How AI can help kill bugs.
As populations increase, so does the demand for food production. While geneticists are determining the best breeds of corn and conditions to produce the highest yields using artificial Intelligence (AI), pest hunters are using AI-based image recognition technology to identify and eradicate unwanted intruders.
According to the UN Food and Agricultural Organization, 20 to 40 per cent of all global crops are lost each year because plant pests and diseases aren’t managed properly. Saving crops and avoiding huge losses depends on speedy and accurate identification and solutions – which is where AI-based image recognition technology comes in.
This highly sophisticated system for identifying and treating infestations of pests depends on a simple smartphone. And while there are a few apps around that allow users to upload photos and information to gain expert advice, there aren’t many drawing on the benefits of AI and machine learning.
Pest control technicians, for example, are currently testing an Android mobile app to identify bugs and rodents. The technician simply takes a photo of the perpetrator and runs the app, which uses Google’s image classification and machine learning software to sift through images and identify the bug or rodent. After a positive identification, the app provides pre-populated solutions to help the technician decide treatment plans. One of its distinct advantages is that it tailors the treatment to the different pests. And of course, each time a technician takes a snap, it improves the recognition and classification capabilities of the machine learning algorithms.
Plantix is another mobile crop advisory app that also uses a smartphone picture and image recognition capabilities. It provides farmers instant diagnosis, and feedback and solutions, as well as offering preventative measures. It's being used widely in India, Brazil and North Africa.
With the help of an app with its solid base of machine learning, the pest control inspector or farmer can tailor treatment to the pest or problem. Korbinian Hartberger, one of the co-founders of the startup that developed Plantix, says that users are less likely to use the wrong pesticides after consulting the app.
Meanwhile, Resson in Canada and San Jose claim that their image recognition algorithms can detect and classify plant pests and disease more accurately than a trained human.
The machine learning factored into these apps means that the farming data is becoming progressively richer and more robust. Monsanto and Second Genome, for example, are collaborating on insect-control research based on big data analytics and machine learning.
When agricultural productivity and food quality are so important, data might well find new solutions to age-old problems and pave the way for the future.
How AI can help kill bugs.
Form submitted successfully
Is Australia ready to embrace the Internet of Things (IoT) revolution? Download our white paper summarising key findings and opportunities in the food industry.