The Human-O project focuses on adapting optimization algorithms to the needs of cyber-physical and social systems, in particular, information flows and human presence. The aim is to design new algorithms that learn both how optimization problems evolve over time, due to changing conditions, and which user-specific goals and constraints to incorporate into the problem itself.
The research stages of the project focus on learning the dynamic system underlying the optimality conditions, on integrating human presence by designing and learning user-specific costs and constraints, and on allowing users a choice among potential decisions, thus giving humans an active role in the optimization process.
The ultimate goal is to shape the emerging field of cyber-physical and social systems, where humans are an integral part of a complex and constantly evolving environment, but which nevertheless requires optimization algorithms to be able to give solid guarantees of safety, reliability and performance. This will change the way humans perceive and interact with technology.