Performing Department
(N/A)
Non Technical Summary
Rising input costs, particularly for fertilizer, have been an ongoing concern for agriculture in general but more recently have become increasingly serious as global supplies have been disrupted. For specialty crop farmers in particular, this adds pressure to an industry that faces multiple challenges that, if left unaddressed, could jeopardize the viable production of many vegetable, fruit and nut crops in the country. We propose to demonstrate a plant-by-plant precision fertilization capability for cool season vegetable crops that will increase harvest yields while reducing the overall amount of fertilizer applied. The solution will help farmers increase margins in the face of escalating production costs while meeting increasingly stringent environmental regulations.The overarching objectives of our variable rate application (VRA) approach are to determine and apply an optimal amount of fertilizer for each plant, to adjust for the inherent field and seed variations so that most plants meet customer specifications and are sold at harvest.An additional benefit of the anticipated capability will be less food waste, employing the EPA Food Recovery Hierarchy's most preferred method of reducing waste where the food is produced.Our innovation includes obtaining a priori information on a field's plant size distribution from high resolution aerial imagery of the entire field. The size distribution will be used to determine an application prescription, which will be sent to a mechanical controller on a tractor-mounted sprayer. Over time, associated plant growth models will be augmented to account for the effects of additional fertilizer, possibly even accounting for different, future varieties that are more drought tolerant or less fertilizer hungry. We will leverage recent experimental results indicating that different fertilizer application amounts lead to different growth rates of iceberg lettuce, so that applying different amounts to different sized plants will increase field uniformity at harvest and also reduce total required fertilizer amounts per crop cycle. The proposed effort will validate and extend these preliminary findings to develop a unique capability for vegetable production. During Phase I, we will determine geolocation accuracy requirements and evaluate if system components can meet them. We will assess timing and coordination requirements of aerial imagery collection, processing and conversion to a prescription map, and we will evaluate spray controller requirements and performance for the VRA approach. We will assess the overall technical feasibility of the proposed concept and perform a preliminary cost-benefit analysis for VRA commercialization in the vegetable industry. We plan to sell the precision fertilization capability as an add-on to subscription services we have already developed and commercialized for the fresh produce industry.
Animal Health Component
60%
Research Effort Categories
Basic
0%
Applied
60%
Developmental
40%
Goals / Objectives
The principal objective of this Phase I SBIR project is to demonstrate the feasibility of applying fertilizer to a given field to adjust for inherent field and seed variations and produce a higher yielding crop at harvest (more plants meeting buyer specifications) with less total applied fertilizer. A feasible solution will provide a sufficiently compelling value proposition for growers, in terms of cost savings in fertilizer and increased profits from more harvestable product meeting buyer specifications, to warrant prototype development in Phase II for rapid commercialization. Following are the specific technical objectives towards that end and the associated questions to be answered to determine the technical and commercial feasibility of the proposed concept:Determine geolocation accuracy requirements and evaluate if the system components can meet them.Assess timing and coordination requirements of aerial imagery collection, processing and conversion to a prescription map for an operational system.Evaluate spray controller requirements and performance for the variable rate application (VRA) approach.Assess the overall technical feasibility of the VRA approach.Perform a preliminary cost-benefit analysis for VRA commercialization in the vegetable industry.
Project Methods
Our plan to determine feasibility of the proposed VRA system is to evaluate, both in terms of vendor-published specifications and with field testing, capabilities of sprayer systems to accept prescription maps generated from drone imagery-derived plant sizing. We will work with two spray controller providers to modify and extend system capabilities to meet the requirements for vegetable production. We will focus on iceberg lettuce during Phase I, given the interest of our producer-partner and GeoVisual's prior experience with lettuce from developing its harvest forecasting capabilities.We will collect aerial imagery and process it using existing capabilities to provide assessments of plant size variability, running parallel A/B tests to isolate the effects of fertilizer application from other variables affecting growth rates and in-field variability. In-field monitoring with GeoVisual's mobile application will provide geopositioned photos and notes recorded in test fields as additional observational data to assess the effects of different fertilizer application rates and timing.Our execution plan is designed to coordinate with major growing seasons, which are roughly May-October in California, and November to April in Arizona. Assuming our Phase I effort begins in May of 2023, we will target data collection during the 2023 season in California.