Datalab Logo

Pon Logistics – Dynamic Assembly

July 16, 2021


Pon Logistics is facing the same challenge everyday: how can all goods be stored and transported in an efficient manner, from the point of origin to the point of utilizationThis process requires detailed planning and execution and can be optimized via multiple routes. One of these routes regards managing the types of orders made by dealers and predicting the required transport containers per order. 

To aid this process, we designed an algorithm that calculates the best fit for every incoming order. The heuristic supports for all item weight and sizes and multiple container sizes. It can rotate the item in 6 positions in order to find the best fit. Besides that, the algorithm can deal with optional weight limits per container to match righteous working conditions. It returns the optimal container selection, based on the lowest container volume and the highest filling ratio per container. 

Why is dynamic assembly essential? Forecasting the correct container type per dealer per day optimizes the logistical process on 3 levels. It selects the best fit right away and, thus, saves time and energy repacking an order to a bigger or smaller container. By selecting the smallest fitting container, it reduces the amount of ‘air’ shipped per container and, subsequently, per truck. Eventually, it could reduce the amount of containers that are used per dealer per day, and therefore reduce the number of trucks that are driving in and out everyday. 

The results of our innovation project resulted in the following implementations. Pon Logistics will now pile all order types in one container for one dealer instead of splitting all order types into different containers. Furthermore, Pon Logistics will purchase a new container type for their container selection, based on the optimal filling ratio.

  • Ellen Mik

    Data Scientist
  • Kasper Meijjer

    Senior Data Scientist
  • Charlotte Koning

    Data Scientist