Aligning parking policy with real-world driver behavior—through flexible sessions, smarter pricing, and optional reservations—can significantly improve curbside efficiency and reduce delivery friction. Urban planners and transport policymakers should consider these insights when modernizing curbside management to support sustainable and efficient last-mile delivery.
Urban logistics professionals know all too well the challenges of finding parking for last-mile deliveries. In dense city environments, curb space is scarce and highly contested—often leading to inefficiencies, congestion, and even illegal parking. Traditional “first-come, first-served” parking policies, typically with rigid 60-minute slots, fail to reflect the actual behavior and needs of delivery drivers. A recent study proposes a more innovative approach: using flexible parking sessions and reservation systems tailored to real delivery patterns.
Using real-world data from delivery operations in New York City and modeling techniques such as Markov Decision Processes and multinomial logit models, researchers evaluated how different parking policies affect curbside utilization and revenue generation. Under the current 60-minute static policy, only 27% of awarded parking time is used. In contrast, introducing flexible parking sessions—which better match drivers’ actual dwell times—increases utilization to an average of 70%.
This improvement is not just theoretical. By dynamically adjusting session lengths and prices, cities can reduce illegal parking while generating comparable or even higher revenues. Importantly, most of the revenue in the flexible session model comes from the time that is actually used—unlike the static model, where significant payment covers idle curb space.
The study also explored advanced reservation systems, allowing drivers to book loading zones up to 10 minutes in advance. Although reservations are sometimes dismissed as impractical, the analysis shows that they can increase average parking space utilization by 5% when combined with flexible session times. Moreover, holding time for reserved spaces often overlaps with active use under the flexible model—reducing the waste seen in static systems.
In terms of revenue, flexible parking with increasing block pricing and modest holding fees proved competitive. Even when driver choice (i.e., price sensitivity or reluctance to park far from delivery points) was taken into account, flexible models maintained strong performance. Although raising holding fees may slightly reduce willingness to park, the resulting revenue gains often outweighed the losses in usage.
The study acknowledges its limitations, including its focus on a single loading zone and the use of Northeast U.S. data. Still, the results provide strong support for broader implementation, particularly when informed by local insights. Extending the model to networks of loading zones could unlock even greater gains by minimizing driver detours and walking time.
Source: Amaya, J., & Reed, S. (2025). Space management policy for urban last-mile parking infrastructure: A demand-oriented approach. Transportation Research Part E: Logistics and Transportation Review, 200, 104185. https://doi.org/10.1016/j.tre.2025.104185