14TSB_ESAP_AUTOPIC_35950-267128_Autonomous Strawberry Harvesting and Management Robot

Lead Research Organisation: Harper Adams University
Department Name: Engineering

Abstract

The AUTOPIC project addresses potential future shortages of fruit pickers and their increasing cost in the UK
and overseas by researching the essential elements of autonomous agile fruit picking vehicles. It will deploy
advanced sensing and pattern recognition and robotic technologies that direct picking heads, adapted to many
fruit or berry types, to inspect, pick and pack soft fruit in optimum condition and/or identify and remove pest
infected, rotten, over ripe or mouldy fruit delivering crop hygiene. AUTOPIC will develop robotic inspection and
manipulation systems for the care and harvesting of soft fruit; it will seek to produce systems with ultimate
operational costs comparable to human pickers but with reduction in errors, improvement in consistency,
allowing more predictable farm resource management. The project brings together state of the art knowledge
of innovative robotic arms, specialised fruit picking heads and the systems integration of multiple visual and
other sensors/cameras for positioning the vehicle as well as the necessary deep horticultural and agricultural
machinery knowledge to maintain the harvest potential of soft fruit, plants, trees or bushes.

Technical Summary

AUTOPIC is a multi disciplinary project aimed at mechanising the harvesting of soft fruit through the use of
autonomous vehicles and robotics. Partners include Harper Adams University, the Shadow Robot Company,
Interface Devices Limited and the National Physical Laboratory. The project is timely since the source of
migrant seasonal fruit pickers is no longer supported by the Seasonal Agricultural Workers Scheme and in
general migration is being discouraged by government policy. This has had the net effect of creating a crisis of
there not being sufficient workers to pick the soft fruit we take for granted in our supermarkets and potential
increased reliance on imports. Further, labour issues are not confined to the UK so that if the project is
successful there will be a significant export market for the project output. There are likely to be many benefits
from the use of the AUTOPIC autonomous vehicle and its robotics and we believe that the new technology will
be transformative for a new UK industry.

Planned Impact

The UK market for strawberries alone is worth around £300M per annum and fruit and vegetable retail
sales value is around £10 Billion. Overseas markets are of the order of 10 times this amount for Europe and 25
times for the Americas. The market is characterised by multiple producers in a market dominated by
supermarket buyers with growth limited by seasonal labour constraints. The wholesale price of fruit delivered
to supermarkets is in the range of £2,500 to £3,000 per tonne and over the 9 month picking season, a seasonal
picker picks around 2 tonnes of fruit per month at a labour cost of approx £1,380/month. (see Appendix A also)

The intended project outcome will ultimately replace the current labour-intensive fruit production system with
a technology-based solution, using high-tech UK manufacturing and providing new high-skilled employment
both in manufacture and technical support services to farmers in the UK, European Economic Area (EEA)
countries such as Spain and further afield with California being a specific target. Our goal for AUTOPIC is to
meet or improve on human fruit picking performance with greater consistency and a capability of 24/7 working
at lower operational cost while lessening the management overheads associated with seasonal labour.

Equally, for hygiene of the crop, AUTOPIC should to be able to detect and eliminate every infected, mouldy or
rotten fruit more efficiently than a skilled worker. The value of autonomous inspection and early treatment
cannot be over stated as with s.w.drosphila (SWD) a crop can rapidly become uneconomic to harvest.

The strategic route to market of the AUTOPIC system is to use this project to demonstrate that it can pick
strawberries efficiently and at a rate at least comparable to a human picker of 10 strawberries per minute and
that the vehicular platform can work up and down the rows without intervention almost continuously.

If the project can show this and show that the system can be produced in quantor a price/performance that
will reduce overall operational costs and improve quality, then the expected demand will create interest in
manufacture, as the system could quickly take a major proportion of the market and release the market from
its current labour constraints. It should also be stressed that the technologies developed by the project will be
highly transferable to other areas in agriculture (and industry) with significant Intellectual Property resulting.

Based on worker productivity figures, we estimate the replacement of 4 workers per AUTOPIC and thus a
minimum UK market size of 1200 units for strawberry production alone, to be followed by export to other fruitpicking
regions and the development of additional capabilities to pick other fruit. The target sale price of the
AUTOPIC autonomous vehicle unit is £40k with an operating value of £50k per annum (ie. less than the labour
replaced). This initial market represents a return on investment of more than a hundred times the project cost.

By the end of the project, with all the elements of the system working together, the project will deliver a clear
estimate of the cost of developing the concept to market, the cost of adapting the system to multiple fruit
types although it may identify factors to be overcome by further development with estimated costs to resolve.

Publications

10 25 50
 
Description A robotic strawberry harvesting machine has been developed which is capable of picking strawberries at the commercially acceptable rate of one ripe strawberry every 4 seconds. The research prototype system comprises high-speed actuation and grasping, machine vision, fruit ripeness sensing, and an integrated mobile base vehicle capable of navigating commercial strawberry stands.
Exploitation Route The research prototype system can be taken forward and developed, firstly into a commercial prototype and then into a manufactured product, through one or more of the following funding routes:
1) Late stage development collaborative R&D funding from channels such as Innovate UK; 2) Business Angel funding; 3) Industry funding; and 4) Private funding.
Other elements of research conducted during the project that require further research and new unknowns that have been discovered during the project can be taken forward through the following funding routes:
1) BBSRC and/or EPSRC funding; 2) The Wellcome Trust; 3) Innovate UK Early Stage funding.
Sectors Agriculture, Food and Drink

 
Description The commercial partners developed an understanding of the constraints that the agricultural environment imposes on designing robotic solutions for agricultural applications. The partners have achieved several successes in this project and are encouraged to seek further opportunities in agriculture. Successes include: 1) A novel system to solve the problem of picking strawberries at a commercially acceptable rate; 2) An integrated robotic base vehicle and strawberry recognition system.
First Year Of Impact 2016
Sector Agriculture, Food and Drink,Digital/Communication/Information Technologies (including Software)
Impact Types Economic,Policy & public services