Allocation flow patterns and cost-aware routing in packaging networks

Abstract

Background – Scania a major manufacturer of heavy commercial vehicles, operates a European returnable packaging network composed of four interconnected pools that supply transport items to suppliers across Europe. Internal allocation orders between these pools generate substantial transport activity that the company's current planning department handles through manual decisions made under limited visibility of the underlying cost structure. In a network where pool-to-pool allocation movements account for approximately 11 per cent of the analysed transport loads, and where contracted prices vary substantially across destinations and shipment sizes, ensuring that direct and allocation-based routing decisions are made on a cost-aware basis is important for both operational efficiency and environmental performance. Purpose – The thesis aims to develop a data-driven decision support concept that helps planners evaluate cost-aware transport routing options now of planning. The work has two interconnected aims. The first is to characterize the structural patterns of internal allocation flows between Scania's four packaging pools (Eschweiler, Angers, Järna and Oskarshamn) and to compare how pool-tosupplier transport costs differ across destinations and shipment sizes when historical and contracted pricing for standard, mega, and MSRT trailers are placed side by side are placed side by side. The second is to translate these insights into a prototype tool that demonstrates the analytical logic for a future production system to build on at the case company. Method – The study combines quantitative analysis of historical transport data with engagement with the contracted rate sheet and qualitative input gathered through recurring discussions with transport planners at the case company. The empirical foundation consists of approximately 35,000 transport records covering March 2025 to March 2026 together with the company's rate sheet defining contracted prices across fourteen load meter brackets for more than 1,000 poolto- destination combinations. Analysis was conducted at network, lane and shipment level. A spreadsheet-based prototype was developed iteratively to demonstrate how empirical findings can be operationalized at the level of the individual routing decision. Results and Discussion – The analysis reveals that allocation flows between the four pools are highly asymmetric. A single corridor between Angers and Eschweiler carries close to half of all internal. Pool roles remain stable across every month of the studied period, indicating that the patterns are structural rather than transient and they always have differed demands in all the studied period of time frame. The comparison between historical and contracted pricing shows that historical weighted averages exceed rate sheet full truck prices on 83 per cent of comparable lanes, with average premiums ranging from 44 per cent at Järna to 6 135 per cent at Eschweiler. A separate allocation-order dataset reveals a boomerang pattern in which bilateral movements occur on the same corridor within the same week, indicating that approximately 14 per cent of the analysed allocation transport cost may be linked to potentially avoidable opposite-direction movements under a conservative minimum-overlap assumption. The decision support concept integrates these findings into a single planner facing tool that displays both pricing perspectives and ranks the several possible routing options for each supplier need and surfaces the information that is currently less visible at the moment of planning. Recommendations for Scania – Based on the results of the analysis, Scania is recommended to develop the prototype into a production version built on a platform such as Power BI or internal Scania transport management tool (NILE) that connects directly to transport management and environmental reporting systems. A KPI capturing bilateral redundancy at the lane level should be introduced to make the boomerang pattern operationally visible. Cross-pool coordination should be embedded organizationally so that planners can see each other's activity rather than relying on individual initiative. Once route-level cost differences are made visible, Scania can use the results as input for future discussions with carriers and pool operators about pricing, responsibilities and incentive structures.