The integration of prior information in learning-based methods is an important research direction, enabling the usage of these methods when the available data is sparse. In control problems we often have access to prior information in the form of constraints derived from physical laws which the systems under study are known to satisfy. A number of methods have recently been developed to enforce these types of constraints in learning-based methods, usually referred to as physics-informed learning. The aim of the project is both to develop new approaches within physics-informed learning which are appropriate for solving complex problems in control, and to evaluate the performance of these approaches in challenging control applications, such as improving charging capabilities of lithium-ion batteries or investigating how, in a human brain, a large network of neurons is capable of complex information processing with ultra-low power.