Share this post:
You may have heard of the Irish potato famine in the 1800s, but do you know what caused it?
Late blight, which can affect both potatoes and tomatoes, is one of the most common and devastating diseases in agriculture. It was the reason for the famine that caused the death or emigration of more than 2 million people from Ireland. It’s also responsible for the emergence of the field of plant pathology.
Humans are completely dependent upon plants, not only for food and medicine, but also for the air we breathe. That is why we established PlantLink, a research network in the area of plant sciences joining Lund University and the Swedish University of Agricultural Sciences (SLU) at Alnarp together in the south of Sweden. The organization’s objective is to improve crops and food products, as well as to enable the production of materials, medication and energy from plants in a sustainable way.
One of the regional project uses precision agriculture techniques to detect late blight.
Precision agriculture with IBM Watson
To combat late blight, conventional farmers typically spray their crops with fungicide as a preventative measure, even if late blight isn’t obviously present. This is costly and damaging to the environment. Organic farmers can’t do anything other than hope late blight doesn’t strike early in the season. If it does, the afflicted plants must be removed as soon as possible to prevent the spread of the disease.
Scientists at SLU were looking for an automated, efficient and environmentally friendly way to detect late blight early. IBM helped the scientists develop a prototype decision-support system that combines visual and near-infrared image analysis with climate data to predict how likely it is that late blight will strike.
The images are analyzed by Watson Image Recognition provisioned through IBM Cloud. Weather data is also incorporated through The Weather Company API. Scientists can see the data compiled in a GUI dashboard interface on a mobile app.
The app determines, based on the image analysis that Watson has done, the likelihood that late blight exists in the field and uses geo-referencing to pinpoint its location.
Scientists with decades of experience have trained Watson with a few thousand images and will continue the training to help Watson distinguish between late blight and other types of damage.
The precision agriculture solution has its roots in an IBM Design Thinking workshop. Aside from the scientists and IBM consultants, other constituents included potato growers and drone pilots.
A more objective view
Traditional and organic farmers can benefit from early detection of late blight. Traditional farmers spray when infection strikes and only the plants that are affected. Organic farmers get the early warning they need to remove sick plants before they infect the entire crop.
Additionally, because potatoes grown worldwide on a large scale are crossbred to introduce resistance to disease identified in nature, breeding companies are interested in efficient identification of new resistant lines. Plant breeding is the art and science of improving agricultural plants: plants might be selected for larger seeds; tastier fruits; and other valuable traits, such as, in this case, their resistance to late blight.
Estimation of late blight severity in the field for scientific research and breeding purposes has typically been a subjective process performed by humans using visual inspection on the ground. Using drone technology and imaging is a more objective and reliable way to cover larger areas and perform more frequent analyses.
According to the International Potato Center, potatoes are the third most important food crop globally after rice and wheat, as well as a fundamental element in the food security for millions of people across South America, Africa and Asia. Finding a way to detect late blight with precision agriculture techniques will have a significant impact on potato farming; and, thus, for the billion-plus people worldwide who enjoy their potatoes.
See what you can do with APIs delivered through the IBM Cloud.