Hyperspectral UAV imagery for precision water management in smallholder crops
A UAV-based hyperspectral imaging system that monitors crop water stress in baby corn and french beans using vegetative indices and machine learning. Designed for small-scale irrigation farms in arid regions, it enables data-driven, sustainable water management for smallholder cooperatives seeking to optimize yields and meet international compliance standards.
Overview
This solution leverages Unmanned Aerial Vehicle (UAV) mounted Hyperspectral Imaging (HSI) technology to monitor water stress levels in baby corn and french beans, enabling Precision Water Management (PWM) for smallholder farmers in water-scarce regions. By capturing spectral data across the 400 nm to 1,000 nm wavelength range, the system derives vegetative indices, such as the Normalized Difference Vegetation Index, that quantitatively reflect plant physiological health. Machine learning algorithms then translate these indices into actionable predictions of crop water stress, supporting timely and efficient irrigation decisions.
The research is conducted at Sultan Green Farm Technologies in Makueni County, Kenya, a region where rain-fed dependency makes sustainable water use critical. The longitudinal study spans two farming cycles over approximately one year, generating robust field-validated insights. For agricultural cooperatives and international buyers, this technology integrates seamlessly with existing compliance and traceability infrastructure, strengthening the documentation of sustainable farming practices.
Technical specifications
Imaging system:
- Hyperspectral camera capturing wavelengths from 400 nm to 1,000 nm at approximately 4 nm resolution
- UAV-mounted platform for aerial field data collection over small-scale irrigation farms
Data processing pipeline:
- Pre-processing of captured imagery using statistical software packages
- Pixel extraction to convert spectral images into structured tabular datasets
- Dimensionality reduction via Principal Component Analysis (PCA) to streamline vegetative indices
Analytical approach:
- Derivation of quantitative vegetative indices correlated to plant water stress levels
- Machine learning models trained to predict water stress from spectral signatures
- Longitudinal validation across two complete farming cycles
Validation basis:
- Builds on prior controlled studies demonstrating that spectral profiles across five differentiated water treatment groups correlate reliably with plant stress levels
Technology readiness level
The technology is currently at an early-to-mid stage of development, transitioning from controlled-environment validation toward field deployment. Prior research using spectral imaging on apple trees has established proof-of-concept for optical sensor-based plant stress detection. The current phase involves active field data collection at a small-scale irrigation farm in Makueni County using a UAV-integrated hyperspectral camera. The study design, spanning two farming cycles over one year, provides a rigorous longitudinal validation framework. A key dependency is the procurement of a UAV equipped with HSI capability, which represents the primary partnership opportunity for advancing this technology toward operational readiness.
About Sultan Green Technologies
Sultan Green Technologies builds data infrastructure and compliance systems designed to assist East African smallholder farmers in accessing international markets. The company develops supply chain verification systems that collect and structure data on pesticide compliance, chain of custody, and traceability. By capturing field-level information and pre-filling compliance documentation, the technology helps producers meet the rigorous standards required by global buyers, such as those mandated by the EU Deforestation Regulation. The platform aims to make complex certification processes achievable for smallholder cooperatives, enabling them to provide audit-ready evidence of sustainable farming practices without needing to replace formal certification systems.
For customers and partners, Sultan Green Technologies addresses the significant compliance gaps that often exclude smallholder farmers from high-value global trade. The company partners with cooperatives in Kenya's highland coffee regions to develop and prove its methodology, with plans to expand into other value chains like horticulture. By building verified sourcing records, the company helps farmers market their adherence to soil health and regenerative practices, ultimately enabling them to command better prices for their food. This infrastructure is essential for cooperatives that meet quality standards but lack the digital tools to document their performance to the satisfaction of international buyers.
delivered to your inbox.