"The new technology allows real-time identification of the areas overgrown with cow parsnip, evaluation of the weed density, automatic transfer of the information, and creation of treatment routes with effective herbicides. After the treatment, the territories are continuously monitored to assess the effectiveness of the applied means," noted Igor Bevzyuk, CEO of 2050.digital. "Our comprehensive system was tested due to the Ministry of Agriculture and Food of the Moscow Region, and the Administration of the Naro-Fominsk urban district initiative."
Treatment of cow parsnip with herbicides is effective during the period of active weed growth in May-June and then in early autumn. To detect the cow parsnip accumulations at early growth stages, the system by 2050.digital implements multi- and hyperspectral survey and special image analysis programs making active use of artificial intelligence and machine-aided learning technologies. Subsequently, the treated areas are surveyed to assess the effectiveness of herbicides usage. If necessary, spraying drones can re-treat the remaining weed specimens.
Treatment maps with data on cow parsnip Sosnovsky growth density are drawn up based on satellite information and surveys performed by unmanned systems for accurate calculation of the required agrochemicals amounts and the intensity of treatment. These data are automatically transmitted to the spraying drones control system.
In general, the applied set of technologies allows saving 20-30% of funds spent on agrochemicals, performing weed control in hard-to-reach areas, and identifying areas requiring additional treatment.