European Commission logo
español español
CORDIS - Resultados de investigaciones de la UE
CORDIS

The first high-performance robotic system for automated harvesting of vegetables greenhouses

Periodic Reporting for period 1 - GRoW (The first high-performance robotic system for automated harvesting of vegetables greenhouses)

Período documentado: 2018-07-01 hasta 2018-12-31

The growing world population is calling for a new wave of innovations in agriculture, with the objective to produce more food in a sustainable way.
Thus, farmers all across the world are under increasing pressure to feed the growing world population, which will need 50% more food to be produced within 2030.
The increasing demand of fresh products also reflects into the growing adoption of greenhouse farming, which allows to extend the production of fruits and vegetables throughout the year.
However, to keep up with the demand, farming (and greenhouse farming) must increase the level of automation in all field operations, allowing effective and continuous planting and harvesting of vegetables.
Robotics is seen as a key resource to streamline farming practices, but issues as speed, accuracy, autonomy and cost are keeping most solutions still at the demonstrator stage.
Although several ongoing attempts are being made to automate greenhouse farming all over the world, none of them is being successfully applied to harvesting of soft vegetables, due to the intrinsic issues of automatically recognizing ripe fruits (artificial vision) and picking them with the necessary repetition rate and accuracy.
The main problems are the capacity to discriminate ripe from unripe fruits through fast and accurate machine vision systems and then to pick soft and pliable fruits without damaging them.
Currently, no solution on the market exists that is able to reach such performances, which in turn would open the gates for a 1B€ market by 2020
MetoMotion, an Israeli start-up, has developed a multi-purpose robotic system for labor demanding tasks in greenhouses, called GRoW (Greenhouse Robotic Worker).
GRoW is a technology platform open to host solutions for pruning, pollinating, de-leafing and harvesting of different crops, like eggplants, cucumbers, peppers, tomatoes, just to name but a few.
For the first time in robotics for agriculture, GRoW delivers a fully autonomous solution that can be easily integrated into current greenhouse practices, from the identification of the ripe fruit up to picking and packing it into the box.
By doing so, it also generates and processes a continuous data flow relating to the plant growth and plant stress, thus assisting the grower through decision making and planning.
Thanks to its proprietary advanced vision system and computational algorithms, flexible picking system and autonomous driving, GRoW is poised to fully meet the market requirements, proving to be accurate (error rate < 5 %), fast (harvesting rate > 720 kg/hour), easy to use and cost-effective, with an expected payback time for customers point of about 3 years. Furthermore, GRoW is a flexible technological platform adaptable to a variety of crops and agricultural operations.
As first application the harvesting of tomato in high-tech greenhouses has been chosen because most grown crop worldwide (35% of all vegetables).
The objectives of the feasibility study was linked with:
1. The understanding of specific technological criteria to be achieved by GRoW in order to be compliant with the growers' requirements.
2. The definition of the most suitable commercialisation strategy and entry points.
During the Phase 1 project, Metomotion has ascertained all main performance requirements as to the harvest speed, harvest accuracy, harvest cost per Hectare.

Starting from these data, the preliminary laboratory scale design of the GRoW system was re-engineered, by adding additional components and tools to increase harvest speed and to adopt standardized components to make GRoW compatible with any greenhouse facility in the EU.

Moreover, the concept design of the automated boxing system was conceived and validated by users.

A preliminary selection of suppliers of industrialized and certified components was done.

The Bills of Materials for the commercial scale was compiled thus enabling the COGs projection and thus a more accurate Profit&Loss calculations.

Detailed exploration of the markets and their economic potentials was done, finding out that the Agricultural Robotics segment is one of most pushing within the global farming scenario, experiencing 21% CAGR.

In this context, Metomotion has carried out a detailed enquiry over the most suitable early adopters of GRoW in European and extra-EU markets, slightly changing the initial commercialization strategy.

In fact, main findings are related to the fact that several EU countries are more prone to adopt automated solutions (like the Netherlands, Belgium and France) because already widely adopting high-tech solutions in the greenhouse management.

Whilst countries like Spain and Italy, even if producing large volumes of greenhouse tomatoes, are not technologically ready to adopt GRoW.

The business model of GRoW was extensively validated by users and other stakeholders, highlighting their interest in the fastest possible introduction of GRoW to the market.

Overall, the feasibility study has enabled the achievement of the commercial scale GRoW device and the finalization of a strong business plan.

The results were disseminated through participation at events of the sector world wide. As a direct consequence, Metomotion has received over 30 expression of interest to join the development and the sales of GRoW in different countries on the global scale.
Currently, there is no industrialized solution for the automated harvesting and boxing of the tomatoes.

This is due to the complexity of the harvesting operation requiring precision and delicate touch, in order not to damage the plant, nor the fruit.

Metomotion's team, after a decade of multi-disciplinary research and development, has achieved to develop different proprietary components jointly enabling an accurate and fast harvesting of the tomatoes:

- proprietary harvest tool

- proprietary machine vision algorithms
Working GRoW prototype