Space management policy for urban last-mile parking infrastructure: A demand-oriented approach

Abstract

Finding parking to make deliveries in urban areas is a challenge. Urban parking infrastructure is limited due to high premium land value and increasing demand for curb space. We propose that better space management of on-street loading zones may be achieved by designing policies that incorporate behavior patterns of last-mile delivery drivers. To assess space management of the parking infrastructure, we model loading zone utilization as a Markov Decision Process with the baseline being the current first-come first-serve policy with 60 min static parking sessions. The baseline results in only 27% utilization of awarded minutes on average. We introduce flexible parking sessions to account for the demand distribution of drivers and achieve the same levels of mean utilization as 60 min parking sessions as well as improve space management by utilizing 70% of awarded minutes on average. We extend the analysis to show that flexible parking sessions may achieve competitive revenues by utilizing an increasing block pricing scheme and holding fees for advanced reservations. Furthermore, we utilize a multinomial logit model to determine willingness of drivers to park at available loading zones. The results show that our proposed parking space management policies can reduce illegal parking while increasing utilization, confirming that parking infrastructure may be better managed when parking policies match behavior of drivers.