Exhibited at the recent RHS Chelsea Flower Show, where it was awarded a Silver Gilt medal, ‘RoboCrops’ demonstrated its ability to identify hidden signs of disease risk, environmental stress and growth performance in plants.
Created by the University of Lincoln and the Lincoln Institute for Agri-food Technology (LIAT), the technology uses an advanced robotic phenotyping system, called ‘PhenAIx’. Using a combination of artificial intelligence, imaging technology and robotics, it can spot issues in plants long before symptoms become visible to growers.
The researchers believe it could lead to further innovation that supports earlier crop disease detection, improved yields and climate resilient crops.
It has already attracted attention from policymakers, with the Mayor of London, Sadiq Khan, flagging how technologies like PhenAIx could one day help tackle wider food production and climate challenges.
Speaking further on the possibilities of this tech, Professor Elizabeth Sklar of the University of Lincoln told Food Manufacture: “The exhibit focuses on accelerating plant breeding, affording breeders the ability to continually develop new varieties. Given the needs of long-term food security and the threat posed by climate change, plants will need to survive in warmer climates and be more resilient to different types of disease.
“Manually taking measurements, which has been the case for a long time, is repetitive and time consuming. Through the adoption of AI and robotics, we aim to fill this gap and accelerate progress.”
The exhibit shows several examples of measurements which can be taken automatically, including one in which it counts the number of leaves on strawberry plants.
“We are looking to monitor how many leaves the plant has and asses what proportion of the plant’s volume is comprised of leaves versus fruit,” explained Sklar.
“This matters as the plant only has a fixed amount of energy to utilise, and if we are looking to maximise fruit production, we need to calculate the optimal proportion of leaves to do so. This is a challenge that breeders are trying to figure out for different varieties and assessing where we can strike a balance.”
In this particular scenario, the AI was trained to spot stresses associated with strawberry plants, for example powdery mildew or thrips. However, it’s entirely possible to teach it to spot other stresses associated with other plants, opening up a world of possibilities for farmers in the future.
Alongside helping to solve complex challenges within the food system, the researchers are also hoping the exhibit will inspire younger generations to consider a career in robotics, data science or agri-tech innovation.
