The developed PEDOT:PSS/CNT composite thermoelectric thin film, enables to realize self-powered monitoring of smokeless temperature rising with a high sensitivity of S=226.2 μV/K. By capturing Voltage Waveform Fingerprints through the ADS1220 and integrating the TimesNet model with TinyML, it evolves fire safety from passive notification to proactive prediction, establishing a highly efficient and resilient safety network for homes and vulnerable populations.

The developed PEDOT:PSS/CNT composite thermoelectric thin film provides the high-precision heat detection, overcoming the bottlenecks due to battery loss or circuit damage. Compared to traditional sensors that typically suffer from a 10-to-60-second delay, this TimesNet-assisted system achieves millisecond-level instantaneous response through carrier physical displacement imaging and moves beyond broad spatial monitoring to achieve device-level, "zero-distance" sensing.

The core technology lies in the developed PEDOT:PSS/CNT composite thermoelectric thin film, showing high-precision sensitivity on minimal temperature change and resolving the critical pain point of traditional sensors failing due to power interruptions in high-temperature environments. By involving Voltage Waveform Fingerprints" through an AI diagnostic mechanism; ADS1220 with temporal deep learning models, the design enables the real-time alerts and tracking of abnormal temperature rises.