This paper compares the three major maintenance strategies reactive (RM), periodic (PM) and predictive maintenance (PdM) through a mathematical and simulation-based framework. Building on the Weibull-distribution and the cost rate-model, analytical formulas were created for each strategy and extended to multi-component systems where several parts are maintained jointly under unified setup costs. Different cost structures are modelled and numerically presented. All analytical results were validated using Monte-Carlo simulations in Python. PdM generally achieved the lowest average cost rate, particularly for components with reliable predictive detection signals. In the multi-component comparison, semi-heuristic optimal, no and full maintenance-grouping were employed and evaluated. The paper also mentions how the mathematical parameters presented in the analytical chapter influence qualitative detection strategy choices, discusses the applicability of the model to industry scenarios and provides an outlook especially highlighting the influence of advancements in artificial intelligence on maintenance choices. It aims to give a cohesive comparison of different maintenance strategies and groupings in order to enable data-based maintenance decisions.