Overview

In addition to climate change, shrinking arable land, and falling water levels, the global food crisis is escalating and exacerbated by labor shortages. Thanks to advances in technology, artificial intelligence is being called upon to provide tools to solve problems everywhere. Many pilots AI-based solutions have been deployed in the hope of increasing agricultural productivity while reducing costs and labor.

An European company specializing in vine analytics offering vineyard monitoring solutions to help winemakers gain actionable insights from data collected from their farmland, either by drone or, in this case of a more affordable solution, an existing farm truck equipped with cameras supported by an NVIDIA® Jetson-based Neousys NRU Series rugged edge computer.

Four GMSL2 cameras are mounted on the farm truck to effortlessly collect data from the vineyard as the truck traverses the vast area of the farm. The data is then uploaded to the cloud and analyzed using a special algorithm. This allows farmers to easily determine which parts of the grape crop are ready for harvest.

The system also provides farmers with an infrared map with the edge data collected for a comprehensive and detailed representation of crop maturity on their land, allowing them to optimize harvesting decisions and improve overall farm management.


Challenges

  • Budget constraints: For many automated or autonomous agricultural applications, how to achieve the application goals at a reduced cost is always a big challenge. Especially for the land area it must cover, the number of visual equipment needed would be significant.
  • Lack of power and internet connectivity: The lack of power and internet connectivity on a vast farmland is also one of the biggest challenges in agricultural applications. These challenges can also be solved by Neousys’ in-vehicle features.

Solution

In this story, our European customer is wise to use moving vehicles such as trucks as an application platform to eliminate the need to deploy cameras or sensors every 5 or 10 meters in the vast area of farmland. In this case, our customer's solution required only four cameras and an embedded computer to be installed on an agricultural truck.

The Neousys NRU-51V+ rugged edge AI computer, running on an NVIDIA® Jetson system-on-module (SOM) with 8GB/16GB LPDDR5 @ 3200 MHz and a wealth of I/Os for GNSS satellite positioning, GMSL2 cameras and flexible communications support, backed by a number of Neousys patented on-board robustness features, was selected for use on the farm truck to ensure stable operation of the AI application.