Multimodal PCSC Sensors for Real-Time Temperature and Force Detection Using LRTNet.

Area of Science:

• Materials Science
• Robotics
• Sensor Technology

Background:

• Multimodal sensors are crucial for robotics but face challenges like signal crosstalk and slow real-time performance.
• Accurate detection of pressure and temperature is vital for measurement precision in sensing applications.

Purpose of the Study:

• To develop a multimodal sensor with reduced signal crosstalk and improved real-time performance for simultaneous force and temperature detection.
• To introduce a novel algorithm for enhanced temperature sensing accuracy by fusing multimodal sensor data.

Main Methods:

• A multimodal PCSC sensor was designed utilizing carbon quantum dots (CQDs) for pressure sensing (resistance variation) and spiropyran (SP) for temperature sensing (thermochromic properties).
• A lightweight ResNet-Transformer network (LRTNet) was developed to fuse resistance and color signals for accurate temperature detection.
• The sensor and LRTNet were tested on a robotic manipulator for dual force and temperature recognition.

Main Results:

• The PCSC sensor demonstrated signal separation, with force detection showing a 0.4 s response time and temperature detection stabilizing within 7.5 s.
• LRTNet achieved a runtime of 152.08 ms and a temperature sensing accuracy of 95% on the robotic manipulator.
• LRTNet improved overall performance by at least 11% compared to traditional algorithms, enhancing reliability.

Conclusions:

• The developed multimodal PCSC sensor effectively mitigates signal crosstalk and provides reliable force and temperature measurements.
• The LRTNet algorithm significantly improves temperature sensing accuracy and processing speed in multimodal sensor applications.
• This integrated sensor and algorithm approach advances the performance and dependability of multimodal sensing systems in robotics.