Dual-Modal Material Identification Method via MTEG-TENG Synergistic Sensing and Machine Learning Optimization in...

Area of Science:

• Robotics and Sensor Technology
• Materials Science
• Energy Harvesting

Background:

• Robots require advanced material identification sensors for intelligent perception.
• Environmental complexities challenge the accuracy of current material identification sensors.
• Developing robust sensors is crucial for robotic innovation.

Purpose of the Study:

• To propose and validate a dual-modal collaborative material identification method.
• To assess the sensor's performance under various external and contact conditions.
• To integrate machine learning for enhanced material classification.

Main Methods:

• Fabrication of a prototype combining a microthermoelectric generator (MTEG) and a triboelectric nanogenerator (TENG).
• Utilizing MTEG for thermal diffusivity measurement and TENG for electron affinity measurement.
• Establishing a validation system to test material identification under varying conditions.

Main Results:

• The dual-modal sensor effectively distinguishes materials based on thermal diffusivity and electron affinity.
• Performance analysis revealed the impact of external factors like temperature, humidity, pressure, and surface roughness.
• Integration with machine learning achieved 93.54% accuracy in identifying eight materials.

Conclusions:

• The proposed MTEG-TENG dual-modal method offers a robust solution for material identification in robotics.
• The sensor demonstrates significant potential for accurate material classification even under challenging environmental conditions.
• This technology advances intelligent robotic perception and material handling capabilities.