Professor Simon Pearson, director of the Lincoln Institute for Agri-Food Technology (LIAT) and the 2022 winner of the RASE Science and Technology Award, shares insight into the key areas of research at LIAT and what’s on the horizon for use of robotics and AI in agriculture.
Robotics and artificial intelligence (AI) have a broad range of uses in agriculture. These include use for selective harvesting, systems for crop care and phenotyping to identify the highest yielding varieties in a crop.
Robotics for selective harvesting
LIAT has done a lot of work on developing robotic systems which use AI for selective harvesting of fruit, vegetables and salad crops. This has been partially driven by the labour crisis in the UK fresh produce industry. As a result, there is now a lot of funding and efforts going into robotic harvesting of products like soft fruit, tomatoes and mushrooms.Selective harvesting requires identification of fruit that is ready to be picked and a mechanism for picking without damaging the fruit. These robotic systems use AI to locate the fruit for harvesting and can autonomously move around the crop.
Robotics for crop care
Robotics and AI can be used to improve environmental best practice in farming. They can be used to complete many activities relating to growing a healthy and profitable crop, including weeding, fungal control and improving nitrogen use efficiency.Robotics and weed control
There are robotics and AI systems which can move around a crop and recognise a weed and kill it, either with a vastly reduced amount of herbicide compared to applying the treatment to the whole area, or no herbicide at all.LIAT is developing various technologies for weed control, such as robotic spraying machines and robotic steerage hoes. Weed control technologies are going to be really important in the future as herbicide availability goes down. Weeds’ resistance to herbicides is expected to increase, while the number of active ingredients available for use is likely to decline due to environmental pressures.
Robotics and fungal control
LIAT is working with Saga Robotics in the development and use of a robotics and AI system used to control powdery mildew in a variety of crop types. The robot moves autonomously around the crop, dosing it with UV as needed to control powdery mildew without the use of fungicides. The technology can be used in fields, polytunnels or vineyards.Robotics and nitrogen use efficiency
LIAT has started a large project on use of robotics and AI to improve nitrogen use efficiency in wheat.The project is farmer-led, meaning farmers are contributing to the research on how to reduce nitrogen use on wheat in a way that is practical and profitable. The project is using multi-sensors and machine learning to predict crop needs, so it can be applied selectively rather than at a generic application rate, changing the way nitrogen is used.
Robotics and phenotyping
LIAT is also using robotics and AI for crop phenotyping. This involves robots that move around crops and identify the highest yielding varieties at speed, using 5G, generating a vast amount of data to inform crop breeding decisions to optimise varieties for sustainable, profitable production.Explainable AI in farming
One outcome of the continued digitisation of agriculture is there are increasingly vast quantities of data coming from farming systems. The next step towards increasing farm profitability and sustainability by using robotics and AI is to develop systems which simulate the data and provide recommendations alongside logical reasoning for farmers to use as a tool in decision making. This is known as ‘explainable AI’.For example, a farmer deciding when to apply fertiliser could use explainable AI to take all the data on the crop’s development so far, plus the weather forecast to understand the risk of storms or heavy rain. The system would provide a suggested date for fertiliser application and a rational explanation for its recommendation.
Explainable AI could also be used to create rationalised recommendations on other decisions, such as how to change rotations, how to integrate regenerative practices and whether to use organic or inorganic nitrogen. Whatever the decision, it is about pulling data together to improve recommendations.
Co-creation of robotics and AI
AI based technologies are moving at pace. The next challenge is to make the digital transformation as straightforward as possible, so farmers can benefit with minimal inconvenience as new technologies find their place on-farm.Previously, some technologies have been pushed onto farmers which ultimately did not get adopted, as they did not work in practice. This is why co-creation of robotics and AI systems for agriculture is essential. Researchers at LIAT work closely with farmers, asking what they want, what is needed and how they would use different solutions, to ensure that everything is thought about in terms of design and useability. This ensures there is demand for the innovation and results in technologies being adopted and implemented faster and more effectively.
Many thanks to Professor Simon Pearson, the 2022 winner of the RASE Science and Technology Award, for contributing towards this blog. The 2023 RASE Science and Technology Award will be open for applications until 14th May 2023. Find out how to apply or nominate a colleague here.
