MENU

Development of ict fusion smart farm technology for intelligent production and distribution of orange

This project aims to develop low cost unmanned aerial vehicles (uav) and sorting machinery -supplemented by vis + nir multi- hyperspectral cameras- and cloud based decision support systems to help citrus growers and managers to take decisions in real time and easily be customized to serve to the nee

Oranges represent one of the most dynamic agriculture sector where research and development work have done possible the continuous production of new oranges varieties - for fresh consumption or for juice production, the creation of many companies and laboratories which provide advising services for crops management (i.e. nutrition, irrigation, pest and diseases…), thus deciding the most efficient and economically management during each critical moment of the citrus production cycle. Conventional methods to tackle problems such as yield estimation, determination of the optimum harvest time, variety authentication, fruit quality – at pre-harvest and post-harvest- stages are often complicated and destructive, cannot be run in large scale and are still far away from being implemented to large volumes of product or even better to individual piece of high quality fruits -either on the tree or over a conveyor belt or on the shelf of a supermarket. In the last 30 years, researches has been carried out in non-destructive sensors for pre-harvest and post-harvest quality and safety inspection of fruits. Pre-harvest management of citrus fruits have benefited from the advancements in remote sensing (RS): satellite imagery being used in Precision Agriculture applications for the management of spatial and temporal variability of the fields using ICT (Information, Computers and Technology). Precision agriculture’s professionals have been using soil maps and satellite images for a long time. However, despite the broad expansion of Precision agriculture, small and medium farms as well as managers need more affordable data acquisition systems -in addition to the remote sensing satellite imagery- and providing more useful information (i.e. finer spatial resolution that the traditional > 30 m^2 reached by satellite imagery like LANDSAT, for example). There is no doubt that recent research and technology developments has sufficiently demonstrated that in the near future farmers and crop managers will use small unmanned aerial vehicles or drone (SUAVs). SUAVs can carry different types of imaging sensors for collecting field data. Almost all applications of SUAVS in crops management are being developed on the observation of crops in distinct areas of the electromagnetic spectrum, based in the experience gained with RS. Thus, the above mentioned works are commonly done using the visible, the short wave near infrared and the thermal infrared portions of the spectrum. In our opinion, the research and technological developments which are being done using SUAVs and cameras can still benefit from further optimization and from being also intelligently combined with information coming from the so called “full near infrared region” (780nm-2500nm). Near Infrared Spectroscopy (NIRS) have demonstrated to be ideally suited to analyse samples in real time (seconds or milliseconds) and non-destructive manner, for multiple analytical constituents and samples (soil, leaves, fruits and vegetables, cereals and other grains). Our approach will involve the development of a low cost compact multispectral imaging system (MIS), by means of selecting key wavelengths from the full VIS + NIR spectral region (400-2500 nm). This reduction of the data dimensionality will make the data analysis faster thus ultimately will reduce the cost of instrumentation and analysis. The MIS developed will be adapted for “on site” quality inspection and quantitative determination of strategic quality indicators and optical labelling of oranges, at the packaging/sorting citrus lines. Furthermore, the MIS will be also able to fly on board on a low cost SUAV adapted for the purpose, over a citrus crop field to collect hyperspectral images of the desired sites and to provide qualitative and quantitative physiological and biochemical features of the oranges, soils and leaves. Furthermore, in order to ensure the fast update of the projects outputs for producers and processors, a set of programmed utilities based on novel mathematical and statistical algorithms and programs will be developed. Robust and fast computing of raw and predicted data from MIS images will be related to the information gathered from MIS for developing a Decision Support System (DSS) based in “real time “computation using the CLOUD. The DSS will help citrus producers and managers to allow taking “real-time” decision using commonly used devices as mobile and or tablets. Utilities in the APPs format will also settled for real time access for consumers to useful quantitative and qualitative information from the fruits for those are willing to pay an extra price. Some of the quality parameters that will be considered as “objective of study” include: crop yield, disease detection and pest management, optimal harvest time, variety recognition and qualitative indicators (sugar, acidity, pulp yield, juice yield), among others.
Start date: 
01-12-2016
Project Duration: 
36months
Project costs: 
2 050 000.00€
Technological Area: 
AGROFOOD TECHNOLOGY
Market Area: 

Raising the productivity and competitiveness of European businesses through technology. Boosting national economies on the international market, and strengthening the basis for sustainable prosperity and employment.