Optimization-based Learning for Dynamic Load Planning in Trucking Service Networks

Ritesh Ojha,Wenbo Chen,Hanyu Zhang,Reem Khir,Alan Erera,Pascal Van Hentenryck

2024-04-28

Abstract:The load planning problem is a critical challenge in service network design for parcel carriers: it decides how many trailers to assign for dispatch over time between pairs of terminals. Another key challenge is to determine a flow plan, which specifies how parcel volumes are assigned to planned loads. This paper considers the Outbound Load Planning Problem (OLPP) that considers flow and load planning challenges jointly in order to adjust loads and flows as the demand forecast changes over time before the day of operations in a terminal. The paper aims at developing a decision-support tool to inform planners making these decisions at terminals across the network. The paper formulates the OLPP as a mixed-integer programming model and shows that it admits a large number of symmetries in a network where each commodity can be routed through primary and alternate terminals. As a result, an optimization solver may return fundamentally different solutions to closely related problems, confusing planners and reducing trust in optimization. To remedy this limitation, this paper proposes a lexicographical optimization approach that eliminates those symmetries by generating optimal solutions staying close to a reference plan. Moreover, this paper designs an optimization proxy that addresses the computational challenges of the optimization model. The optimization proxy combines a machine-learning model and a repair procedure to find near-optimal solutions that satisfy real-time constraints imposed by planners in the loop. An extensive computational study on industrial instances shows that the optimization proxy is orders of magnitude faster for generating solutions that are consistent with each other. The proposed approach also demonstrates the benefits of the OLPP for load consolidation and the significant savings obtained from combining machine learning and optimization.

Artificial Intelligence,Machine Learning,Systems and Control

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to optimize the Outbound Load Planning Problem (OLPP) in the parcel delivery service network. Specifically, the paper focuses on how to adjust the loading and flow plans in response to changes in demand forecasts over time while ensuring that the service network can efficiently transport parcels from their origin to their destination. The key challenges of this problem are:

1. **Determining the quantity and type of loads**: It is necessary to determine how many trailers are dispatched to other terminals at different times of the day and what their types are.
2. **Allocating parcel volumes**: It is necessary to determine how to allocate parcel volumes to the available routing options while adhering to trailer capacity limitations.

These two decisions together define a "loading plan". The paper points out that the traditional OLPP model has the problem of being sensitive to input data, that is, when there are minor changes in the volume of goods, fundamentally different solutions may be generated, which makes it difficult for the model to gain the trust of planners in practical applications. In addition, this inconsistency also increases the difficulty of implementation because factors such as driver availability and equipment location need to be considered.

To address these challenges, the paper proposes the following solutions:

- **Lexicographic Optimization Approach**: By generating optimal solutions close to the reference plan to eliminate symmetry, thus providing a stable and optimal loading plan close to the existing plan.
- **Optimization Proxy**: Combine machine - learning models and MIP - based repair procedures to quickly find approximate optimal solutions that meet real - time constraints.

Through these methods, the paper aims to develop a decision - support tool to help planners automatically adjust loading plans across the entire network, reduce the time required for manual review and finalization of plans, and improve operational performance. Experimental results show that the Optimization Proxy is approximately 10 times faster than commercial solvers while generating solutions that are highly consistent with the reference plan, demonstrating the effectiveness and practicality of this method.