Artificial Intelligence-Assisted RFID Tag-Integrated Multi-Sensor for Quality Assessment and Sensing

Area of Science:

• Electrical Engineering
• Computer Science
• Food Science

Background:

• Radio Frequency Identification (RFID) is crucial for the Industrial Internet of Things (IIoT) but faces challenges in pervasive monitoring and item-level data recording.
• Existing IIoT systems often require complex, additional sensor networks, increasing costs and complexity.

Purpose of the Study:

• To develop an Artificial Intelligence (AI)-assisted RFID-based multi-sensing technology to overcome limitations in current IIoT monitoring.
• To integrate passive and semi-passive RFID tag-integrated multi-sensors for enhanced sensing capabilities.
• To apply machine learning algorithms for food product quality assessment and sensing (QAS) using RFID data.

Main Methods:

• Development of UHF RFID tag-integrated multi-sensors with boosted charge pumps for improved RF sensitivity and operational range.
• Hardware design optimization, including antenna and energy harvester.
• Integration and demonstration of the NARX (autoregressive model with exogenous input) neural network for analyzing RFID sensing data.

Main Results:

• A novel UHF RFID tag-integrated multi-sensor was designed and tested, showing high RF sensitivity and extended operational range.
• The NARX neural network model achieved high accuracy in ham product quality assessment and sensing (QAS), with an RMSE of 0.007 and R-squared of 0.99.
• Successful demonstration of AI-assisted RFID sensing for real-world food product quality assessment.

Conclusions:

• The proposed AI-assisted RFID multi-sensing technology offers a low-cost, timely, and flexible solution for product quality assessment in manufacturing.
• This technology enhances product quality, optimizes manufacturing lines, and improves supply chain management.
• The integration of machine learning with RFID data represents a novel approach for advanced sensing applications in the IIoT.