Multi-model collaborative early warning for battery systems with integrated fiber optic sensors.

Multi-model collaborative early-warning framework with distributed fiber optic sensors achieves up to 706 seconds warning lead time in laboratory modules and 554 seconds in real-world vehicle operation.

• Why it matters: Current battery management systems are limited by sparse temperature data and monolithic diagnostics, which hinder timely detection of thermal runaway risks. Addressing this gap can improve safety and prevent failures.
• What they did: The authors developed a multi-model framework using distributed temperature sensing, increasing sensing points from 32 to 540, and established a dynamic warning boundary based on temperature Shannon entropy and temperature variation.
• The result: This approach enables early warnings with significantly extended lead times, demonstrating the potential of distributed sensing and collaborative diagnostics for proactive thermal-runaway risk identification.