Project background
Growers managing multiple crops needed a way to standardize and share grow profiles — the combinations of light, nutrient, irrigation, and climate settings that vary across a plant's lifecycle. Ad-hoc spreadsheets were causing inconsistent outcomes.
Challenge
Designing a recipe format flexible enough to capture any crop while remaining strict enough to validate automatically. Supporting stage transitions based on time, sensor thresholds, or operator approval, and versioning recipes so improvements could be traced.
Approach & solution
We defined a structured recipe schema with stages, setpoints, triggers, and tolerances. A web editor allowed growers to build and fork recipes visually, while the engine executed them against live device fleets. All runs were logged for comparison, enabling continuous improvement.
Results & benefits
Recipes became a shareable asset across sites, dramatically shortening the learning curve for new crops. Growers compared runs side-by-side and promoted the best-performing versions to production with confidence.
