A wide-angle robotic vision camera that sees like a human combining stereo depth, high-resolution color, and on-device AI, all powered and connected via PoE.
Bushels per acre of harvest loss can be reduced with the Harvest Vision AI solution.
Harvest loss measurement counts per acre without stopping.
Farmwave is tackling one of agriculture’s most persistent challenges: grain loss during harvest. Traditional methods for measuring loss, like stopping combines to manually count kernels, are slow, inaccurate, and often skipped altogether. This means farmers are unknowingly leaving significant revenue in the field.
Accurately measuring harvest loss in real-time poses serious technical hurdles:
Farmwave’s Harvest Vision system mounts Luxonis OAK‑D‑PoE cameras on the combine’s header and rear. These cameras capture images every few seconds and run kernel-counting algorithms directly on-device enabling real-time, in-cab feedback on where and how much grain is being lost, without stopping the machine or needing cloud access.
The OAK‑D‑PoE offered Farmwave the perfect mix of performance and practicality:
Onboard AI inference ensures fast, reliable detection even in remote areas.
Power-over-Ethernet simplifies install with one cable for power, data, and video.
IP67-rated durability handles dust, vibration, and extreme field conditions.
A compact form factor made it easy to retrofit across a variety of combine models.
Farmwave also benefited from Luxonis’ fast integration and ease of use, making their solution scalable to farms of all sizes without requiring technical expertise from the end user.
A wide-angle robotic vision camera that sees like a human combining stereo depth, high-resolution color, and on-device AI, all powered and connected via PoE.
The Farmwave Harvest Vision system uses camera sensors to measure harvest loss by taking pictures and counting grain in real-time without ever stopping the combine.