Precision agriculture sensing and AI for specialty crop management

Overview

This solution combines geospatial sensing with process-based artificial intelligence to deliver real-time, actionable management support for specialty crop producers. By integrating on-the-go soil sensing, aerial remote imagery, and near-ground canopy measurements, the approach captures the spatial and temporal variability of plant-environment interactions that drive yield, water use, and fruit quality. The result is a precision management framework designed to help farmers intensify production sustainably while conserving water and nutrient inputs.

The technology is particularly relevant for high-value specialty crops in Mediterranean and semi-arid climates, including citrus, pistachio, and almond, where soil texture variability and irrigation decisions strongly influence outcomes. Producers and agricultural partners benefit from improved estimation of crop water use, more targeted fertilization, and better-informed harvest decisions.

Technical specifications

Key features:

• Multi-scale geospatial sensing combining near-ground sensors, UAV-based imagery, and daily high-resolution satellite data
• On-the-go soil mapping using gamma-ray, apparent electrical conductivity, and terrain features collected by an all-terrain vehicle platform
• Near-ground canopy sensing through inexpensive lidar and multi-spectral sensors mounted on the same vehicle to capture canopy geometry and spectral properties
• Data fusion of ground-based and remote sensing layers to characterize in-season tree physiology, soil-water-plant relationships, and their implications for yield and fruit quality
• Process-based AI models calibrated against sensor measurements to provide real-time decision support for irrigation, fertilization, and harvest management
• Downscaling capability that translates coarser, more expensive measurements such as evapotranspiration into finer, more actionable spatial information

Technology readiness level

The underlying methods have been validated over a decade of research by the lead scientist and three years of lab-level development, including field campaigns in California citrus, pistachio, and almond systems. Recent work demonstrated that integrating on-the-go soil sensing with aerial plant information significantly improves estimation of yield and crop water use, while highlighting the need for near-ground canopy sensing to better connect UAV imagery with ground-truth leaf and fruit measurements. Future validation will add lidar and multi-spectral sensors to the existing all-terrain sensing platform, conduct system testing at university research farm citrus groves, and integrate daily high-resolution satellite imagery to refine soil-water-plant models. The approach is at a stage suitable for collaborative field validation and pilot deployment with grower and industry partners.