Dear Optimisation group members,
Please note the following Biomodelling talk tomorrow (Friday 22 September) that may be of your interest.
Speaker: Nigel Clay
School of Science, RMIT University
Title: Ensuring Blood is Available When it is Needed Most
Date and time: Friday 22 September 2017, 3:00–4:00pm
Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus
Abstract: A robust supply Red Blood Cells (RBCs) is needed to meet patient demand for transfusion. The quantity of blood transfused is doubly stochastic as both the number of patients needing transfusion and the quantity of blood a patient needs are random. We define a shortage as the situation where RBCs are needed for transfusion but none are available. Shortages may arise because unused RBCs perish, and/or because the supply of RBCs is not sufficient to cover demand. For hospitals, the supply of fresh RBCs comes from the Blood Bank. However, the Blood Bank has to rely on donations which are also stochastic. Further, once a donor has given whole blood they are ineligible to do so again for 84 days. Therefore, to ensure a robust supply of RBCs is available is important to ensure that a sufficient quantity of donors are available to give blood.
In previous research the volatility of RBC inventory levels and orders has been assumed to be a function of volatility in demand and supply. But this does not capture all the volatility present. This research uses system dynamics to show how the structure of the blood supply chain itself can induce additional volatility. While the supply chain may be altered to mitigate this volatility, this does not imply that the resulting system will be optimal. We use discrete time stochastic control to find an order quantity that minimises the probability of shortages and out-dates. However, the state space of this problem is extremely large making it intractable for computation. To address this, we use sample average approximation to reduce the computational complexity. The approach requires estimation of plausible demand trajectories and we describe how these can be generated.
Bio: Nigel Clay is a final year PhD student with over 25 years of experience working in industry as a quantitative specialist. Prior to joining RMIT as a PhD student Nigel was a Senior Manager at the National Australia bank and prior to that a Senior Manager at ANZ bank.