Conference Agenda

A low attendance rate for urgent, high-consequence treatment is prevalent in the developing world. In poorer countries, ignorance about the illnesses people have, allied to the financial costs of attending for treatment, are two of the main reasons why they fail to show up for medical appointments. Yet, for many diseases, patients may need surgical treatment soon after diagnosis to prevent severe disability. This project aims to identify ways to increase the attendance rate of patients who need emergency high-consequence treatment. Traditionally, in care systems across the world, information about why patients must undergo a particular treatment or care plan is communicated by emphasising medical aspects of the disease, typically in scientific terms. This research, in contrast, analyses the effect of providing emotionally engaging information about how a patient’s life experience can be changed by their medical condition. The study runs a two-stage, randomised controlled trial at the world’s largest eye hospital, the Aravind Eye Hospital in India. This study hopes to improve the chances of poorer populations receiving the care they need and to help reduce needless blindness. Furthermore, it hopes to inspire other healthcare providers globally to adopt the proposed methods.

In this paper, network-aware clearing algorithms for local energy markets (LEMs) and local flexibility markets (LFM) are proposed to be sequentially run and coordinate assets and flexible resources of energy communities (ECs) in distribution networks. In the proposed LEM clearing algorithm, EC managers run a two-stage stochastic programming while considering random events by scenario generation and network constraints using linearized DistFlow. As one of outcomes, maximum available up- and down-regulations provided by ECs are estimated in LEM and communicated to LFM. In the distributed LFM clearing algorithm, an iterative auction is designed using a dual-decomposition technique (Augmented Lagrangian) which is solved by consensus alternating direction method of multipliers. The LFM algorithm efficiently dispatches the flexibility provided by ECs in operating time while considering flexibility local marginal price as pricing method. Network constraints are included in the algorithm with an AC distribution optimal power flow for dynamic network topology in which branches and buses are decomposed to solve the problem in distributed fashion. The designed LFM algorithm can respond to exogenous and endogenous signals for flexibility requests. The simulation results in a test case display effectiveness of two proposed LEM and LFM algorithms for an efficient provision of flexibility.

Global supply chains have been optimized in terms of economic efficiency for many years. Driven by tight regulations as well as changing expectations of customers and other stakeholders, ecological and social sustainability aspects are becoming increasingly important in supply chain network design. The mathematical formulation of the sustainable network design problem leads to a multi-objective optimization model in which several sustainability indicators need to be balanced. We leverage activity analysis for modelling the sustainability impacts of individual supply chain processes and integrating them into the optimization problem. A special feature of this approach is the consideration of different types of sustainability indicators in the model’s objective function based on the decision-makers’ individual preferences. The optimization of this multi-objective model with an "a posteriori" approach results in a Pareto front from which the decision-makers can select their preferred solution. We demonstrate the application of the developed model combined with a state-of-the-art solution algorithm using the supply chain for lithium-ion batteries in the field of electric mobility. In this field, the model can be used, for example, by battery manufacturers to optimize the supply chain regarding the sourcing of raw materials, or by regulators to determine the impact of regulations on the optimal supply chain network.

We present an optimization model for the strategic management of a system of water reservoirs, when the goal is to sustainably satisfy vital demand. Given that both standard finite horizon and discounted infinite horizon approaches seem inadequate here, we optimize the expected shortage costs over cycles. A cycle starts and ends when all reservoirs in the system are full. To account for the unpredictable effect of climate change on future water supply, we consider a robust model where nature chooses the most adversarial inflow distribution. This leads to a stochastic shortest path problem under ambiguity. We discuss structural properties of our model as well as important policy insights. To overcome the curse of dimensionality, we present a scalable heuristic together with lower and upper bounds on the (optimal/associated) expected costs, allowing to solve the policy evaluation problem. The results of a case study for the Sacramento River Basin in Northern California complement the talk.