(604h) Multi-Site Supply Planning for Drug Products Under Uncertainty | AIChE

(604h) Multi-Site Supply Planning for Drug Products Under Uncertainty


Zavala, V. M. - Presenter, University of Wisconsin-Madison
In the pharmaceutical industry, the goal of a supply planner is to make efficient capacity allocation decisions that ensure uninterrupted supply of drug products to patients and to maintain product inventory levels close to the target stock. This task can be challenging due to the limited availability of manufacturing assets, uncertainties in product demand, fluctuations in production yields, and unplanned site downtimes. It is not uncommon to observe uneven distribution of product inventories with some products carrying excess inventories, while other products may be close to a stockout. Maintaining high stock levels can have economic repercussions due to the risk of expiry of unused products. Whereas products facing a stockout can adversely affect the treatment regimen of patients. The network complexity of pharmaceutical supply-chains coupled with regulatory constraints and siloed planning systems force supply planners to rely on manual (error-prone) decision-making processes. Such an approach results in suboptimal capacity allocation and inventory management decisions. The overall equipment effectiveness (OEE) of pharmaceutical companies is only 31% compared to the 60% OEE in consumer-packaged goods [1].

In this talk, we present a success story of Genentech in developing and applying a stochastic optimization model for the production scheduling of multiple drug products in lyophilization units across multiple sites. The framework leverages information obtained from historical and forecast data to generate scenarios of uncertain parameters (e.g. yield, demand, and downtimes) that can realize in the future. The optimization model determines a product filling schedule that maintains product stock levels close to targets under diverse scenarios. We show that this approach helps avoiding reactive scheduling and in maintaining a more stable production plan than deterministic procedures (which ignore uncertainty). Stochastic optimization models for production planning and schedule have been extensively studied in literature [2, 3–5]. The majority of stochastic optimization models capture uncertainties in product demand [3, 5], product prices [6], and processing times [4]. Uncertainties in product yield can also be modeled in similar manner. To the best of our knowledge, none of the existing studies account for the loss in production capacity due to unplanned equipment downtime (or batch rejections) while making production scheduling decisions. We also note that there are no studies reported for the successful application of stochastic optimization approaches for capacity disruptions in an industrial setting. Even the leading commercial supply chain management tools used in industry such as SAP APO (26.6% market share [7]) and Oracle’s (13.7% market share[7]) do not offer any capabilities for handling uncertainty associated with loss in production capacity due to unplanned equipment downtime.

Our modeling framework seamlessly integrates stochastic optimization models into the supply planning structure of an industry, without overhauling the current industrial decision-making structure. State-of-the-art models can capture the impact of a downtime and batch rejection implicitly through a multistage stochastic program [4, 8] that solves the model at multiple intervals in the planning horizon once some of the uncertainties are realized e.g. taking a recourse action after an unplanned downtime has occurred. For supply planning in pharmaceutical industries, a multistage approach is not a suitable decision-making approach due to the long frozen horizons encountered in practice (time interval during which no changes are made to the production plan). During this frozen horizon the manufacturing sites prepare the raw materials and equipment required for manufacturing the next drug in the production plan. Crashing the frozen horizon (or taking a recourse action) whenever a unplanned downtime or a batch rejection occurs can disrupt the manufacturing operations and result in sub-optimal performance. In addition, frequent adjustments in the supply plan takes considerable planning resources (in terms of time and money) from the planning resources. We will conclude the talk by applying the modeling framework to determine the optimal filling decisions for drug products at Genentech. We will also showcase the improved performance achieved by the stochastic model over a deterministic counterpart.


[1] Philip Berk, Mark Gilbert, and Gideon Walter. Rethinking the pharma supply chain: New models for a new era available at https://www.bcg.com/en-us/publications/2013/biopharmaceuticals-operations-supply-chain-management-rethinking-the-pharma-supply-chain-new-models-for-a-new-era.aspx, 2013. [Online; accessed 18-June-2019].

[2] Zukui Li and Marianthi Ierapetritou. Process scheduling under uncertainty: Review and challenges. Computers & Chemical Engineering, 32(4-5):715–727, 2008.

[3] MG Ierapetritou and EN Pistikopoulos. Global optimization for stochastic planning, scheduling and design problems. In Global optimization in engineering design, pages 231–287. Springer, 1996.

[4] J Balasubramanian and IE Grossmann. A novel branch and bound algorithm for scheduling flowshop plants with uncertain processing times. Computers & chemical engineering, 26(1):41–57, 2002.

[5] J Balasubramanian and IE Grossmann. Approximation to multistage stochastic optimization in multiperiod batch plant scheduling under demand uncertainty. Industrial & engineering chemistry research, 43(14):3695–3713, 2004.

[6] Iddrisu Awuduand Jun Zhang. Stochastic production planning for a biofuel supply chain under demand and price uncertainties. Applied Energy, 103:189–196, 2013.

[7] Louis Columbus. SAP Leading The Fast-Growing SCM Market With 26% Share available at https://www.forbes.com/sites/louiscolumbus/2018/07/28/sap-leading-the-fast-growing-scm-market-with-26-share/#6572642470df, 2018. [Online; accessed 23-March-2020].

[8] Georg Ch Pflug and Alois Pichler. Multistage stochastic optimization. Springer, 2014.