Abstract

The truckload industry faces a serious problem of high driver shortage and turnover rate which is typically around 100\%. Among the major causes of this problem are extended on-the-road times where drivers handle several truckload pickup and deliveries successively; non-regular schedules and get-home rates; and low utilization of drivers dedicated time. These are by-and-large consequences of the driver-to-load dispatching method, which is based on point-to-point dispatching or direct shipment from origin-to-destination, commonly employed in the industry. In this dissertation, we consider an alternative dispatching method that necessitates careful design of an underlying network. In this scheme, a truckload on its way to destination visits multiple relay nodes and the driver and/or tractor are switched with a new one at these locations so that each driver stays close to their home domicile. In this respect, we evaluate the project in three different parts in which we address strategic (long-term), tactical (medium-term) and operational (short-term) decisions to design, and examine the proposed network.

In the first part of this research, we study a tactical design of a relay point network (RP-network) that may potentially help to alleviate this problem. Some specific design characteristics include the possibility of both direct and RP-network shipments, multi-route assignments, fixed relay costs, limited route circuity, and coverage required for relay points. We present a MILP model capturing these characteristics and a solution procedure based on strengthened Benders decomposition framework further enhanced by efficient heuristics. The solution approach is able to solve the large-scale problems, considering realistic inputs, in a reasonable time and helps us to examine the performance of the RP-network. Computational results demonstrate the performance of the algorithm.

In the second part, we investigate the strategic design of an RP-network under uncertainty in demand which can be more prominent for long-term planning. We use two-stage stochastic programming approach to model the RP-network designing problem in this situation. The setting of our problem of interest builds on the deterministic RP-network design problem addressed in the previous chapter. In this chapter, we extend this model by considering uncertainty in demands. The setting of our problem of interest builds on the deterministic RP-network design problem while we extend the model and the solution approach to address demand uncertainty. In order to address the computational difficulties specially occurring in this setting, we develop Progressive Hedging- Strengthened L-shaped algorithm. We show that the suggested solution method can effectively solve different classes of test instances and its effectiveness increases by increasing the size of the instances.

In the third part, we develop framework to study and test different truckload transportation concepts in an operational setting of our problem. This framework enables us to simulate day-to-day operations in TL transportation as closely as possible from the load dispatching and networking strategy perspectives. Using this simulation environment, we compare different network strategies including point-to-point (PtP), RP-network and hybrid PtP-RP-network and different dispatching approaches comprising dispatcher-based dispatching approach and a collaborative dispatching paradigm taking inputs from drivers as well. In this context, we develop and embed an optimization model for dispatching into our simulation environment.

Keywords: Relay network design, Benders decomposition, Two-stage stochastic programming, L-shaped method, Progressive hedging, Simulation-optimization

Degree Date

Fall 2019

Document Type

Dissertation

Degree Name

Ph.D.

Department

Engineering Management, Information, and Systems

Advisor

Halit Uster

Subject Area

Industrial/Manufacturing Engineering, Civil Engineering, Computer Science

Notes

The defense date: Friday, August 23, 2019

Number of Pages

133

Format

.pdf

Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

Share

COinS