Keynote Lectures

Available Soon
Oliver Rose, Universität der Bundeswehr München (University of the Federal Armed Forces Munich), Germany

Available Soon
Peer-Olaf Siebers, University of Nottingham, United Kingdom

The Optimum is not Enough: Robustness, Alternatives, and Dealbreakers
Francesco Contino, UCLouvain, Belgium

Abstract
Available Soon

Abstract
Available Soon

Abstract
Designing the optimal solution is at the core of many engineering and modelling activities. Yet a single optimum is fragile: it is only optimal for the assumptions behind it, and a change in costs, demand or technology performance can turn it into a poor choice. Including uncertainties and propagating them to the outputs reveals a wealth of information: which parameters drive the result, how wide the range of outcomes is, and which designs remain good across many possible futures. But what happens when we do not have all the objectives, or when some cannot be formulated upfront? Decision makers rarely want a single answer. They prefer alternatives they can compare in light of objectives that are difficult to quantify, or that only reveal themselves once the options are on the table. They would also prefer not to have surprises, at least not bad ones. Identifying dealbreakers, the conditions under which a solution fails, and the policies that mitigate them is therefore as important as finding the optimum. In this talk, I will show how robust optimization, uncertainty quantification and the generation of near-optimal alternatives can be combined to support decisions rather than replace them. I will illustrate these methods with applications to energy systems, from the micro scale (a factory, a district) to the macro scale (a continent), and discuss what they change for modellers and for those who use their results.