WestRock is one of the world’s largest paper and packaging companies with $15 billion in annual revenue and 42,000 employees in thirty countries. WestRock’s Recycling Division is responsible for procuring fiber from various vendors to supply its internal paper mills where fiber will be converted to large, brown paper rolls. In the project, we analyzed and improved WestRock’s fiber allocation process with the goal of reducing costs.
The goal of the project was to lower overall cost. There was an opportunity to reduce cost within the Recycling Division by addressing how WestRock makes fiber allocation decisions. Improvements can be made by considering material and transportation costs simultaneously when making allocation decisions.
Another opportunity was to reduce transportation costs strategically by analyzing transportation modes used on highly utilized lanes. The high transportation costs result from having to procure carriers to pick up excess, unforeseen loads from supply points. In such cases, the Logistics team has to secure non-dedicated carriers at rates much higher than market rates. When loads are lighter than forecasted, WestRock loses money on dedicated carriers that are scheduled to pick up loads. Therefore, there was an opportunity to do a financial analysis of transportation costs and provide cost-saving recommendations on strategic transportation plans.
I. The design solution for improving the weekly fiber allocation process is an optimization model. The model is a mixed-integer linear model, and it’s objective is to minimize overall cost when allocating weekly supply to demand points. The decision variable is the allocation decision of the number of truckloads to ship from a supplier to a mill.
The model’s constraints are as follows:
Demand must be met at each mill.
The amount sourced from each supply point cannot exceed the supply available.
The grade being supplied by a supply point must be a material grade that the mill can accept.
Requirements on the minimum quantity of material to be sourced by suppliers must be met.
The optimization model is run in the programming language Python using a package called PuLP. It has an integrated linear programming solver which, based on the specific model, picks the fastest algorithm to minimize run-time.
II. To provide recommendations for strategic transportation plans, another what-if analysis was conducted on changing only transportation modes. In this analysis, transportation shipments data from January 2018 to July 2019 (19 months) for 20 mills in the South and North regions were analyzed. An analysis was done to study whether there is a potential for savings in transportation costs by changing the shipment mode of highly utilized lanes. The change is from higher-cost transportation mode to a lower cost based on historical average cost data.
Methodology for the transportation cost analysis is as followed:
For each mill, select lanes with more than five weekly average shipments.
If both dedicated and spot carriers are utilized for the lane, analyze it to see if there's a potential for cost savings from changing transportation mode. If there's a potential for cost-savings, make appropriate recommendations (i.e., investigate increasing/decreasing dedicated carriers)
Approximate potential savings are calculated assuming 50 percent of the average number of shipments from the higher cost transportation mode is changed to the lower cost.
For lanes where only one mode of transportation is utilized, make a recommendation to consider another mode.
An optimization model tool with a GUI. The tool output optimal fiber allocation that minimizes overall costs while enforcing necessary constraints. WestRock is currently working on integrating the optimization tool into its fiber allocation process and is analyzing model recommendation to understand where the savings lie.
The second deliverable is an Excel Worksheet with analysis of transportation mode cost analysis and recommendations. WestRock plans to use the analysis as a conversation starter in interdepartmental discussions and effort for strategically reducing transportation costs.
Value and Impact
WestRock can save approximately $400,000 weekly in overall costs by following the optimization model’s allocation recommendations.
WestRock can further saves an approximate $3 Million in transportation costs from implementing transportation mode change recommendations.