Harvester Case Study

OBJECTIVE

Demonstrate the capabilities and performance of the Farmwave Harvest Vision during soybeans harvest in some of the harshest harvest conditions in agriculture, everything from deployment and product durability to results based decisions making in the cab.   


APPROACH

Deploy custom built prototype units on various positions of the combine with specific focus on the header and rear discharge area. Visually capture, count, and return loss data in near real time via a prototype application based interface.   


RESULTS

As an aftermarket solution, a three unit vision kit suitable for a single combine can be deployed in less than 10 mins total. These units communicate in a closed loop infrastructure while returning results to an incab display allowing operators the ability to make decisions to optimize the machinery and reduce loss during the harvest season. Current functional AI on header loss is operating with about 85-90% accuracy. We are continuing to improve accuracy in the cleaning area models and full header automation. In an integrated system, the units will talk to each other to average a total machine loss in real-time while moving through the field. 


KEY LEARNINGS

During our time in Brazil we had the opportunity to validate Farmwave’s capability and scalability in more extreme farming conditions. In addition to the focus on harvesting, we pushed forward on demonstrating capabilities in planter and sprayer automation as well. Future work on the same AI will focus on the cleaning area, rear of machine, and even pre-header loss. The Farwmave technology is rapidly scalable and can be deployed in various ways for all products and brands within agriculture.