Embedded Files
Pickup and drop-off (PUDO) facilities have long been important to support passenger transportation and goods delivery. Typical examples of passenger PUDO facilities are expected:
at inter-mobility stations or terminals where passengers switch between different transportation modes,
outside the stadiums where large-scale sports events or concerts are hosted, and
at schools where students are dropped off in the morning and picked up in the afternoon.
Meanwhile, the last mile of the freight delivery system has garnered unprecedented and ongoing attention since the boom in e-commerce and time-sensitive delivery services, leading to frequent pickups and drop-offs of small shipments near their end consumers. Passenger pickup and drop-off, as well as the loading and unloading of goods, are frequently observed at designated curb spaces or in parking lots. The recent increase in adoption of the above-mentioned transportation services, and the anticipated continuation of this trend with the introduction of self-driving vehicles, imply that PUDO facilities will become more widespread and substantial.
However, have you ever observed those processes of pickups and drop-offs closely? Take ride-hailing services for example. The transportation facility was not constructed to meet this increasing demand for pickup and drop-off services. In a busy urban context, congestion can be observed alongside those services. From the perspective of road management, every single stopped vehicle can delay the entire lane. There is limited curb space or curb length, so is it capable of handling all pickup and drop-off services?
For example, Portland International Airport has launched a pilot project starting in May 2019 to test a new late-binding process. In this process, riders join a queue and receive a code from the ride-hailing apps instead of knowing their driver immediately. When a rider becomes the first in line, he or she rides a car and shows the driver the code for trip information. When demand rapidly builds up in a confined area, road congestion worsens. The above late-binding process was proposed to address the excessive delay riders and drivers experience when trying to identify each other in a highly congested area. However, as the following picture (click the dropdown button to see it) from Portland suggests, this process alone may not be sufficient to address the congestion.
Thus, my research aims to design and evaluate the throughput capacity of PUDO facilities. Efficient computational tools are developed to provide insights into the design of facility layouts and operational policies. Different methods are introduced for different types of PUDO facilities, namely,
(1) the passenger PUDO facilities with a uniform spot size,
(2) the passenger PUDO facilities with regular and handicapped parking spots (as an example for multiple spot sizes), and
(3) a delivery bay (with flexible spot sizes).
My doctoral dissertation presents mathematical and computational models that optimize the long-term performance metrics for various types of passenger and freight transportation facilities. In summary, my work:
(i) evaluates the throughput capacity of PUDO facility layouts and operational policies, making provisions for long-term demand shifts and uncertainty,
(ii) formulates this design problem as Markov decision processes (MDPs), verifies any modeling assumptions, and characterizes the optimal design decisions,
(iii) conjectures practical, interpretable, and implementable approximate policies or near-optimal solutions, and
(iv) develops microscopic discrete-event simulation frameworks with realistic vehicle trajectories for validation.
Page updated
Google Sites
Report abuse