A humidity-stable thermoelectric hybrid material is developed for a self-powered triple sensing system. The material, composed of poly(3,4-ethylenedioxythiophene) (PEDOT):polystyrene sulfonate (PEDOT:PSS) and multi-walled carbon nanotubes (CNTs), enables simultaneous and accurate pressure and temperature sensing in a single sensor. The enhanced planarity of PEDOT and charge transfer between PEDOT:PSS and CNTs, facilitated by strong π-π interactions, contribute to the high humidity stability. The sensor demonstrates 97.78% accuracy in intelligent object identification. Combined with a triboelectric nanogenerator (TENG), the system offers potential for industrial applications in intelligent classification without visual input. The sensor is flexible, cost-effective, and environmentally friendly, with high sensitivity and a wide temperature detection range. It can also detect sound and human motion. The triple sensing system, integrated with PDMS-based TENG, enables self-powered operation and is suitable for applications in soft robotics and human-machine interaction. The sensor's humidity stability is achieved through the combination of DMSO and CNTs, which enhance the electronic pathway and reduce ionic conduction. The sensor's performance is validated through various characterization techniques, including GIWAXS, Raman, and FTIR. The system shows excellent repeatability and durability, with over 2500 cycles of compression and release. The sensor is capable of detecting objects and identifying them with high accuracy using machine learning. The study highlights the potential of the triple sensing system for real-time monitoring and intelligent object identification.A humidity-stable thermoelectric hybrid material is developed for a self-powered triple sensing system. The material, composed of poly(3,4-ethylenedioxythiophene) (PEDOT):polystyrene sulfonate (PEDOT:PSS) and multi-walled carbon nanotubes (CNTs), enables simultaneous and accurate pressure and temperature sensing in a single sensor. The enhanced planarity of PEDOT and charge transfer between PEDOT:PSS and CNTs, facilitated by strong π-π interactions, contribute to the high humidity stability. The sensor demonstrates 97.78% accuracy in intelligent object identification. Combined with a triboelectric nanogenerator (TENG), the system offers potential for industrial applications in intelligent classification without visual input. The sensor is flexible, cost-effective, and environmentally friendly, with high sensitivity and a wide temperature detection range. It can also detect sound and human motion. The triple sensing system, integrated with PDMS-based TENG, enables self-powered operation and is suitable for applications in soft robotics and human-machine interaction. The sensor's humidity stability is achieved through the combination of DMSO and CNTs, which enhance the electronic pathway and reduce ionic conduction. The sensor's performance is validated through various characterization techniques, including GIWAXS, Raman, and FTIR. The system shows excellent repeatability and durability, with over 2500 cycles of compression and release. The sensor is capable of detecting objects and identifying them with high accuracy using machine learning. The study highlights the potential of the triple sensing system for real-time monitoring and intelligent object identification.