State of health is an essential parameter for the proper function of large battery packs. A wide array of methodologies has been proposed in the literature to track state of health, but they often lack the proper validation needed to be universally adaptable to large deployed systems. This is likely induced by the lack of knowledge bridge between scientists, who understand batteries, and engineers, who understand controls. In this work, we will attempt to bridge this gap by providing definitions, concepts and tools to apply necessary material science knowledge to advanced BMS. We will address SOH determination and prediction, as well as BMS implementation and validation using the mechanistic framework developed around electrochemical voltage spectroscopies. Particular focus will be set on the onset and the prediction of the second stage of accelerating capacity loss that is commonly observed in commercial lithium ion batteries.