To test the quality of their soil and water, farmers typically send samples out to labs. The process can be time-consuming -- results may be outdated by the time they're received. And for small farmers, it can be a costly process.
To address this challenge, IBM researchers in Brazil have developed an AI-powered prototype to help farmers easily conduct their own chemical analyses -- on location, in real-time. The prototype "could revolutionize digital agriculture and environmental testing," IBM's Mathias Steiner wrote in a blog post. Given that family farms produce about 80 percent of the world's food, low-cost tools that bring AI to small farms could indeed have a major impact.
The AgroPad is a paper device about the size of a business card. It has a microfluidics chip inside that can perform a chemical analysis of a water or soil sample in less than 10 seconds. A farmer simply puts his sample on one side of the card, and on the other side, a set of circles provides colorimetric test results.
Using a dedicated smartphone app, the farmer can receive immediate, precise results. The app uses machine vision to translate the color composition and intensity into chemical concentrations, with results more reliable than those that rely on human vision.
The current prototype measures pH, nitrogen dioxide, aluminum, magnesium and chlorine, though the research team is working on extending the library of chemical indicators. AgroPads could be personalized based on the needs of the individual farmer.
Once the test results are in, the data can be streamed to the cloud and labeled with a digital tag to mark the time and location of the analysis. Results for millions of individual tests could be stored. "This is an important feature for monitoring, for example, the change in fertilizer concentration in a particular region throughout the year," Steiner wrote.
Artificial intelligence is already a growing part of the agricultural industry. John Deere recently acquired a Silicon Valley-based AI startup called Blue River to incorporate machine learning, deep learning, and robotics into its farm equipment. Along with crop analysis, there are a number of other applications for AI in farming. For instance, robotics can help improve crop harvesting, while predictive analytics can help farmers anticipate weather conditions that will impact crops.