfactory degree and utilization rate.