Baloch, GohramGzara, Fatma2020-05-252020-05-252020-04-01https://doi.org/10.1016/j.ejor.2019.09.007http://hdl.handle.net/10012/15907The final publication is available at Elsevier via https://doi.org/10.1016/j.ejor.2019.09.007. © 2019. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/MedAvail Technologies Inc. is a healthcare technology company that develops new technologies for self-serve pharmacy solutions. The technology, called MedCenter, is a pharmacy kiosk that provides 24/7, easy, and reliable access to pre-packaged prescription drugs and over the counter medications. To meet its business goals of having the right medication in the right kiosk at the right quantity, MedAvail faces several challenges related to assortment and stocking decisions of medications in the kiosk limited by kiosk capacity. This research addresses these decisions through an analytics project aimed at analyzing pharmaceutical sales, determining optimized kiosk storage capacity and service levels, and recommending assortment, stocking, and supplier-driven product substitution guidelines. We developed several mixed integer optimization models that use sales data to obtain robust solutions with respect to randomness in demand. We perform extensive testing using real as well as randomly generated data, and under multiple substitution rules, replenishment guidelines, and demand prediction strategies. Our results show that supplier-driven product substitution could save up to 9% in storage capacity depending on the desired service level and characteristics of product demand. We also propose a column-generation based heuristic approach that, on average, obtains near optimal solutions within 1.1% of optimality gap while reducing computational times by a factor of three.enAttribution-NonCommercial-NoDerivatives 4.0 InternationalOR in health servicescapacity planningassortmentcolumn generationArticle