The agriculture production system increasingly becomes data-driven and data-enabled based on the cyber-physical management cycle. This paper describes several IoT-applications of the EU-funded IoF2020 project in which data and data-sharing plays a crucial role. It provides an integrative framework aiming at cross-fertilisation, co-creation and co-ownership of results. Technical integration, business support and ecosystem development are key mechanisms to realize this.
APPLICATION OF BIG DATA IN ENHANCING EFFECTIVE DECISION MAKING IN AGRICULTURAL PRODUCTION
1. Application of Big Data in enhancing effective
decision-making in Agriculture Production
Sjaak Wolfert, Senior Scientist
International Agricultural Congress, 13-15 Nov. 2018, Kuala Lumpur, Malaysia
2. 2
Interview with
Johan Bouma in
Resource 4 Oct. 2018
p. 18-19
1. Multidisciplinarity
2. Collaborative process
3. Agile development
3. Important ICT Trends
Mobile/Cloud Computing – smart phones, wearables, incl. sensors
Social media – Youtube, Facebook, Twitter, etc.
Location-based monitoring – GPS, remote sensing, geo information, drones
Internet of Things – everything gets connected in the internet (virtualisation,
M2M, autonomous devices)
Block Chain – Tracing & Tracking, distributed ledgers, smart contracts
Big Data - Web of Data, Linked Open Data, Big data algorithms
Next step: Artificial Intelligence – Deep learning, Machine learning, etc.
anywhere
High Potential for unprecedented innovations!
7. Redefining Industry Boundaries
3. Smart, connected product
+
+
+
2. Smart Product
1. Product
Adapted from: Porter and Heppelmann, Harvard Business Review, 2014)
8. 5. System of systems
farm
management
system
farm
equipment
system
weather
data
system
irrigation
system
seed
optimizing
system
field
sensors
irrigation
nodes
irrigation
application
seed
optimization
application
farm
performance
database
seed
database
weather data
application
weather
forecasts
weather
maps
rain, humidity,
temperature sensors
farm
equipment
system
planters
tillers
combine
harvesters
4. Product system
Adapted from: Porter and Heppelmann, Harvard Business Review, 2014
9. 5. System of systems
farm
management
system
farm
equipment
system
weather
data
system
irrigation
system
seed
optimizing
system
field
sensors
irrigation
nodes
irrigation
application
seed
optimization
application
farm
performance
database
seed
database
weather data
application
weather
forecasts
weather
maps
rain, humidity,
temperature sensors
farm
equipment
system
planters
tillers
combine
harvesters
4. Product system
Your company
10. 5. System of systems
farm
management
system
farm
equipment
system
weather
data
system
irrigation
system
seed
optimizing
system
field
sensors
irrigation
nodes
irrigation
application
seed
optimization
application
farm
performance
database
seed
database
weather data
application
weather
forecasts
weather
maps
rain, humidity,
temperature sensors
farm
equipment
system
planters
tillers
combine
harvesters
4. Product system
Your company
Farmer:
How many platforms must
I use?
Developer:
On how many platforms
should I offer my
solution?
Platform owner:
How many connections do
I need to maintain?
11. The Landscape of Data for Farming and Food
Farming
Data
Food
Data
See: Wolfert et al., Agricultural Systems 153 (2017) 69–80
Processors
Ag
Business Tech
Companies
Tech
Start-up
Tech
Start-up
Ag Tech
Retail
Venture
Capitalists
Data
Start-up
Data
Start-up
13. Creating a collaborative infrastructure
Scenario: get expert advice for spraying to
handle disease on tomatoes
State AuthorityFranz Farmer Ed Expert
Spraying
(follow advice)
Create
Advice
Approval
Request
Advice
CollaborativeBusinessProcess
1
2
3
FIspace App
‘Weather
Information’
FIspace App
‘Spraying
Expert Advice’
FIspace App
‘Spraying
Certification’
Back-EndSystems
Farm
Management
Systems
Sensor Network
in the Greenhouse
Agronomist
Expert System
Regulations &
Approval
System
product type, etc.
sensor data
(access details)
suggested
chemical
advice details
certification
details
13
14. Intermediate conclusions
Agri-Food chains become more
technology/data-driven
Probably causes major shifts in
roles and power relations among
different players in agri-food chain
networks
There is a need for a facilitating
open infrastructure
Two extreme scenarios:
1. Strong integrated supply chain
2. Open collaboration network
Reality somewhere in between!
15. Governance
● privacy, security, stakeholders...
Business models
● fair share, new opportunities
Infrastructure
● open versus closed, integration
Ecosystem Development
● establishing critical mass
...which are often intertwined!
Current key issues and challenges
16.
17. Objective:
Large-scale uptake of IoT in the European
farming and food sector
• Business case of IoT
• Integrate and reuse available IoT
technologies
• User acceptability of IoT
• Sustainability of IoT solutions
17
Internet of Food and Farm 2020
Innovation Action: 2017 - 2020
30 M€ funding by DG-CNCT/AGRI
18.
19. THE INTERNET OF ARABLE FARMING
1.1 Within-field Management Zoning (potato)
1.2 Precision Crop Management (wheat)
1.3 Soya Protein Management (soya)
1.4 Farm Machine Interoperability
20. THE INTERNET OF DAIRY FARMING
20
2.1 Grazing Cow Monitor
2.2 Happy Cow
2.3 Silent Herdsman
2.4 Remote Milk Quality
21. 3.1 Fresh Table Grapes Chain
3.2 Big Wine Optimization
3.3 Automated Olive Chain
3.4 Intelligent Fruit Logistics
THE INTERNET OF FRUIT
21
22. 4.1 City Farming for Leafy Vegetables
4.2 Chain-integrated Greenhouse Production
4.3 Added Value Weeding Data
4.4 Enhanced Quality Certification System
THE INTERNET OF VEGETABLES
22
23. 5.1 Pig Farm Management
5.2 Poultry Chain Management
5.3 Meat Transparency and Traceability
THE INTERNET OF MEAT
23
24. Soil map based variable rate applications and machine automation in potato production
UC1.1. WITHIN-FIELD
MANAGEMENT ZONING
Coordinators: Peter Paree (ZLTO) & Corné Kempenaar (WUR)
25. SOIL MAP SERVICE
VARIABLE RATE
APPLICATION MAP
AUTOMATION & MACHINE
COMMUNICATION
Product Impressions
26. IoF2020 - Trial: The internet of Arable
Farming
Use case 1.1: Within-field management zoning
Short description and location
Sensing and actuating devices are used to gather data, mainly related to potatoes, predict
yields, define management zones, monitor and optimize growing potatoes’ behaviour,
optimize use of herbicides, and optimize farm management. (NL, DE)
Domain application areas addressed
Management zoning of arable fields; Crop protection; Yield prediction.
(Farming, Logistics)
IoT Devices
30 sensors for soil moisture, Veris soil scanner,
machine control, yield sensors, indoor climate,
crop quality, 4 weather stations, 3 GEO-localization
units, NDVI Sensor
IoT Platforms and Software
Initiatives and platforms: FIWARE,
FIspace, EPCIS, AgroSense, Apache Cassandra,
Apache Flink, Apache Spark
IoT Applications
Weather forecast service, Growing crops,
Akkerweb agro-eco algorithms; GIS, zoning and
T&T modules; Control fertilize machines; Control
irrigation systems; Measure soil temperature and
water potential
IoT Technologies and Standards
Lora Network, 365FarmNet, Zoner, Crop-R and
Akkerweb platforms, Cloudfarm FMIS, ISOBUS.SW/HW Infrastructure
Cropfield sensors platform,
Agriculture combination (e.g.,
tracktor), Manufacturer Cloud
with cloud storage, FMIS Cloud,
Prediction Model Cloud
Architecture View
Partners
ZLTO (NL); Kverneland Group (NL);
KPN (NL); Bayer CropScience AG
(DE); Van den Borne Aardappelen
(NL); Grimme Landmachinen-fabrik
GmbH & Co (DE); Wageningen
University & Research (NL).
27. Major Challenge Here is what we aim to improve (KPIs)
Yield by better
plant distribution
Variable planting distance map –
Validation in 2017 and 2018. Nov. 2018
portal where maps can be ordered.
Variable rate herbicide use map -
Validation in 2016 and 2017. May 2018
portal where maps can be ordered.
Quality by better
plant distribution
Reduction
pesticide use
Core Product Features
Variable Rate
Application Map Service
Customer & Provider
Uses soil maps and agronomic knowledge to create
crop management task map based on variability in
soil characteristics like organic matter and/or clay
content, water storage capacity, tramlines, shade,
etc..
Smart application of resources: seeds,
pesticides, fertilizers +4%
+5%
-23%
Better distribution of plants leads to +5% kilos and +5% better
quality (more potatoes in desired size). Taking soil characteristics
for weed growth into account: -23% less herbicide and +2% more
yield.
Enriching canopy index with soil characteristics lead to -10% less
additional N fertilizer (2nd phase).
These values derive from comparison of a standard farm’s performance
prior to the installation of our system and after.
Reduction
fertilizer use
-10%
Product Factsheet
Existing variable rate maps are often based on tweaking
expert judgement and lack a certain level of precision in
tasking / lack of validation.
Farmers and advisors
Price per unit, added value
LoonwerkGPS,
soil analysis labs,
FMIS providers VRA additional N spraying
June 2018 on Growth + Soil Maps.
High spatio-temporal monitoring dashboard
28. IoT tools for sustainable wine production, wine quality management and shipping monitoring
BIG WINE OPTIMIZATION
Some KPI’s: Pesticides -10% | Production costs -10% | Wine quality +10% | Shipping costs -5%
30. IoT Product Impressions
sensors in
the vineyard
display devices,
agronomic parameters
and weather forecast
Temperature/RH
logger
with data
transmission
NIR spectrometer
% alc., sugar,
etc.
31. IOF2020 ECOSYSTEM & COLLABORATION SPACE
WP1ProjectCoordination&
Management
GENERIC APPROACH & STRUCTURE
WP2 Trials/Use cases: Knowledge & App development
Lean multi-actor approach
3. EVALUATION
1. CO-DESIGN
2. IMPLEMENTATION
P1
P2
LARGE
SCALE
P3
WP3 IoT Integration WP4 Business Support
WP5 Ecosystem Development
33. SmartAgrihubs – another 20M€ project
33
Consolidate and foster EU-wide network of Ag Digital Innovation Hubs
Start: 1 November 2018, duration: 4 years
34. Specific Objectives
Build network covering all EU regions including
technology, business, sector expertise + relevant
players
Critical mass of multi-actor Innovation
Experiments
Financial support 3rd parties by open calls –
various public/private funds
Ensure long-term sustainability incl. business
plans + attracting investors
Promote DIH’s full innovation accelerating
potential
34
35. Concepts and coherence
35
• Layered network of Competence
Centers and Digital Innovation Hubs
organized in Regional Clusters
• Multi-Actor Innovation Experiments
interacting with DIH’s innovation
services
• Innovation Services Maturity Model
developing the DIHs
• Innovation Portal supporting
Ecosystem Development
37. ECOSYSTEM & COLLABORATION SPACE
ProjectCoordination&
Management
Multidisciplinary, Collaborative, Agile Approach
Trials/Use Cases: Knowledge & App development
Lean multi-actor approach
3. EVALUATION
1. CO-DESIGN
2. IMPLEMENTATION
P1
P2
LARGE
SCALE
P3
Data Science &
Information management
Business Modelling,
Governance & Ethics
Ecosystem Development
38. Thank you for your
attention!
More information:
sjaak.wolfert@wur.nl
nl.linkedin.com/in/sjaakwolfert/
Twitter: @sjaakwolfert
http://www.slideshare.net/SjaakWolfert
38
Editor's Notes
The management or decision support cycle increasingly becomes a cyber-physical cycle system monitored by advanced sensor networks and controlled by data and computer-based algorithms, tightly integrated with the internet and its users. In this way the cycle increasingly becomes autonomous, with less human intervention. This development can be applied to all parts of the food supply chain, while at the overall level data and high-tech are also driving new ways of traceability. But also public decision-making is increasingly supported by data-driven models and algorithms in order to make decisions on e.g. food safety or environmental control. Consequently, the data and related infrastructure for private and public purposes increasingly becomes intertwined in which the same data is used for multiple purposes, both private and public. However, this raises various issues on data ownership, -access, privacy, ethics etc. and progress will highly depend on creating trust and partnerships.
Against this background we can distinguish four sub-themes that can be addressed by projects within these themes.
This slide provides an overview of the project aim and objectives.
Through these projects we have developed a success formula in approaching the challenge of ICT and Information Management in Agri-Food :
Trials and use cases form the core, in which we jointly develop as research and business organisations, knowledge and application through a lean multi-actor approach
This means that we quickly develop minimum viable products with involvement of all relevant stakeholders and upscale these through several cycles of development
In parallel we create synergy by
Technical integration: open architectures, standard that can be used as generic building blocks in the trials and use cases
Governance and business modelling: solve issues that arise from the trials and use cases regarding ownership, privacy, trust, etc. and support the businesses in developing sustainable business plans for the apps, services and organization structures that are being developed
Ecosystem Development – support the trials and use cases in embedding their solutions in global ecosystems and upgrading them to a large scale
Project coordination and management is trivial, but we have shown that Wageningen University and Research is very capable to fulfil this role in large public-private projects
This integrated approach will guarantee long-term, sustainable results from these projects.
This has become our general project approach in many projects...