Christophe GUIZARD IRSTEA Sensors to monitor vineyard Tendencies UMR ITAP Montpellier christophe.guizard@irstea.fr www.irstea.fr InnoVine Final symposium Toulouse, 16-17 of November 2016
InnoVine Final symposium Toulouse, 16-17 of November 2016 PRESENTATION PLAN What is the Need? Emerging technologies Examples Conclusion
What is the Need? (Not exhaustive) InnoVine Final symposium Toulouse, 16-17 of November 2016
Farm Ecosystem Farm size World market Human resources An ecosystem very complex Lot of actors Lot of rules Lot of exchanges New business New demands InnoVine Final symposium Toulouse, 16-17 of November 2016
InnoVine Final symposium Toulouse, 16-17 of November 2016 What We Need New tools? YES, but do do what? Simplify the farmer life Meet regulatory and business requirements How to do? Use new technologies to have an objective information for better decision Technologies must be Easy to use, to install, to maintain Efficient Cost effective With a fast Return Of Investment
InnoVine Final symposium Toulouse, 16-17 of November 2016 Future Look like? Statistiques
Emerging technologies (Not exhaustive list) InnoVine Final symposium Toulouse, 16-17 of November 2016
InnoVine Final symposium Toulouse, 16-17 of November 2016 Emerging trends Emerging trends on sensors 1. Optical sensors, lasers... Contactless, non-contaminating, Real time Can «reproduce» human feeling Able to "see" the invisible (IR, NIR, UV...) 2. New information vectors Smartphones UAV, satellites IOT(sensors networks) New services Ressources management Cartography (biomass, soil, nitrogen...) Traceability Etc. System Heart
InnoVine Final symposium Toulouse, 16-17 of November 2016 Emerging trends Farm Lab Observation SENSOR Plant vigor Pedestrian, Satellite tools Observation Soil Physicochemical components Plant stress Etc. Robots UAV Observation Action Weeding Spraying Tillage
Examples (Not exhaustive list) InnoVine Final symposium Toulouse, 16-17 of November 2016
InnoVine Final symposium Toulouse, 16-17 of November 2016 Place of sensors Knowledge and Understanding Diseases Maturity Load Diseases, Water Stress Traceability, Risk management, Resource management, Production management Berry Cluster Vine Vine plot Farm Territory Acting & Controling Pruning, Leafing, Treatments... Precision spraying, Maintenance inter row Harvesting, Logistics...
InnoVine Final symposium Toulouse, 16-17 of November 2016 Pests control BEECAM Pheromone dispenser for IPM (Canada) ZTRAP : Insect trap Orchards (http://spensatech.com) Src: http://www.advansee.com/bee_home SRC : http://www.wfs.org/blogs/daniel-castro/thirty-plus-ways-internet-things-changing-world SRC : http://www.iotjournal.com/
Pedestrian optical sensors GREENSEEKER Trimble QUALIRIS GRAPPE Sodimel (Prototype) SPECTRON Pellenc DUALEX Force-A SMARTGRAPPE (prototype) MULTIPLEX Force-A vitisflower InnoVine Final symposium Toulouse, 16-17 of November 2016
Example Lab on the chip Allergy Test with a smartphone Test counterfeit (ie pills) Cholesterol test Colorimetric test Source :UCLA Henry Samueli School of Engineering and Applied Science MIT Techology Review (Photo courtesy of Stratio) Cornell University InnoVine Final symposium Toulouse, 16-17 of November 2016
Embedded solutions InnoVine Final symposium Toulouse, 16-17 of November 2016
Embedded solutions InnoVine Final symposium Toulouse, 16-17 of November 2016
Vegetative Canopy measurement InnoVine Final symposium Toulouse, 16-17 of November 2016
InnoVine Final symposium Toulouse, 16-17 of November 2016 Embedded solutions Solution for traceability for spraying system Src : IRSTEA
InnoVine Final symposium Toulouse, 16-17 of November 2016 Wireless network sensors Lot of solutions on the market System TracoVino Humidity Soil Humidity Temperature Solar Radiation Can be extended : Additional sensors for Leaf Wetness, Soil PH Values and Nutrient Levels
InnoVine Final symposium Toulouse, 16-17 of November 2016 UAV / Satellites «Aerial» imagery for vineyard management Oenoview (ICV) resolution <= 2,5 m UAV example «TerraDrone» For a centimetric precision New European satellites "Sentinels provide "free" images with coverage frequency of 5 days (forecast), and metric resolution SEQUIOA sensor Parrot
New sensors for new services Conclusion Changing practices, methods, and often the organization Next revolution? New IOT are everywhere «from the vineyard to the bottle» A gateway to robotics in the vineyard Towards true collaboration Human & Machine (Smart Machines) New challenges for Research Improve sensors performance Build new agricultural knowledge with new data from new technologies Ensure safety of users data InnoVine Final symposium Toulouse, 16-17 of November 2016
Mobility changes everything in business, but it doesn t replace everything. Thanks for your attention Contact Christophe.guizard@irstea.fr www.irstea.fr La France Agricole InnoVine Final symposium Toulouse, 16-17 of November 2016
InnoVine WP5 Toulouse, 16 & 17 of November 2016 Christophe GUIZARD IRSTEA Smartgrappe A smartphone in a vineyard
Smartgrappe concept Support + Mobile + software = SMARTGRAPPE P a t e n t e d a p p a r a t u s L o w c o s t s u p p o r t Controlled lighting conditions Constant distance to the scene Reference targets inside Work on any smartphone (IOS, Android, Win10 ) S o f t w a r e Berries size, color and surface defect Number of berries / surface unit 2
Cloud solution Statistiques Applications 3
Robust detection Blooming Grappe variety Berries Size 4
Results Images Samples 5
Conclusion Smartgrappe A simple and low cost solution to help you to manage your vineyard with your smartphone 6
THANKS! See you later christophe.guizard@irstea.fr 7
GRAPEVINE OPTICAL SENSORS AND DECISION SUPPORT TOOLS FOR PRECISION VITICULTURE NAÏMA BEN GHOZLEN JEAN-LUC AYRAL INNOVINE International Symposium 17 November 2016 1
ACTIVITIES Expertise in optical technologies: fluorescence and plant-light interactions Manufacturing of proximal plant sensors DUALEX MULTIPLEX Sales of plant sensors to Research Institutes Sales of diagnostic services to wine growers
A FULL RANGE OF VINEYARD DIAGNOSIS Vineyard Management Actions Spring FA-vigor ----------- Leaves Vigor - Physiological balance In-season Nitrogen fertilization Summer Autumn FA-vendange ----------- Grapes Maturity monitoring Harvest Quality analysis Zones optimization Selective harvest Homogeneous batches Winter FA-wood ------------ Pruning woods Vigor Pruning, fertilization, grass cover
DATA PROCESSING AND DISPLAY Data display on 2 complementary interfaces SENSOR Capteur : Real-time Indices en vine temps indices réel FA-SERVER Mapping and dynamic zoning 4
GRAPEVINE VIGOR MAPPING FA-vigor Foliar density map Diagnostic map Vigor map (NBI index) Cru classé 1855, appellation Margaux Display the balanced zones and the origins of the unbalances (N carency, )
MASTERING GRAPE MATURITY FA-vendange Anthocyanin evolution Measurement of anthocyanin content Plots classification Intermediate High potential Low potential Optimize harvest logistics and understand wine phenolic potential
MASTERING GRAPE MATURITY FA-vendange Zone Anthocyanin mapping mg/l Anthocyanin zoning along the rows Cru classé 1855, appellation Margaux Anthocyanin level % area Average value High 50 4170 mg/l Low 50 3655 mg/l
Anthocyanins MASTERING GRAPE MATURITY FA-vendange Zone «Premium» Vigor Lower Optimum Higher Combine the leaf vigor and grapes quality maps to identify zones for premium wines Higher Lower Cru classé 1855, appellation Margaux
THANK YOU FOR YOUR ATTENTION INNOVINE International Symposium
InnoVine Final symposium Toulouse, 16-17 of November 2016 New tools and applications for phenotyping and vineyard monitoring Javier Tardaguila and Maria P. Diago
New grapegrowers need more data
From visual assessment to data-driven viticulture.
Vineyard monitoring and plant phenotyping
New technologies for phenotyping Hyperspectral imaging Spectroscopy Computer vision
Hyperspectral imaging, a powerful tool for phenotyping VIS-NIR Hyperspectral camera
Classification of cultivars and clones using hyperspectral imaging Diago, M.P., Fernandes, A.M., Millan, B., Tardaguila, J., Melo-Pinto, P 2013. Computers and Electronics in Agriculture 2013 Fernandes, A, Melo-Pinto Millan, B., Tardaguila, J. Diago, M.P., Journal of Agricultural Science 2015
Using hyperspectral imaging to fingerprint individual anthocyanins Predictive models based on modified partial least squares (MPLS) were built for 14 individual anthocyanins with coefficients of determination of cross validation (R 2 cv) ranging from 0.70 to 0.93
New tools to non-destructively estimate the number of flowers and berries using computer vision
Computer vision in viticulture
Estimation of berry number per cluster by computer visión in the field Original image Extraction of berry candidates Final result after false positive filtering
Actual number of berries per cluster Estimation of berry number per cluster by computer vision 300 250 y = 1.4112x + 15.114 R² = 0.81** 200 150 100 50 0 0 20 40 60 80 100 120 140 160 180 Number of berries visible in the image
Fruit-set assessment by computer vision Goal: To investigate differences on fruit-sett among cultivars Non-invasive, inexpensive, image analysis as powerful tool to estimate fruit set
On-the-go assessment of canopy porosity and exposed clusters in precision viticulture
Multi-sensor mobile platform was developed Televitis mobile lab Speed: 5 km/h Illumination for night time image acquisition GPS for data georeferencing Inductive sensor and controller for autonomous camera triggering
Image aquisition and processing 1.Image obtained on-the-go using a quad 2.Image segmentation Porosity (White) Clusters Leaves
% porosity from image analysis on-the-go On-the-go assessment of canopy porosity 60 50 y = 0.7427x + 8.1439 R² = 0.91 40 30 20 10 0 0 10 20 30 40 50 60 % porosity with Point Quadrat (reference value)
On-the-go assessment of wood pruning using machine vision
On-the-go imagen adquisition and processing Original image Segmented image
Pixel number of wood On-the-go assessment of pruning weight 600000 500000 y = 431775x + 24408 R² = 0.92** 400000 300000 200000 100000 0 0,0 0,2 0,4 0,6 0,8 1,0 1,2 Pruning weight (Kg/vine)
Mapping Embedded sensors with GPS for data georeferencing Pruning weight (kg/vine)
Javier Tardaguila María Paz Diago Juan Fernández Borja Millán Salvador Gutiérrez televitis.unirioja.es
Thank you for your attention
L innovation au service de la Protection du Vivant Spatial Presentation of WINEO s products - Using Very high resolution remotely sensed images for precision viticulture Environnement Développement durable Innovation technologique
WINEO products NOVELTIS proposes the acquisition of suitable remotly sensed images over your vineyard; NOVELTIS proposes to digitize vineyard boundaries. High resolution images Mosaic over the vineyard and digital boundaries of the parcels NOV-9DAC-FRA-SL-1014 NOVELTIS 2015 Ce document est la propriété de NOVELTIS et ne peut être reproduit ou communiqué sans son autorisation 2
Several innovative products have been developed and validated by NOVELTIS for precision viticulture: Map and characterization of vineyard rows ; WINEO products Map of missing crops. Percentage of missing crops within 5% NOV-9DAC-FRA-SL-1014 NOVELTIS 2015 Ce document est la propriété de NOVELTIS et ne peut être reproduit ou communiqué sans son autorisation 3
Several innovative products have been developed and validated by NOVELTIS for precision viticulture: Map of vine vigour only on detected vines; WINEO products High Medium Low Classification along homogeneous areas. High vigour Low vigour NOV-9DAC-FRA-SL-1014 NOVELTIS 2015 Ce document est la propriété de NOVELTIS et ne peut être reproduit ou communiqué sans son autorisation 4
WINEO products Several available output formats. KMZ/KML files (can be visualized on Google Earth) Webmapping interface Login / Password Your layers http://innovine.noveltis.fr NOV-9DAC-FRA-SL-1014 NOVELTIS 2015 Ce document est la propriété de NOVELTIS et ne peut être reproduit ou communiqué sans son autorisation 5
Contact NOVELTIS creates added value for its customers by supplying innovative, tailored solutions. Florian POUSTOMIS Scientific engineer Email: florian.poustomis@noveltis.fr Tél. : +33 (0)5 62 88 11 43 NOV-9DAC-FRA-SL-1014 NOVELTIS 2015 Ce document est la propriété de NOVELTIS et ne peut être reproduit ou communiqué sans son autorisation 6
vite.net a DSS for sustainable vineyard management Powdered by
Mission Increase the value of research by transferring the technological innovation to practical agriculture. Core activity Development of Decision Support Systems (DSSs) for sustainable crop production based on new Information and Communication Technologies (ICTs).
Structure Information on crop & environment feeds the DSS in real time Web applicaiton Holistic approach Usefulness Strengths
How it works
Structure Web applicaiton No software installation, constant updating of the functionalities Holistic approach Usefulness Strengths
Web-application 7/7 24h
Structure Web applicaiton Holistic approach Considers all the key cultivation problems Usefulness Strengths
u Vine growth & development u Ripening & yield Plant u Berry mouth u Leafhopper u Mealybugs Protection u Downy mildew u Powdery mildew u Grey mold u Black-rot u OTA u PPPs u Optimal dose u Protection dynamic u Resistance risk Canopy management Stress u Frost u Drought u Topping u Leaf removal u Cluster thinning Cropping u Fertilization u Weed control u Irrigation Functionalities
Structure Web applicaiton Holistic view Usefulness Complex and multi-criteria decisional processes are turned into easy, clear & reliable decision supports Strengths
Frequency (%) 0 20 40 60 80 100 Less time spent Easier decision-making Better awareness Better knowledge Better decisions Score 1 2 3 4 5 Friendliness Speed Clarity Confidence Usefulness Usefulness: user s feedback
Same protection than the grower s schedule with important savings: Organic viticulture Kg copper/season: - 37% Cost of protection: -195 /ha IPM TFI / season: - 30-40% Cost of protection: - 250-300 /ha Usefulness: economic benefits
Conventional vite.net IPM vite.net organic Usefulness: environmental benefits
14 3 1 5 5 20 2 1 4 18 3 29 2 12 14 18 5 2 2 + Spain Portugal Greece Israel 1 1 3 4 1 1 5 6 9 2014: 66 users 2015: 180 users 2016: 300 users
Thank you for your attention