Proton Exchange Membrane (PEM) fuel cells are important for clean energy, but they are not widely used yet because they can wear out early and become unreliable over time, even from small, hard to detect early-stage degradation. Traditional diagnostic methods often lack the sensitivity and spatial resolution to detect incipient failures before performance significantly declines. We propose developing a non-invasive optical early warning system utilizing femtosecond laser-induced Thermal Lens Spectroscopy (TLS), leveraging its ultra-sensitivity and depth-resolved capabilities. PEMs are thin and difficult to probe with traditional methods. TLS allows measuring in-plane or through-plane thermal diffusivity by analyzing the time-resolved thermal lens signal. This system will enable real-time, localized monitoring of thermal and optical properties within specific PEM fuel cell layers (e.g., catalyst layer, membrane). By correlating depth-resolved thermal lens signatures with component health metrics, TLS aims to identify and differentiate subtle degradation modes, such as catalyst degradation and membrane dry-out, at their earliest stages. The system's unique ability to probe thin PEM structures with high spatial (10-100 μm) and temporal (ms-μs) resolution, without disrupting normal operation, provides a significant advantage over conventional electrochemical techniques. Our ultimate vision is to create a physics-based "failure signature library" that predicts impending failures with substantial lead time, enabling proactive maintenance strategies and significantly extending PEM fuel cell lifetimes, thereby addressing a major challenge in fuel cell technology.