Cognitive Technologies develops 'world's first' model for autonomous ag machinery
Autonomous Vehicle Technology announces 2019 ACES Award Winner in Autonomy | Controllers
Cognitive Technologies has launched what it says is the world’s first industrial, universal control system model for autonomous agricultural machinery. The C2-A2 Agrodroid is an artificial brain equipped with a cradle—a universal device for fast connection with agricultural machinery such as harvesters, tractors, and sprayers. The cradle makes it possible to move the artificial brain from one machine to another without purchasing a new system each time. The C2-A2 is based on Cognitive’s Agro Pilot autonomous driving system for agricultural machinery, but the key new innovation is the state-of-the-art convolutional neural network (CNN) that was modified for agricultural purposes. It is the first product of its type based on standard Nvidia Jetson TX2 with deep neural networks.
Compared with existing autonomous driving systems for agricultural machinery based on GPS, the neural network-based C2-A2 opens a new class of systems that can protect equipment and people from all possible collisions. It can perform autonomously on plowed and unplowed ground, mowed and unmowed fields, along the rigs, and along the rolls. It can conduct autonomous U-turns at the end of the field, perform autonomous unloading of crops into a bunker or truck, and work at night and in poor weather conditions. The system uses only one (vs. three to four) video camera, allowing a 3-5x reduction of cost.
The final cost of the C2-A2 will be about $3000, about 1.5% of the cost of a combine harvester and about 3% of the cost of a tractor.
Click to view AVT’s 2019 ACES Award Winners best in class of