Machine Learning algorithms are beginning to revolutionize modern agriculture. Enabling farmers to combat pests and diseases in real time, the technology is improving crop production and profits, while reducing waste, greenhouse gas emissions, and pesticide use.
Around 6% of the world’s CO2 emissions come from farming. And every year, up to 40% of crops are lost due to pests and disease. For farmers already operating in a low-margin industry, critical resources wasted on unused crops makes surviving, let alone thriving, that much harder.
But a new AI-powered platform from startup Fermata offers farmers a way to mitigate the impact of pests and crop diseases while also making farming more sustainable, and worker-friendly.
The new ML-powered computer vision system, named Croptimus, continuously scans crops 24/7. When it detects pests or early signs of crop disease, the platform immediately alerts farmers, enabling them to rush resources to impacted crops and keep the threat localized.
The platform is trained on high-quality data, enabling its computer vision software to accurately differentiate between healthy and at-risk crops and quickly identify pests or diseases. The system gives farmers real-time analytics and provides annotated, 360-degree, augmented reality maps showing up-to-the-minute reports on the health of crops.
Fermata trains its models using PyTorch with NVIDIA cuDNN on on-prem devices. For inferencing, it uses a combination of both cloud and on-prem compute, including NVIDIA T4 GPUs running on AWS cloud, and NVIDIA Jetson Nano code optimized with NVIDIA TensorRT for accelerated performance.
The Croptimus system is deployable in a variety of configurations in large greenhouses and in outdoor farms. Cameras to scan and analyze crops can be mounted on tall poles, greenhouse ceilings, integrated into aerial drones, or attached to mobile robots that regularly traverse rows of crops.
The AI-powered model doesn’t replace a farmer’s existing routines—it augments traditional workflows and makes them more targeted.
For instance, farms typically rely on trained scouters to manually inspect crops. But high-quality scouters are increasingly in short supply, and when they are available, they’re expensive. Additionally, humans get tired or make mistakes. When scouters fail to detect crop diseases or pests, they can quickly proliferate, which can lead to widespread crop spoilage and waste.
By contrast, AI systems like Croptimus are always scanning for issues. When the system flags a potential problem, it sends an alert to farmers who can then direct humans to inspect crops and, when necessary, intervene before pests or diseases can spread. Early intervention reduces crop loss and helps farmers use human labor in a more targeted way.
Another upside to the system is that farms can detect and mitigate pests and diseases early and have less need for pesticides. Which not only saves farmers money, but also reduces the negative effects of pesticides leaking into the environment.
Read the full story about this ML-powered farming system in Future Farming.
