Collaborative Fruit Retrieval Using Intelligent Transportation

The value of soft fruit production in the UK is £724M (Agriculture_in_the_UK_DEFRA_2020), covering 11,000 hectares or nearly 7% of the land area used for horticultural crops in the UK. Domestic production of all fresh fruit as a percentage of total new supply was at 16% in 2019 and 2020, with soft fruit accounting for about a third of that.

Harvesting soft fruit is a backbreaking task, requiring skilled labourers to work long hours in polytunnels (the predominant setting for growing soft fruit in the UK). The decline in available seasonal labour due to Brexit restrictions on work permits, the devaluation of the pound and negative views towards the UK since the 2016 referendum has made it difficult for soft fruit farmers to find enough workers to harvest their crops. Covid-19 restrictions on travel and social distancing regulations further exacerbated the situation, which is additionally compounded by the 2022 rise in National Living Wage. While programmes (e.g. "Pick for Britian") attempted to recruit domestic workers for harvesting, a range of problems such as lack of skill, waning physical stamina and unwillingness/inability to reside on farms meant that very few British workers successfully filled the gaps (Life_without_EU_workers_The_Guardian_22_Nov_2021).

The Co-FRUIT project proposes an innovative approach to harvesting--using collaborative human-robot teams, where tasks are allocated to maximize efficiency: humans are assigned to tasks for which they are best-suited (e.g. picking berries) and robots are assigned to tasks for which they are best-suited (e.g. transporting berries from the polytunnels to the packing stations). This project merges and advances three key technologies to demonstrate this solution: (1) an accurate and adaptive model of individual human workers ensures that the right people are assigned the right tasks and robot "transporters" can predict when and where their services will be needed; (2) a responsive and safe methodology for autonomous navigation ensures that robots can move to where they need to be, when they need to be there, without crashing into each other, human workers, fruit plants or polytunnel infrastructure; and (3) a robust and affordable mobile platform ensures that trays of fruit are transported without damage, though the terrain may be muddy and uneven.

Co-FRUIT respects the contributions of skilled human workers while deploying a robot co-workforce and demonstrating a cost-effective and efficient collaborative harvesting solution.