Skip to main content

Intelligent decision from vineyard robots

Periodic Reporting for period 3 - VineScout (Intelligent decision from vineyard robots)

Reporting period: 2018-12-01 to 2019-11-30

Wine is a very important product in the European agricultural sector. To obtain good wines with high added value, it is essential to harvest grapes with high quality. In order to achieve this, the vines need to be monitored to assess the optimal growth and ripening state for harvesting. So far, the available monitoring methods were visual observation or random sampling of individual grapes. These methods do not provide completely reliable information on the crop. There are accurate methods such as using multispectral sensors mounted on drones or airplanes. However, either the low resolutions of the data obtained in this way or the low capacity of users to decide the acquisition time when data is most needed, limits their usefulness for wine growers. Vinescout robot is the evolution of Vinerobot which was created to tackle the aforementioned problem. As the robot collects the data in the field, it processes higher resolution images taken at less than one meter from which more accurate information can be gathered. The robot scouts the vineyard and takes data autonomously. It is equipped with two crop sensors which take canopy temperature and geo-referenced multispectral information of the vines. Maps on canopy temperature and vegetative growth are obtained by computing vegetative indices. These data are transferred to the wine grower that can monitor the needs and status of the vineyard and take decisions on the best moment to irrigate, apply treatments, or select different zones to harvest the grapes more efficiently. Vinescout uses electrical power from lithium batteries and solar panels
1. Optimization of mechanical design and external appearance. A new external cover has been designed for VineScout prototype, taking into account the harsh working conditions in the field, and the importance of an attractive look regarding marketing movements. Regarding traction abilities and power sufficiency, there have been some reinforcement in the mechanical torque and batteries.
2. Industrialization of internal electronics: Fail-safe capabilities and environmental endurance are being achieved by designing a suite of modular electronic blocks that have been fabricated, and tested. This rational design enhances the seamless integration of software and electromechanical devices.
3. Maps validation and sensing capabilities: Maps built by the robot need to deliver truthful data, so it can be statistically compared to alternative measurements. Getting truthful data means having taken data with the appropriate sensors, thus, sensors were tested in field, not only integrated in the robot, but also independently from it, to check the values given.
4. Software refinement, optimization, and market preparation: Efforts were done to improve software performance in terms of runtime velocity and fail-safe response. As new sensors have been incorporated, modifications in programming were necessary.
5. Construction of three prototypes with growing capabilities: An active iterative process is necessary to converge to the optimal solution, which for agricultural environments requires heavy testing in actual environments. For the first year of the project, a prototype was delivered on 21 August 2017 (first milestone achieved), and the second prototype is planned to be delivered earlier than the past year (planned for January).
6. Demonstration: the first Steering Week (SW1) of the project took place from 28 August 2017 to 1 September 2017. The Agronomy Day was on 30 August, during SW1, where the Consortium invited external people to see, touch, and even manage the robot, while they could make questions about either the robot or the project itself.
7. Market introduction tactics, end-user acceptability, and dissemination: The most important activities for the project regarding visibility is done during the Steering Weeks. In the Agronomy Day, people with different backgrounds can give their opinion and tell the Consortium their suggestions, which are extremely valuable for the success of the project, as people can talk about their experiences with other robots.
8. Attendance to trade shows on robotics and agricultural equipment shows that interest in agricultural robotics is high, but competition is not too intense. The service robot market continues growing as well as the agricultural robots market. The Agronomy Day revealed that the main concern is the retail price of the robot, being reliability the second.
The potential direct competitors found during the first year of the project in conferences, tradeshows, commercial catalogues, scientific journals, and dissemination magazines in the farm machinery or wine production sectors are: Romovi (with a subjective assessment of its potential risk to the VineScout project of low risk, because it only focuses on the specific environment of terraces with only one row, envisioning a localization solution based upon a full-custom wireless positioning system), Vinobot (with a low risk as the objective of this project is plant phenotyping, and they are not looking at a commercial solution), Vitibot (this one with a medium risk. It is autonomous, solar and electric, which are features that compete with VineScout), and TED (with medium risk, TED probably represents the most direct competitor. It is made by a European company that just started selling agricultural robots. This platform is nice and robust to withstand farm environments, and can operate at 4 km/h). The fact that the robotics market is beginning to bloom together with the advent of a multiplicity of crop sensors at competitive cost is rising the interest of viticulturist and vineyard-owning wineries to implement digital technologies in their management strategies.

-Societal impact. An important side effect of the successful introduction of robotics in European vineyards is the attraction that new technologies pose to young farmers. The average age of farming population is currently near retirement age, with very few growers under 35. The lure of electronics and automation will likely help to counterweight the negative effect of an aging population in agriculture.
-Economical-Social impact. Socio-economic studies have firmly proved that automation creates more jobs for the overall economy than it eliminates. It reduces the number of low-skill repetitive tasks while creating a whole set of higher skill jobs in relation to manufacturing, data support for decision-making, service, and financial industries, with entirely new industries, in many cases, leading to greater economic prosperity.
-Environmental impact. The deployment of agricultural robots entirely run by renewable energy, that is, rechargeable batteries and solar panels, implies a paradigm change in agricultural machinery, where farm vehicles have been only powered by diesel fuel over the last 100 years. VineScout robot is meant to be constructed in its final prototype with recyclable materials in most of its parts. Both approaches will make it one of the greenest solutions for food production involving autonomous self-propelled machines.
-Technological impact. A mapping agent for a decision support tool that requires reliable and quite precise positioning information offers an ideal milieu for promoting the new satellite positioning system GALILEO. The economical efforts made by the EU to set up the European GNSS will revert to end-users with better management decisions and in consequence higher profits.
Participants in the VineScout II Agronomy Day - 29 august 2018