Data Fusion &
Deterministic Engine
Hazel-Guard operates at the intersection of global meteorological satellite data and hyper-local IoT edge-computing. We eliminate guesswork from agriculture.
The 3-Layer Architecture
How we ingest, fuse, and process data to generate a definitive "Spray or Wait" decision in real-time.
Macro Datasets
NASA POWER & Copernicus Satellite grids. Provides 8-day precipitation forecasts and regional weather models.
Micro IoT Network
Hazel-Guard edge gateways deployed in orchards. Provides live leaf wetness, wind speed, and canopy temperature.
Data Fusion Engine
Calibrates the low-resolution satellite models using our high-resolution ground truth IoT data via localized data fusion algorithms.
Deterministic Decision
Runs against agronomic thresholds to output the exact "Golden Window" for chemical application.
Rainfastness Window Calculation
Chemical pesticides do not work instantly. They require significant time to absorb into the leaf cuticle. If rain falls before this absorption process (Rainfastness) completes, the chemicals are washed directly into the soil.
"Most systemic fungicides and insecticides require a continuous dry period of 4 to 8 hours post-application for optimal leaf cuticle penetration without risk of runoff."
— Based on Willoughby et al. (2024)
Our Processing Scenario:
- 1
Our satellite link forecasts rain starting at 18:00.
- 2
The engine applies the 8-hour rainfastness offset, closing the viable spraying window exactly at 10:00.
- 3
If live IoT wind speed exceeds 3 m/s before 10:00, the system immediately cuts off the window to prevent ILO drift risks.
Science-Backed Foundation
Hazel-Guard's deterministic rules are strictly built upon peer-reviewed agricultural, IoT, and climatology research.
Hazelnut Quality Degradation
"Effect of Brown Marmorated Stink Bug, Halyomorpha halys Damage on Total Phenolic, Total Flavonoid and Antioxidant Activity in Hazelnut Kernel".
Scope 3 GHG Reduction
Determination of eco-efficiency and optimization of input utilization in hazelnut production of Türkiye. Proves the carbon-reduction claims.
IoT Edge Computing in Ag
Exploring IoT-enabled machine learning approaches for soil quality monitoring in agriculture. Validates our edge-computing infrastructure.