Introduction
The Less-Than-Truckload (LTL) sector, especially in Europe, faces challenges due to increased service requirements and rising shipment volumes. Efficiently remodeling cross-docking terminals and choosing optimal terminal shapes for logistics optimization are crucial.
This paper presents a tool designed to help small and medium-sized enterprises (SMEs) plan and optimize LTL terminals. The tool combines automatic model generation and generic modeling in LTL planning for all terminal shapes.
This approach enables efficient layout planning and resource allocation without the need for significant financial investments or specialized knowledge.
Simulation model
A generic simulation model for LTL terminal logistics optimization was developed previously. It was designed to be accessible without simulation proficiency and customizable to fit unique SME I-Shapes of LTL terminals.
In the table below, SME requirements needed for the model, such as input parameters, design criteria, and KPIs, are listed. The criteria are defined for all shapes of LTL terminals. In this approach, users can individually set and modify input parameters before model generation.
In this logistics optimization study based on previous work, the authors used a parametric approach for high-level user input, suitable for similarly structured LTL terminals. Generic modeling, combining discrete-event and agent-based simulation modeling, is applied for terminal processes.
The proposed concept involves user input with predefined building blocks and parameters, which are transferred to the automatic model generation component. This generates an executable simulation model with predefined process models and agents. Experiments are guided by the user.
The concept is divided into three layers for flexible usability, catering to different user knowledge profiles:
- Layer 3: No simulation knowledge required, but deep understanding of LTL terminals.
- Layer 2: Broad understanding of LTL terminals and simulation knowledge for creating and adapting building blocks.
- Layer 1: Deep simulation experience required for automatic model generation, with optional knowledge of LTL terminal processes.
This layered approach ensures a wide range of users can utilize the solution effectively.
The research simplifies generic process modeling to focus on automatic modeling of terminal shapes. It includes three main processes: unloading, handling, and loading.
The automatic model generation process involves creating predefined building blocks, rotating elements if necessary, and positioning terminal components based on user input. Connection paths and walls for free-moving vehicles are then generated. Finally, the elements are integrated into the modeled processes, such as pickup and storage areas and paths.
The concept is implemented using AnyLogic 8.8, which supports automatic model generation, predefined building blocks, and process modeling. AnyLogic integrates databases, provides GUIs as web apps, and combines discrete-event simulation with agent-based modeling. Java is used for automatic model generation, with agents and AnyLogic libraries handling building blocks and processes.
A pixel-based interface is created for data input, defining the terminal in three steps: setting yard capacity, defining terminal shape, and determining dock positions.
Results
The validation of the simulation model is conducted in two phases to ensure it accurately represents real-world processes:
- The first phase verifies the model’s infrastructural completeness based on a defined requirements catalog.
- The second phase involves developing an experimental plan to assess the model’s ability to produce realistic results.
The model successfully meets all design criteria, demonstrating its capability to simulate various LTL terminal shapes and accurately map material flow and forklift paths and achieve logistics optimization.
Validation shows that the tool’s results are comparable to those of manually built simulation models, making advanced planning techniques accessible and fostering competitiveness and sustainability in the LTL sector.
The simulation results compare the traveled distances of forklifts for different terminal shapes and sizes relative to the I-shape. The results confirm the hypothesis: the I-Shape is most efficient for terminals with fewer than 69 docks, the T-Shape for intermediate sizes up to 104 docks, and the X-Shape for larger terminals. Additional key performance indicators, such as forklift utilization, system load, and cycle times, were also determined using the simulation approach.