Problem:
A pet supply company faced significant challenges in designing an automated distribution center. The goal was to ensure it could efficiently handle the flow of products like dog food and cat litter. The complexity of managing dynamic demand patterns, throughput uncertainty, capital expenditure, and system downtime posed a significant risk to the facility's performance and cost-efficiency.
Solution:
MOSIMTEC and DCS implemented a highly flexible distribution center simulation model using AnyLogic to replicate the automated pallet-building process. The warehouse modeling included autonomous mobile robots (AMRs), gantry cranes, pallet wrapping stations, and SKU slotting strategies.
Results:
- Validated order fulfillment as the distribution center successfully met daily order targets within a 20-hour window, even with shifting demand conditions.
- Reduced AMR requirements by 30%, lowering initial cost estimates without compromising throughput.
- Improved system design with identified and corrected SKU slotting imbalances, reducing congestion and increasing efficiency.
Introduction: automated distribution center project for a leading pet supply company
A leading pet supply company, operating more than 1,000 retail stores across all 50 U.S. states, needed to design a highly automated distribution center (DC). The goal was to efficiently manage the flow of products, such as dog food, cat litter, and other pet essentials, from manufacturers to stores.
To ensure that the facility could meet operational goals, the company enlisted Design Conveyor Systems (DCS) and MOSIMTEC to build a distribution center simulation model using AnyLogic—before construction even began.
It was planned that the DC would receive single-SKU manufacturer pallets and then prepare them into multi-SKU store-ready pallets using autonomous mobile robots (AMRs) and automated gantry cranes.
Problem: complexities in warehouse modeling for efficient design
The difficulty of distribution center automation and fulfillment system posed several key challenges:
- Throughput uncertainty: With a mix of products, varying demand profiles, and new automation tech, it was difficult to predict whether the system would meet daily order fulfillment targets.
- Dynamic system behavior: Frequent changes in SKUs, demand patterns, and order sizes made it difficult to anticipate how the system would perform over time.
- Capital expenditure risk: Substantial investment in AMRs and other automated components made understanding actual equipment needs critical to avoiding overdesign and excess cost.
- Design complexity: With product induction, palletization, and robotic task coordination all coinciding, the system required precise balance and synchronization.
- Downtime considerations: Breaks, mechanical delays, or downtime had to be factored into the analysis to ensure system resilience.
single-SKU pallets from manufacturers are transformed into multi-SKU pallets
Explore the recent release blog post, which highlights AnyLogic 8.9.4 features, including enhanced maintenance modeling.
Solution: distribution center simulation model with AnyLogic
MOSIMTEC developed a flexible, dynamic AnyLogic simulation model representing the automated multi-SKU pallet-building process. The simulation aimed to validate the system’s performance and guide investment decisions by uncovering hidden risks and optimization opportunities.
The simulation model was built to reflect the system's physical and logical realities. Warehouse modeling incorporated autonomous mobile robots (AMRs), gantry cranes, pallet wrapping stations, induction zones, and the rules and constraints governing their operation. The team carefully planned where to place each product, making sure items were in the right zones to build pallets easily and avoid delays.
One of the distribution center simulation model’s key strengths was its flexibility. Rather than hard-coding assumptions, MOSIMTEC enabled users to modify inputs such as SKU demand profiles, order sizes, zone configurations, robot speeds, and more. This adaptability was critical given the dynamic nature of retail demand and the expectation that the system would need to serve a growing and evolving network of stores.
A lightweight 3D visualization of the distribution center was also integrated into the simulation to support communication and analysis.
In addition, the simulation was tied to a Microsoft Excel interface, which allowed for easier scenario configuration and more accessible analysis for non-technical stakeholders. Through this tool, the client could run “what-if” scenarios and view key metrics, such as crane utilization, robot path congestion, zone dwell time, and order completion rate. This provided immediate and actionable feedback for design refinement.
Results: efficiency gains and cost savings with warehouse modeling
The distribution center simulation model delivered highly valuable insights and tangible business benefits:
- Validated order fulfillment: The system successfully fulfilled daily store orders within the 20-hour target, even under shifting demand conditions.
- 30% reduction in expenses on AMRs: Initial estimates on the number of AMRs required were cut by almost a third without compromising throughput.
- Improved SKU slotting strategy: The model revealed imbalances in SKU zone assignments, allowing teams to reallocate SKUs and reduce crane congestion.
- Uncovered non-obvious behavior: The simulation captured system dynamics that would have been missed in traditional spreadsheet models.
- Stress-tested downtime scenarios: The system maintained performance even when factoring in mechanical delays and crew breaks.
Now, the simulation model can be used as a long-term planning asset as the company brings new retail locations online and consumer purchasing patterns evolve. Warehouse simulation modeling enables ongoing optimization and scenario testing, ensuring the distribution center remains efficient and responsive beyond its initial launch.
The case study was presented by Geoffrey Skipton from MOSIMTEC and Justin Willbanks at the AnyLogic Conference 2024.
The slides are available as a PDF.
If you are interested in other MOSIMTEC case studies of leveraging AnyLogic for warehouse optimization, check out how they designed a material handling system with simulation for Walmart, the world’s largest retailer.
