UC Berkeley

The last several decades have seen revolutionary advances in our understanding of nonequilibrium thermodynamics. Designed originally to study macroscopic systems manipulated slowly, these new approaches to thermodynamics have extended the applicability of the theory to the small, fluctuating systems most relevant to biology and nanoscale engineering. With this powerful framework in place, a significant effort has been made to not only understand but control and manipulate such systems to achieve desired goals, and to do so optimally. In this thesis, after summarizing relevant history and background, we explore two major facets of this thermodynamic control problem. First, we investigate shortcut engineering, a control problem aimed at addressing rapid system transformation from an initial state to a target state. Second, utilizing a geometric approach, we study the control of slowly-driven systems, with particular focus on the optimal, i.e. maximally efficient, control of mesoscale thermal machines. Finally, we demonstrate a novel connection between these two disparate topics: although seemingly disconnected, the optimal control problem for certain shortcutting settings is identical to that of slowly-driven systems. We conclude by commenting on both the scientific merit and pragmatic calculational utility of this connection and on potential implications for future work.