InnoProm 2026

When applying plant protection products (PPPs), such as in fruit growing and viticulture, as well as in the prevention of bark beetle infestations in forestry—the PPPs are currently applied to plants via spraying. However, this method of application is heavily influenced by environmental factors such as wind, rain, and heat. In addition to reduced effectiveness, this leads to drift or runoff, which contaminates the environment and, above all, groundwater. Recent research in the field of biochemistry has yielded PPPs that are injected into the plant via the xylem. This approach can reduce PPP usage by 90%, thereby significantly lowering the environmental impact. In contrast to spraying pesticides, injection has so far been performed manually for each individual plant and is therefore not suitable for the large-scale applications mentioned above.
As part of the proposed doctoral research, the automation of the injection process using an agricultural robot equipped with an injection unit will be investigated. The injection process must meet the following requirements:

• Varying topography and applications across different crop types
• Varying weather and light conditions
• Dynamic Behavior of the Agricultural Robot for Fast, Precise, and Energy-Efficient Injection

The scientific focus of this work lies in environmental perception, determining a suitable injection site on the plant, and controlling the injection robot. This will be investigated and prototyped using large language models and reinforcement learning. Clemens, an SME specializing in agricultural equipment, is very interested in the research results and, if the doctoral project is approved, will support the project by drawing on its long-standing expertise (particularly in viticulture) and by providing test equipment.