Location: Sustainable Agricultural Systems Laboratory
Project Number: 8042-30400-001-000-D
Project Type: In-House Appropriated
Start Date: Sep 3, 2025
End Date: Sep 2, 2030
Objective:
Objective 1: Develop a modular, scalable plant sensing platform for crop and weed detection, identification, and performance mapping (species, biomass, and stress) for crop production and breeding researchers and farmers.
Sub-objective 1.A - Develop and curate a high-resolution, open-access, annotated crop and weed plant image repository from semi-field and field grown plants.
Sub-objective 1.B. Calibrate and validate PlantMap3D across cover crop-based corn, soybean, and cotton-producing regions of the U.S. for species and biomass mapping of crops and weeds.
Objective 2: Develop and test new technologies and management strategies that improve crop productivity and stability, as well as organic and herbicide-resistant weed management.
Sub-objective 2.A. Adapt PlantMap3D for high throughput phenotyping to enhance and accelerate grass and legume cover crop screening and improvement for higher biomass, winter hardiness, soft seed, disease resistance, and allelopathy.
Sub-objective 2.B. – Integrate PlantMap3D with precision fertilizer and herbicide actuators in on-station and on-farm trials to quantify the interaction between climate, soil, and agronomic management on weed suppression and nitrogen management.
Approach:
Our research will develop integrated weed management (IWM) solutions that leverage state-of-the-art machine vision and learning, real-time data aggregation, advanced analytics (traditional statistics and AI/ML), and cloud and edge computing to refine multi-tactic weed management frameworks and develop farmer-focused decision support tools (DSTs). Key efforts include creating a national plant image repository (AgImageRep) and PlantMap3D, a platform-agnostic system integrating hand-held, robotic, and tractor-mounted cameras with machine learning algorithms to map cash crops, cover crops, and weed composition and biomass. These technologies, linked to decision support tools, enable high-resolution mapping and facilitate variable-rate nutrient and weed management solutions based on crop yield stability maps and underperforming field components, promoting precise agricultural management at both field and landscape scales. These advancements will enhance knowledge for site-specific IWM applications, deliver real-time precision weed management technologies, reduce costs, and improve research efficiency. By advancing IWM theory and applications, this research aims to make food systems more productive, profitable, efficient, sustainable, and resilient. Emphasizing ecological and nonchemical tactics like cover crop use and other cultural control strategies, these solutions conserve existing herbicide chemistries, support long-term conservation tillage for improved soil health, and ensure the viability of weed management strategies while mitigating agriculture's environmental impact. Ultimately, this approach de-risks weed management, counters practices that drive herbicide resistance and increased costs, and enables both conventional and organic farmers to scale up production effectively.