Redefining plant monitoring with innovative tactile sensor technology

By Dr. Chinta SidharthanReviewed by Lexie CornerNov 26 2024

In a recent study published in Device, researchers in China developed a tactile sensory system using iontronic interfaces to classify plant species and their growth stages. By measuring interfacial capacitance and resistance from plant leaves and applying machine learning algorithms, the system demonstrated high recognition accuracy.

Image Credit: lovelyday12/Shutterstock.com

Background

Plants exhibit surface features such as texture, hardness, and water content that vary across species and growth stages. These characteristics are important for ecological studies, smart agriculture, and plant health monitoring.

Existing plant monitoring technologies primarily rely on visual methods or wearable electronic sensors. While widely used, visual techniques are affected by environmental conditions such as light and weather, whereas wearable devices cannot measure mechanical properties like texture and hardness.

Recent research has shown that tactile sensors can perceive the physical characteristics of plant surfaces. While microstructured tactile sensors can detect properties like hardness and texture, they are less effective at distinguishing between similar leaves at different growth stages.

Non-steady-state iontronic interfaces, which use ions as signal carriers and respond to physical stimuli, have been effective for collecting biological data but remain underexplored in plant monitoring applications.

About the study

In this study, researchers developed a tactile sensory system with a non-steady-state iontronic interface to classify plant species and growth stages. The system integrated an iontronic interface with a built-in piezoresistive (BIP) sensor to measure resistance, interfacial capacitance, and contact force.

The interface, inspired by human skin, utilized asymmetric conductive electrodes fabricated through modified magnetron sputtering. When embedded in plant leaves, these electrodes enabled pressure-sensitive measurements by forming a responsive iontronic interface. Surface properties such as texture and water content were captured as capacitance and resistance signals.

The system was further enhanced with a machine-learning algorithm for data classification. A neural network architecture processed the interfacial data, achieving high recognition accuracy.

Bauhinia leaves and other plant species were used for testing and validation, covering a wide range of morphological variations and growth stages. Additionally, contact forces recorded by the BIP sensor refined the data processing capabilities of the system.

To optimize sensory capabilities, a three-dimensional array of parameters—force, resistance, and capacitance—was recorded, enabling precise clustering of plant samples for classification tasks.

The interface design was biocompatible and reversible, as confirmed through mechanical testing and cytotoxicity assays. The system was further used to compare plant species with similar morphologies and to analyze growth status through repeated testing.

Major findings

The study demonstrated that integrating iontronic interfaces with tactile sensing allows for the accurate classification of plant species and growth stages based on the interfacial properties of leaves. The system achieved 97.7 % accuracy in identifying ten plant species and 100 % accuracy in distinguishing growth stages within a single species.

Analysis revealed that the system effectively measured interfacial properties such as texture, water content, and hardness. Fresh leaves showed higher capacitance and lower resistance due to greater deformability and ion conductivity. Using machine learning models, the system successfully clustered interfacial data for plant identification with minimal sample overlap.

Further testing demonstrated the system’s ability to differentiate plant leaves with similar morphological features, maintaining high accuracy rates and confirming its robustness. The system was also used to track growth status before and after watering, showcasing its adaptability to dynamic conditions.

Mechanical evaluations indicated that the device maintained reliability over thousands of contact cycles, demonstrating high durability. Biocompatibility tests confirmed the safety of the materials, ensuring suitability for biological applications. By integrating advanced sensing capabilities with machine learning, the system offers a scalable and precise method for agricultural and environmental studies.

Conclusions

Overall, the study highlighted the effectiveness of combining iontronic tactile sensing with machine learning for classifying plant species and growth stages. By capturing interfacial properties and contact forces, the system achieved high accuracy and robustness. Its adaptability to diverse plant morphologies and growth conditions highlights its potential for precision agriculture and ecological monitoring.

The intronic tactile sensory system represents a notable advancement in plant monitoring technologies. It enables advanced, noninvasive analysis and contributes to sustainable agricultural practices.

Journal reference

Chen, M., Song, Z., Liu, S., Liu, Z., Li, W., Kong, H., Li, C., Bao, Y., Zhang, W., Niu, L. (2024). Iontronic tactile sensory system for plant species and growth-stage classification. Device. DOI:10.1016/j.device.2024.100615, https://www.cell.com/device/fulltext/S2666-9986(24)00570-2