Supply scenario simulation and optimisation in a global complex multi-echelon consumer healthcare supply chain

Haleon is a modern consumer healthcare company with a rich and diversified portfolio of products, distributed to millions of customers utilizing a variety of selling archetypes (including direct-to-retailer, distributor, or direct-to-pharmacy models). To achieve this, Haleon relies on a complex global multiechelon supply chain (taking into consideration raw materials supply, manufacturing, logistics and distribution). In such supply chain models, customers represent demand nodes (in which the demand levels are typically stochastic, having varied service levels and pricing agreements). The underlying supply chain scheme consists of thousands of raw material suppliers, tens of internal (worldwide) manufacturing plants, and hundreds of (local) contract manufacturing partners, several cross-docking and consolidation logistic hubs, distribution centers, warehouses, etc. Further complexities arise by considering additional regulations and product categorizations across different markets, as well as stability and perishability properties of particular products.

To address the challenge of orchestrating such a complex multiechelon supply chain, we will consider the interaction between supply planning and volume-to-value planning activities to meet the forecasted demands, under profitability constraints (enabling the achievement of a-priori-set financial objectives). By utilizing modern integrated business planning platforms (as a means to supply scenario modelling and simulation), we aim at developing associated optimization methods (and related sequential decision-making strategies) to obtain realistic and competitive plans that not only meet the financial and profitability targets, but also respect several other (often conflicting) objectives, such as capacity utilization, efficiency, reliability, and environmental impact, while at the same time respecting supply chain operational constraints and uncertainty.

In this project, we plan to develop novel multi-objective sequential decision-making algorithms on graphs (which are intended to represent the underlying multiechelon supply chain of Haleon with varying granularity levels). The development of the said algorithmic strategies must rely on the utilization of bespoke simulation software and Haleon’s real-world supply chain data to produce realistic and competitive strategies in a timely manner. The project will focus on the development of multistage and multi-objective optimization algorithms able to adapt to stochasticity, while respecting any associated operational constraints. At the same time, the construction of an underlying hierarchy of optimization models will be informed by Haleon’s real-world multiechelon supply chain.

The project will involve investigations at the forefront of multiechelon supply chain and multiobjective stochastic optimization research, and the development of associated Python software that will be subsequently tested on real-world supply chain data. A strong emphasis of this work will be to produce realistic, practical, and actionable outcomes.

The project is expected to involve close collaboration with colleagues from the Department of Mathematics of KCL (with primary supervision by Dr. Spyridon Pougkakiotis), and with partners from Haleon’s Data Science team (with external supervision by Dr. Gueorgui Mihaylov). Ultimately, the end-goal for the student is to compose an individual PhD thesis that meets the requirements for a degree as set out by the formal policies of KCL.

Key outcomes of this PhD project:

· Analysis of the Haleon supply practices, specific operational reality and constraints;

· Formulation of a realistic modelling framework for supply scenario simulation and quantitative assessment (optimal flow on a hierarchical graph model, with stochasticity determined by execution of actions/transitions affected by uncertainty);

· Formulation of a relevant, potentially multi-objective, optimization framework (definition and assessment of adequate objective functions to describe the relevant dynamics, implementation of the relevant operational, financial and contractual soft and hard constraints);

· Realization of a prototype in collaboration with the Haleon Data Science Team;

· Strategic assessment of the optimization AI enterprise capabilities necessary to enable full scale adoption of state-of-the-art solutions in the space of supply scenario simulation and optimization (infrastructure, computational resources, knowledge and skills).