University of Lincoln using AI to grow millions of strawberries
One university is using AI to grow millions of strawberries to what it regards as “perfect ripeness”.
The Lincoln Institute for Agri-Food Technology (LIAT) at the University of Lincoln has joined forces with the London-based AI data engine, V7.
The partnership enables robots to analyse millions of strawberries throughout their growth cycle.
It monitors weather conditions and berry appearance to predict each plant’s “perfect” harvest date.
Strawberries
Through this new technology, robots will run through strawberry greenhouses in the north of England. They gather video footage that AI analyses as the robot moves along.
The AI can spot each strawberry, including the unripe green ones hidden amongst the foliage and every berry’s location.
This has enabled the LIAT team to predict the day of perfect ripeness. It can also estimate the total number of punnets a greenhouse might create six weeks ahead of current forecasting systems.
Artificial intelligence emulates human intelligence by learning from multiple examples.
Robots learning from human knowledge
For a robot to learn to detect the perfect strawberry across all angles, lighting conditions, berry varieties, and maturity states, it must learn from human knowledge.
This ‘knowledge’ is what the partners call training data. The process of humans ‘teaching’ AI by producing training data is annotation.
This involves drawing boxes or shapes around objects on software such as V7s and then classifying the item, such as a strawberry of a certain variety or ripeness level. The more training data, the better the algorithm.
Once it has collated sufficient training data, AI models can be trained that ingest this data and test themselves against un-seen images.
What has set V7’s technology apart is the ability for humans and AI models to work side-by-side on annotation challenges.
LIAT’s wheeled robot captures imagery of the strawberries and loads them onto the V7 platform.
That enables humans to begin distinguishing different strawberry types. After a few hundred examples are provided, the teaching is sped-up as early AI models begin pre-completing training data to cover cases where their confidence is high.
This data is combined with weather forecasts in algorithms to predict yield six weeks ahead of existing systems.
The impact of annotating using V7’s AI-assisted approach also helped increase the algorithm accuracy from an 85% human-only baseline to 95%.
Picking labour accounts for 40% of horticulture production costs. Much of the remaining costs derive from picking berries at the wrong time, leading to spoilage or sour under-ripeness.
Add value to foodchain
LIAT’s mission is to develop technologies that add value or solve challenges across the food chain.
In this project, they started with a small strawberry farm located near Riseholme Campus.
The team uses vehicle-mounted cameras to identify and count individual fruits and estimate their weight and maturity.
Other articles on That’s Farming:
