Robust Flexible Electret Tactile Sensor for Identification on Mushy Material in Harsh Environment

Abstract

In traditional food industry, the assessment of mushy materials plays an important role in high-quality food production, which still relies heavily on human tactile perception. In this work, to address this issue for enhancing food production efficiency, a flexible electret tactile sensor that can mimic expert touch is developed. The sensor consists of a pair of electrodes, a microstructural spacer, and a pre-charged electret. Benefiting from the electrostatic induction-based working mechanism, the sensor attains high sensitivity and is ideal for precise sensing in actions similar to those of skilled Baijiu distillers. Due to its hermetic and electromagnetic interference-resistant encapsulation with polypropylene/aluminum/parylene films, the sensor remained durable in vinasse in the real distillery, with its high water and alcohol content, for over 21 days. This demonstrates its long-term stability in harsh environment. Based on the proposed flexible electret tactile sensor, an automated and intelligent vinasse identification system is built, mimicking the actions of Baijiu distillers in vinasse assessment. Combining tactile sensing data with machine learning, the system can distinguish 8 kinds of vinasses with different ingredient ratios, achieving an accuracy of 98%. This work significantly demonstrates the practical potential of the sensor in the food industry.