HarvaAI offers autonomous harvesting robots for commercial strawberry farms, using dual‑camera vision and force‑feedback grippers to detect and gently pick ripe berries with high accuracy. The system navigates rows via GPS‑augmented SLAM, streams detailed harvest data to a cloud dashboard, and integrates with existing farm‑management platforms to lower labor costs and increase throughput for medium‑to‑large growers.
Funding
Funding not disclosed
Founders
Product
Problem
Commercial strawberry farms face high labor costs and seasonal worker shortages, leading to inconsistent harvest timing and increased fruit bruising from manual picking. The lack of scalable automation limits yield optimization and profitability for growers operating at large scale.
Solution
HarvaAI delivers autonomous harvesting robots that combine high‑resolution computer‑vision pipelines with machine‑learning models to locate ripe strawberries in real time. A six‑axis robotic arm equipped with force‑feedback control gently extracts each berry, minimizing damage while maintaining a steady pick rate. The system operates outdoors, navigating rows using GPS‑augmented SLAM and adapting to variable lighting conditions. Collected data on fruit size, location, and yield are streamed to a cloud analytics platform, enabling growers to monitor performance and integrate results with existing farm‑management software. By replacing manual labor with a continuously operating robot, growers can achieve higher throughput and lower per‑kilogram harvesting costs.
Target Audience
Primary customers are medium‑to‑large commercial strawberry growers and agribusinesses that manage acreage exceeding 20 ha and seek to automate labor‑intensive harvest operations.
Features
- Dual‑camera vision stack (RGB + hyperspectral) with deep‑learning inference optimized for edge GPUs, achieving >95 % detection accuracy under sunlight and shade.
- Force‑controlled, soft‑grip end‑effector that applies <0.5 N pressure to prevent berry bruising.
- Autonomous navigation using RTK‑GPS and LiDAR‑based SLAM for row following and obstacle avoidance.
- Real‑time edge analytics that tag each harvested berry with GPS coordinates, size, and ripeness metrics.
- Cloud‑native dashboard with API hooks for integration into existing precision‑ag platforms (e.g., FarmLogs, Granular).
- Modular hardware design allowing quick retrofitting to existing planting layouts and easy part replacement.
- Weather‑sealed chassis rated to IP66 for operation in rain, dust, and temperature extremes.
- Remote diagnostics and OTA software updates to keep vision models and control algorithms current.