Reviewed by Lexie CornerNov 14 2024
A study published in Device details how researchers have developed a robot capable of identifying different plant species at various growth stages by "touching" their leaves with an electrode.
The robot can measure surface texture and water content, providing insights that visual techniques cannot. It successfully identified ten plant species with an average accuracy of 97.7 %, including the leaves of the flowering bauhinia plant at different growth stages.
Zhongqian Song, an associate professor at Shandong First Medical University and Shandong Academy of Medical Sciences and one of the study's authors, notes that the robot could eventually be used by large-scale farmers and agricultural researchers to monitor crop health and growth. This could help inform decisions on water and fertilizer usage as well as pest control strategies.
It could revolutionize crop management and ecosystem studies and enable early disease detection, which is crucial for plant health and food security.
Zhongqian Song, Study Corresponding Author and Associate Professor, Shandong First Medical University
Current devices typically use optical methods to collect data from plants, but these approaches are limited and sensitive to factors such as background interference, changes in lighting, and weather conditions.
To address these challenges, Song and his colleagues developed a robot that interacts with plants using a mechanism inspired by human skin. This design enables the robot to gather information through touch.
The robot measures several factors, including the charge that can be stored at a specific voltage, the resistance to electrical current passing through the leaf, and the contact force when the robot grasps the leaf. These measurements provide insight into the plant's characteristics. Since different values for these metrics correspond to specific plant species and growth stages, machine learning is used to process the data and identify the plant.
According to Song, while the robot shows significant potential for applications in areas like ecological research, precision agriculture, and plant disease detection, several issues still need to be addressed.
For example, the robot is not yet adaptable enough to reliably identify plant species with complex morphologies, such as burrs or needle-like leaves. Song suggests that improvements to the robot's electrode design could address this limitation.
Song added, “It may take a relatively long period to reach large-scale production and deployment depending on technological and market developments.”
The researchers aim to expand the robot's ability to identify a broader range of species by gathering data from more plant types and strengthening the database used to train the algorithms.
Additionally, Song mentioned plans to improve the integration of the sensor so that it can display data in real-time without requiring an external power source.
Journal Reference:
Chen, M., et al. (2024) Iontronic tactile sensory system for plant species and growth-stage classification. Device. doi.org/10.1016/j.device.2024.100615.