Designing and evaluating the use of crop signaling markers for fully automated and robust weed control technology

Abstract. The primary technological aspect that prevents the development of a fully automated and robust system for weed control in specialty row crops, like vegetables, is the lack of a reliable weed and crop plant detection system that can allow effective mechatronic removal of weeds growing between crop plants. In a weedy field with moderate to high weed density, existing robotic weeding machines become confused since they cannot interpret the complex natural scenes of weed-crop mixtures at higher weed densities. A novel solution to the this problem has been developed at the University of California, Davis, named Crop Signaling that provides reliable weed-crop differentiation in the field. Crop Signaling is a unique, machine-readable signaling technique that marks crop plants at planting. It creates optical signatures that are unique and significantly simplify and ensure the differentiation task. To be effective and commercially viable, the development of the Crop Signaling technology with unique optical signals is required to meet certain technological criteria: robust and applicable for use in an outdoor agricultural environment, sufficient signal-to-noise ratio to enable real-time signal detection while moving, satisfactory signal lifetime (i.e., environmental stability) of the optical signal, and minimal leakage of the applied material from crop plants to weeds and soil. This study focuses on designing and evaluating the use of Crop Signaling compounds and physical markers to achieve satisfactory signal detection and biodegradability.