Optimising Packaging for eCommerce Businesses

Packaging is necessary for the protection and transportation of products. However excessive packaging increases both material and transportation costs, and in turn has a negative impact on the environment. It is therefore important to minimise packaging as far as possible. In the United Kingdom the impetus towards reducing packaging has recently grown due the new Extended Producer Responsibility (EPR) regulations which require businesses to report data on how much packaging they use and pay associated waste management fees. This challenge is particularly important for eCommerce businesses which use large amounts of cardboard packaging to dispatch thousands of orders each day. For logistical reasons such as limited physical space, such businesses usually only have a limited number of boxes available that they can use to dispatch orders. This selection of boxes must be carefully chosen. On the one hand, the selection of boxes must contain boxes large enough to pack the vast majority of orders. On the other hand, the selection of boxes must contain smaller boxes so that smaller orders can be packed efficiently. What makes this problem particularly complex is that the range of possible orders is extremely large, consisting of combinations from a range of hundreds or possible thousands of different products. Previous research has resulted in a mathematical optimisation framework which is able to optimise the selection of boxes for dispatch. The framework takes as inputs a large set of order data and a large set of potential boxes. It analyses which boxes can be used to pack each order, and then selects the subset of boxes which would then optimise packing efficiency. This framework has been implemented as a software tool and been demonstrated to produce significant improvements over previously used packaging for several real world data sets. However, the tool, which is implemented as a suite of command line programs and application programming interface, requires a high level of technical expertise to set up and use, and a relatively high amount of computing power to run. The aim of this project is to therefore create a web-based graphical user interface (GUI) for this framework. The GUI will lower the technical barrier required to use it, and it being hosted in the cloud would enable it to be used from any computer with internet access, including low-power devices. This GUI will therefore open up the box optimisation tool to a larger range of users and empower them to experiment with it directly. For example, the tool could be used to investigate the effect of allowing different numbers of boxes to be used for dispatch, and the solutions can be customised according to which key performance indicators are most important. It will thus serve as an effective decision support tool for optimising packaging options. This will allow businesses to reduce to the costs associated with packaging including both direct costs from packaging materials, and the waste management fees associated with the EPR regulations. Ultimately, this work will benefit the environment by reducing the total of packaging being used.