• machine learning
• optimization
• ergotropy
• quantum mechanics
• quantum thermodynamics

Traditionally, thermodynamics deals with systems composed of a large number of particles. With the continuous miniaturization of electrical user devices, we necessitate to take into account quantum mechanics effects. In the first part of the thesis, we analyze the maximum extractable energy from systems with continuous spectrum Hamiltonians. We identify an explicit state, attainable through unitary evolutions, enabling us to extract all the energy from an arbitrary initial state subject to the free particle Hamiltonian. Building on this achievement, we extend our exploration to assess the work extractable from other systems with continuous spectra. We find that the state used for the free particle Hamiltonian permits the complete extraction of energy. In the second part, we analyze the optimal work extraction process. We develop a numerical algorithm employing machine learning techniques to achieve a unitary evolution from an arbitrary initial state to the previously determined state with minimum energy.