(61k) A Framework for Resilient Multi-Product Supply Chains: An Application to Healthcare | AIChE

(61k) A Framework for Resilient Multi-Product Supply Chains: An Application to Healthcare


Sarkis, M. - Presenter, Imperial College London
Papathanasiou, M., Imperial College London
Shah, N., Imperial College London
Supply chain management aims at identifying successful strategies for the timely delivery of products to end-customers, by coordinating procurement of raw materials, transformation of intermediates into finished products and distribution activities [1]. As business environments continuously change due to globalization, regulatory aspects, emerging information technology tools, competition, and mergers and acquisition or outsourcing contracts, decision-making processes become increasingly decentralized. This results in planning decisions taken rather asynchronously by the decision makers involved, based on the information available and aimed at fulfilling company-specific objectives. In this context, end-to-end supply chain responsiveness is hindered and the risk of shortages may increase.

In next generation healthcare, supply chains are becoming increasingly distributed. Individual contract manufacturing organizations typically provide manufacturing expertise and assets for larger companies to produce complex genetic engineering-based raw materials, intermediates, and the end-therapy. In this landscape, key challenges arise from product orders varying on a day-to-day basis, a requirement for patient-specific batches for personalized therapies and short product shelf-lives [2]. Further constraints may result from the inventory kept for raw materials, which are acquired months prior to manufacturing [3]. On the raw material supply end, contractors may serve a range of markets and schedule production campaigns in advance to fulfil mid-term orders. This hinders the potential response of the manufacturing schedules to sudden short-term increases in end-therapy demand, failing to supply raw material and causing backlogs in downstream supply chain nodes. Strategies to mitigate shortage risks may include manufacturing raw materials in-house, which come with additional production costs. In these circumstances, computer-aided tools can help assess trade-offs and alternatives and identify supply chain configurations and operational policies which mitigate risks, ensuring successful end-product delivery (Figure 1) [4].

Here we present a mixed-integer linear programming (MILP) framework for the optimization of multi-product supply chains in personalized and targeted healthcare. Given location-specific and time-varying product demands, operating expenditures, process rates, inventory policies for each supply chain node, the optimization determines suitable candidate supply chain structures, production, and distribution plans.

We demonstrate how the developed tool can be used to assess the impact of network configuration, raw material procurement and supply uncertainty on supply chain performance with respect to (i) cost and (ii) delivery success. This enables the identification of resilient manufacturing strategies for improved supply chain coordination and minimized risk of failure.


[1] Shah, N. (2005), Process industry supply chains: advances and challenges. Computers and Chemical Engineering, 29, 1225-1235.

[2] Papathanasiou, M.M.; Stamatis, C.; Lakelin, M.; Farid, S.; Titchener-Hooker, N.; Shah, N. (2020). Autologous CAR T-cell therapies supply chain: challenges and opportunities? Cancer Gene Therapy, 27, 799-809.

[3] Capra, E.; Gennari, A.; Loche, A.; Temps, C. (2022), Viral-vector therapies at scale: Todays’ challenges and future opportunities. McKinsey & Co.

[4] Sarkis, M.; Bernardi, A.; Shah, N.; Papathanasiou, M. (2021). Decision support tools for next-generation vaccines and advanced therapy medicinal products: present and future, Current Opinion in Chemical Engineering, 32, 100689.


Funding from the UK Engineering & Physical Sciences Research Council (EPSRC) for the Future Targeted Healthcare Manufacturing Hub hosted at University College London with UK university partners is gratefully acknowledged (Grant Reference: EP/P006485/1). Financial and in-kind support from the consortium of industrial users and sector organizations is also acknowledged.