Optimizing a Leading Beverage Company’s Last Mile Delivery Network: How a Supply Chain Consultancy Used Simio to Achieve $66 Million in Value

Introduction

In today’s competitive beverage industry, efficient distribution networks are critical to maintaining profitability and service excellence. When a major beverage bottler found that delivery operations constituted approximately 70% of their overall supply chain costs, they recognized a significant opportunity for optimization. The bottler approached a leading end-to-end supply chain consultancy to help tackle this challenge through innovative simulation techniques.

The bottler operated a substantial network comprising 18 distribution centers and multiple cross-stock facilities serving nearly 20,000 active customers daily. Despite having successfully upgraded order fulfillment in their distribution centers with automation and streamlined processes, their delivery operations remained a significant cost center. Additionally, the bottler was experiencing non-uniform growth across different markets due to expansion and additional merchandizing plans, further complicating the optimization challenge.

This case study explores how the supply chain consultancy leveraged Simio simulation software to develop a dynamic routing tool that optimized the bottler’s last-mile delivery network, resulting in $12.8 million in annual operating cost reductions and a net present value of $66 million over ten years.

Company Background: The Supply Chain Consultancy

The supply chain consultancy is an end-to-end supply chain advisory firm that serves as an extension of their clients’ teams. With over 250 professionals experienced across diverse industries—from groceries and retail to manufacturing and construction—the consultancy brings a holistic view of supply chain operations to every project.

The company defines end-to-end supply chain consultancy as covering the complete spectrum of operations: from procurement and manufacturing through to warehousing and logistics, including all related infrastructure. In simpler terms, they help companies optimize how they buy things, make things, store things, and get them where they need to go.

The consultancy began in Australia and has since expanded globally with offices in New Zealand, Asia, the UK, and North America. The company’s motto, “Invent tomorrow. Today,” reflects their commitment to innovation and tangible results. They combine industry expertise with advanced technologies like simulation software to test solutions and ensure they deliver measurable benefits.

Within their service portfolio, the consultancy’s simulation capability enhances their supply chain offerings by addressing complex challenges that traditional methods cannot solve effectively. Their simulation expertise spans network optimization, automation evaluation, last-mile delivery improvement, distribution center efficiency, inventory management, and manufacturing optimization.

The Challenge: Optimizing a Complex Delivery Network

The beverage bottler approached the supply chain consultancy with a significant business challenge: their delivery costs formed approximately 70% of their overall supply chain costs. This presented a substantial opportunity for optimization, but several factors made this a complex undertaking:

• Network Scale and Complexity: The bottler operated 18 distribution centers and numerous cross-stock facilities serving almost 20,000 active customers daily. Cross-stock facilities are locations where product flows in through one end and out through the other without being stored, adding another layer of complexity to the network.
• Uneven Growth Patterns: The bottler was experiencing non-uniform growth across different markets due to expansion and additional merchandizing plans. This meant that volume was not increasing evenly throughout the network, creating imbalances that needed to be addressed.
• Geographic Diversity: The network included both regional areas with longer travel distances due to less dense routes and metro zones with different routing profiles and constraints. This diversity required a multi-tiered approach to optimization.
• Automation Utilization: The bottler had already invested in automation and streamlined processes for order fulfillment in their distribution centers. However, the existing customer-DC mapping was not optimized to direct volume through these automated facilities, limiting the return on these investments.
• Multi-Mile Transportation: The project needed to account for both middle-mile movements (between DCs) and last-mile deliveries (to end customers), adding another dimension to the routing challenge.

Traditional approaches to transportation modeling often struggle with these complexities due to the numerous constraints involved, including delivery windows, travel speeds, traffic patterns, and asset capacities. Additionally, the question of direct versus multi-stop transportation further complicates the optimization process.

The bottler needed a solution that could find the optimal balance between lead time, fixed costs, and variable costs by evaluating different network configurations while maintaining service levels.
Solution Approach: Dynamic Simulation Modeling

The supply chain consultancy approached this challenge by developing a comprehensive simulation model using Simio software. The project followed a structured 12-week approach divided into three key phases:

Phase 1: Strategic Review (5 weeks)

The initial phase focused on understanding the bottler’s current delivery network through data collection and analysis of business rules such as operating profiles, routing logic, inventory policies, and other constraints. This information was used to build a baseline model that represented the current system.

The baseline model served two critical purposes:

• Verification: Ensuring the model operated as expected
• Validation: Confirming the model accurately represented the network’s performance

This step was crucial for building stakeholder confidence and ensuring buy-in from the client. Once the baseline model was validated and approved, the team could proceed with confidence that any proposed improvements would be based on an accurate representation of the current state.

Phase 2: Future State Optimization (5 weeks)

The second phase focused on developing an optimized future state route and DC configuration. The consultancy’s simulation engineers created a dynamic routing algorithm within Simio that automatically determined which routes would be assigned to which distribution centers based on active locations.

This algorithm was designed to handle the complexities of multi-stop transportation routing while respecting all necessary constraints. The dynamic nature of the model was a significant value-add, as it could ingest different data sets and show how the network would respond—a capability far beyond what could be achieved with static spreadsheet models.

Phase 3: Benefit Analysis (2 weeks)

The final phase concentrated on quantifying the operational, financial, and strategic benefits the bottler would realize. The team outlined optimal routes for each scenario, provided a clear understanding of required investments, developed a comprehensive business case, and offered strategic recommendations for implementation.

Implementation Details: The Dynamic Routing Algorithm

At the heart of the consultancy’s solution was a sophisticated dynamic craft allocation logic developed by their simulation engineers. This algorithm automatically determined which routes would be assigned to which distribution centers based on active locations. The process followed four main steps:

Step 1: Customer-DC Assignment

The model first determined which distribution center each customer should be assigned to:

• Each active customer was assigned to the closest DC
• Customers with limited DC options (e.g., those on the outskirts of the network) were prioritized
• Customers with fixed assignments (e.g., those on islands) were locked to their designated DC

Step 2: Capacity Verification

The algorithm then verified that each DC could handle its assigned volume:

• The model calculated how many pallets needed to be shipped from each DC
• If the volume exceeded a DC’s capacity, customers were reassigned to ensure no DC was overloaded
• Customers closer to cross-stock facilities were redirected there when appropriate

Several DC allocation rules governed this process:

• New facilities were assumed to have infinite capacity since they could be built to the required size
• Automated facilities were prioritized to maximize utilization of these investments
• DC capacity was based on current profiles and operating hours

Step 3: Route Creation

For each DC, the algorithm created delivery routes:

• Orders large enough to fill an entire truck were assigned to single-stop routes
• Multi-stop routes were built for remaining customers, starting with those closest to the DC
• The model ensured that all customers assigned to a DC were included in routes

Step 4: Route Optimization

The algorithm determined the optimal sequence for customer visits:

• For each route, the model identified the next closest customer
• Travel times were calculated using API data that incorporated traffic patterns
• The algorithm verified that total driving time didn’t exceed legal limits for drivers
• The model ensured all orders fit on the truck

Additional routing rules included:

• Partial pallets were rounded up to full pallets for space calculation
• The number of routes a DC could run was limited by its vehicle capacity
• Traffic patterns were incorporated to reflect real-world conditions

Rather than seeking mathematical optimization, which would be computationally expensive, the algorithm accepted routes within an acceptable tolerance. This approach provided achievable and realistic transportation plans that clients could implement with confidence.

The model incorporated real-world constraints such as driver fatigue limits, vehicle capacity, facility operating hours, and traffic patterns. By integrating these factors, the simulation created a realistic representation of the network that could be used to evaluate different scenarios.

Results and Business Impact

The supply chain consultancy ran 12 different scenarios to identify the optimal network configuration for the beverage bottler. Each scenario was evaluated based on fixed costs and variable costs for both distribution centers and deliveries, with the total operating expenses compared to previous configurations.

The simulation results demonstrated that significant savings could be achieved through strategic network reconfiguration, better utilization of automation, and consolidation of operations. Specifically, the analysis revealed that the bottler could achieve:

• $12.8 million in annual operating cost reductions
• $66 million net present value (NPV) over ten years

Based on the analysis, the consultancy recommended four key scenarios for implementation:

Scenario: Dynamic Daily Routing

This scenario maintained the same customer-DC and cross-stock locations but employed dynamic routing on a daily basis to generate more efficient routes. This approach provided immediate efficiency gains without requiring structural changes to the network.

Scenario: Facility Consolidation

This scenario explored the benefits of consolidating operations between two facilities (anonymized as “Stratford” and “Goodwill” in the presentation). By removing the inter-facility transfer leg and reducing product double-handling, this change streamlined processes and reduced costs.

Scenario: Strategic Facility Evaluation

This scenario analyzed whether the potential fixed cost savings from facility changes would outweigh increased travel distances and resulting delivery costs. While this scenario provided financial benefits, the team noted that the trade-off between flexibility and capacity needed careful consideration.

Scenario: Cross-Stock Transformation

This scenario proposed transforming a full distribution center (anonymized as “Spearman”) into a cross-stock site. This change would maintain customer convenience while directing order fulfillment to an automated facility (anonymized as “Goodwill”). While the immediate financial benefit was modest, it was expected to grow over time as the advantages of automation increased.

The simulation model provided valuable insights by enabling rapid testing of different scenarios and visualizing how the network would respond to changes. This approach allowed the team to identify the minimal network cost while maintaining service levels, even when considering facility closures if the savings outweighed additional delivery costs.

Advantages of Simulation Over Traditional Optimization

Throughout the project, the supply chain consultancy found that simulation offered several key advantages over traditional optimization approaches:

Solution Resilience

Simulation allowed the team to test how the model would perform under various conditions and uncertainties. Given the unpredictable nature of today’s business environment—affected by factors like pandemics, trade wars, and climate change—this resilience testing was crucial for developing a robust solution.

Transparency

Unlike the “black box” nature of many optimization tools, simulation provided full transparency into the model’s operations. This transparency built trust with stakeholders and allowed the team to identify exactly how and where improvements were being realized.

Safe Experimentation

Simulation created a safe environment for testing different ideas and hypotheses without risking disruption to actual operations. This “crystal ball” capability allowed the client to explore various scenarios and see their potential impacts before making any changes to their network.

Dynamic Inputs

While optimization software typically works best with static inputs and constraints, simulation could handle the dynamic nature of real-world operations. The model incorporated changing factors like traffic patterns, labor availability, and vehicle capacity to create more realistic scenarios.

Visual Understanding

The visual nature of simulation models made complex concepts more accessible to stakeholders who might not be supply chain experts. This visual representation helped build confidence in the model and facilitated better communication of results.

Data Gap Management

Many organizations have gaps in their data that require assumptions to fill. Simulation allowed the team to test these assumptions and understand their impact on results, building greater confidence in the model’s outputs.

Conclusion and Future Directions

The supply chain consultancy’s project for the beverage bottler demonstrates the power of simulation in optimizing complex last-mile delivery networks. By developing a dynamic routing algorithm within Simio, the team was able to model various network configurations and identify opportunities for significant cost savings while maintaining service levels.

Based on the simulation results, the consultancy’s key strategic recommendations included:

• Consolidating two facilities to streamline operations
• Converting two distribution centers into cross-stock sites
• Optimizing automation processes at a key facility
• Reevaluating the business delivery model

The project highlights how simulation technology can effectively balance service levels with cost efficiency in complex distribution networks. The dynamic nature of the model allowed for rapid scenario testing and provided insights that would have been difficult or impossible to obtain through traditional optimization methods.

Looking forward, the consultancy is working on developing a business interface that would give clients ongoing access to this virtual asset. This interface would allow them to run new data through the model at any time, essentially providing a “second opinion” on routes without having to change settings in their transport management system.

This approach to network optimization through simulation has broad applications across various industries facing similar last-mile delivery challenges. By combining industry expertise with advanced simulation capabilities, companies can identify significant cost savings while maintaining or improving service levels in their distribution networks.