Real-Time Multivehicle Truckload Pickup and Delivery Problems

Abstract

In this paper we formally introduce a generic real-time multi-vehicle truckload pick-up and delivery problem. The problem includes the consideration of various costs associated with trucks' empty travel distances, jobs' delayed completion times, and job rejections. Although very simple, the problem captures most features of the operational problem of a real-world trucking fleet that dynamically moves truckloads between different sites according to customer requests that arrive continuously over time.We propose a mixed integer programming formulation for the off-line version of the problem. We then consider and compare five rolling horizon strategies for the real-time version. Two of the policies are based on a repeated reoptimization of various instances of the off-line problem, while the others use simpler local (heuristic) rules. One of the re-optimization strategies is new while the other strategies have recently been tested for similar real-time fleet management problems.The comparison of the policies is done under a general simulation framework. The analysis is systematic and consider varying traffic intensities, varying degrees of advance information, and varying degrees of flexibility for job rejection decisions. The new re-optimization policy is shown to systematically outperform the others under all these conditions.

Using simulation to evaluate how multi-agent transportation planners cope with truck breakdowns

“…The on-line optimization approach, extensively documented in [7]- [9], is based on the mixed-integer programming model of [4].…”

“…Then at each epoch the existing plan is either revised by heuristic methods or fully re-optimized via the use of a mathematical program. In this domain, [4] demonstrate the superiority of an exact mixed integer programming formulation of the PDPTW. They compare their reoptimization approaches to three heuristic approaches (a simple round robin assignment, an insertion heuristic, and a reordering approach).…”

“…Given this very different heuristic mechanism, we can no longer compare empty-travel and cost metrics recorded for each truck only at decision epochs as in [4]. We must now consider full traces of truck activity across the full time range of the instance in order to make a fair comparison of planning methodologies.…”

A decomposition approach for commodity pickup and delivery with time-windows under uncertainty

“…He also shows that distributed heuristics are comparable to centralized optimization methods in the case of dynamic pickup and delivery problems. Yang et al (2004) consider a real-time (dynamic) multi-vehicle pickup and delivery problem, where requests arrive in real time, and their pickup and delivery windows are known at arrival. They propose formulations for the offline and online contexts of the problem, and describe that the best policy is one that takes some future demand distribution into consideration.…”

A General Vehicle Routing Problem

“…Dynamic real-life problems often require rich models, in most of the literature on dynamic routing problems, however, some simplifying assumptions are made. For example, in the dynamic full-truckload PDP, which recently has received increasing attention, see Fleischmann et al (2004), Yang et al (2004), andPowell et al (2000), each vehicle can only carry one transportation request at a time and cannot load further shipments until all currently loaded shipments are unloaded. The only work known to the authors regarding rich VRP in a dynamic context is presented by Savelsbergh and Sol (1998).…”