Feeding the world

Analyzing growing conditions traditionally takes a month or more. This delay can lead to the planting of unsuitable crops, improper application of amendments, or low yields. In subsistence communities it can also mean the difference between having food at harvest time or not.

Coral-enabled sensors with neural networks trained on soil data can offer near real-time analysis of soil conditions. This allows for faster corrective action and more appropriate planting.

For smallholder farms, the ability to sort product by grades leads to higher price per bushel and faster time to market. Traditionally done by hand, the process can be slow and consume resources that might be used for other work.

Coral boards running object classification models form the basis for cheaper modular sorters that can help farmers easily grade their products and get to market quick.

Accurate diagnosis of plant health, including disease, pests, and environmental stress, is the first step toward improving growing conditions.

Segmentation models can help to separate leaves and plant features from complex backgrounds. When combined with classification models and run on the Coral platform, farmers can create tools that allow for real time diagnosis of diseases in the field.

Aggregating this data over a growing season and across fields improves the models, which then improves future yields and lowers disease risk.

Precision agriculture requires that you know the position, condition, and progression of every plant to provide the right amendments at the right time. Whether on large or small farms, tracking this information can be difficult.

With condition monitoring tools built with Coral local AI and models for identification and classification, farmers can find plants and assess their condition to deliver the right amendments in a targeted way. The precise application of pesticides, fertilizers, and herbicides reduces runoff and adverse effects as well.

In vertical farms where density and safety concerns prevent the use of broadcast techniques, tray-management machines built using Coral can operate independently while sharing knowledge to improve overall plant health.