Thermal runaway of lithium-ion batteries is a critical concern for clean energy storage and use. Various methods have been investigated for providing early warnings for thermal runaway. Changes in parameters such as temperature and voltage, induced by side reactions during thermal runaway, enable early warning through threshold-based monitoring. Among these parameters, strain has proven to be a more effective parameter for early warning, as it remains low under normal conditions but shows significant changes earlier than temperature and voltage in certain circumstances. In this study, simulations were conducted to analyze strain behavior and establish warning thresholds, with a proposed threshold of 80 mu epsilon. We propose a high-precision, wide-range strain sensor specifically for early strain characteristics of battery thermal runaway and validate it through experiments. Surface acoustic wave (SAW) strain sensors, chosen for their sensitivity and reliability, were calibrated on a beam with uniform strain and tested on a battery shell. Experimental results show that the sensors can detect strain changes below 10 mu epsilon with nonlinearity of less than 1%. The SAW sensors perform effectively and can provide early warnings for thermal runaway in various battery types when appropriate thresholds are set.