Chinese researchers from Shandong First Medical University and Shandong Academy of Medical Sciences have developed a robot that can recognize plants at different growth stages simply by touching them with an electrode. The robot can also measure the surface texture and water content of the plant, which should make it suitable for agricultural use.

Anzeige

The electrode with which the robot touches a plant leaf measures several properties simultaneously, the scientists write in the study “Iontronic tactile sensory system for plant species and growth stage classification”which was published in Device. This includes, for example, how much charge can be stored at a certain voltage. It also measures the difficulty of the electric current passing through the leaf. The contact force is also measured as soon as the robot grips a leaf.

AI Recognition

The data thus collected is analyzed using artificial intelligence (AI). The plant can then be classified. It can be assigned different growth phases that correspond to the respective measurements. So far, the robot has been able to identify ten different plant species with an average accuracy of 97.7 percent.

The accuracy of recognition also depends on the complexity of plant structures. For example, it is not yet possible to recognize burdocks or conifers. According to the researchers, more work needs to be done to improve the design of the robot’s electrode.

But first, scientists want to increase the number of plants that can be recognized. To achieve this, they want to train the AI ​​system with more plants. The sensor also needs to be developed so that it provides real-time data, as the system has so far taken some time to collect the information.

Scientists see a possible area of ​​application for this technology in agriculture. Farmers can use it, for example, to monitor plant growth. Decisions on how much water and fertilizer they need could be made more quickly and in a more targeted manner. This would also allow recognition at an early stage if parasites are present and need to be controlled.

The scientists see several advantages in their technology compared to contactless systems that operate solely based on visual and AI methods. They are more sensitive to disturbances such as lighting conditions, climate changes or background noise.

(olb)