Big Data and AI boost production of raw materials from agriculture, forestry and fishery in a sustainable way

1. Big Data and AI
boost production of
bio-resources
Prof. Caj Södergård (@CajSod)
VTT Technical Research Centre
of Finland
AamuAreena at Tieke
8.2.2019
VTT – beyond the obvious 1

2. • Challenges in agriculture, forestry and fishery
• Big data and Artificial Intelligence as enablers
• Practical pilots in the DataBio project ( www.databio.eu )
Content
2

3. • Climate change and extreme weather
• Agriculture: drought, decrease in farmland, shortage of irrigation water
• Forestry: storms, fires, diseases, need for carbon sink
• Population growth
• Incresed need for food raw materials from agriculture and fishery
• Urbanisation, changes in values
•Pollution, less farmland, more meat consumed (despite counter trends),
pressure to use less pesticides and fertilizers
•Conversion towards bio-based materials and fuels
Challenges in agriculture, forestry and fishery
3

4. • Data and AI to boost - in a sustainable way - production within
agriculture, forestry and fishery
- satellites, airplanes, drones
- sensors in fields, air, ocean
- sensors in agriculture machinery, forest harvesters and fishing
vessels
- other data (weather, market prices...)
- integrate, analyze and visualize data
• Increase production, decrease costs and burdon on environment
• Support end user (farmer, forestry owner, fishermen) in descisions
Big Data and AI as enablers
4

5. • Bioeconomy: The utilization of raw materials from
agriculture, forestry and fishery for food, energy and
biomaterials with responsibility & sustainability
• DataBio shows the benefits of Big Data in the raw material
production for the bioeconomy industry
• EU H2020 Lighthouse. Duration 2017 - 2019, Volume 16 M€,
48 partners, VTT biggest partner & Technical Manager
Data Driven Bioeconomy
5
EXAMPLE

6. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
6
Project at a glance
2017 2018 2019

7. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
7
DataBio platform serves the 26 pilots

8. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
8
Pilot: Precision Agriculture in Olives, Fruits, Grapes
• Smart farming pilot focusing on the exploitation of
heterogeneous data, facts and scientific knowledge
to facilitate decisions and their application in the field
• The pilot promotes sustainable farming practices
through the provision of irrigation, fertilization and
pest/disease management advices
• The farmer benefits from the provided big-data
technologies and advisory services by better
managing the natural resources, optimizing the use
of agricultural inputs and increasing farm yields
Stimagka
Chalkidiki
Veroia
3 pilot sites, crop types and advisory services
4 data sources
fieldremote eye farm

9. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
9
Prediction and real-time alerts of diseases and pests breakouts
• Process
• Collect, validate and store farm IoT data
• Combine with EO and historical farm data
• Perform initial processing, monitoring and
cross-checking on the raw data
• Push the validated values to CEP for further analysis
(temporal reasoning) for triggering early alerts in real-time
• Early experiments with olives (left)

10. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
1
0
PILOT INITIAL
RESULTS
http://www.ypaithros.gr/en/yannis-
olive-grove-
reduction-by-30-in-production-costs-
and-parallel-increase-of-sales/
SUCCESS STORIES
Chalkidiki Pilot
Avg cost of spraying
(euros/ha)
810
250
790
232
782
71
VEROIA CHALKIDIKI
Base Value
Target Value (1st year)
Current Value
Avg cost of irrigation
(euros/ha)
870
330
740
280
490
220
VEROIA CHALKIDIKI
Precision Agriculture in Olives, Fruits, Grapes
-40
-20
-20
Nitrogen under-
fertilization(%)
USER INTERFACES

11. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
1
1
T3.2.1 Pilot A1: Oceanic tuna fisheries immediate
operational choices
Item Description
Pilot
leader
Zigor Uriondo, EHU/UPV,
zigor.uriondo@ehu.es
Objective • Improve vessel energy efficiency
• Predict machinery faults
• Vessel loading to minimize fuel consumption
KPIs Fuel vs kg catch & nautical miles sailed, catch efficiency
Trial 1
Focus
Historical data collection, collation and visualization.
Proton deployment on vessel
Datasets
& Status
• ECHEBF – 3 Ships with 3 to 4 years operational data
collected (Izaro, Jai Alai, Euskadi Alai)
• Fish catches
• Fuel oil consumption
117 ship measurements every 10s
• main engine (1)
• flows to/from engine
• propeller (1)
• aux engines (5)
• flows to /from aux engine
• fish hold (20)

12. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
1
2
DataBio platform serves the 26 pilots

13. 13
Climate change
Illegal logging
Forest loss
Sustainable forest use
Forest monitoring
Forest asset management
Forest research
Biocarbon assessment
Stakeholders
• Forest and climate research
community
• Forest owners and managers
• Forest certification organisations
• Regional/national forest
administration
• International initiatives, research
programmes and panels
• International development banks
• Sustainable development NGOs
• UN organisations
• Value adding (SME) industries

14. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
1
4
Piloting a satellite data based forest inventory service
for efficient coverage of large areas in forest management
• Source data: Sentinel-2 optical satellite data (open)
• Reference data: Sample plot data /Finnish Forest Centre
• Platform: Forestry TEP /VTT
• Processing: Envimon, Probability /VTT
• End-user system: Wuudis /MHG
Output: forest parameter estimates
• Stem number; Stem volume – for
pine, spruce, broadleaved and total;
Diameter; Basal area; Height
DataBio Pilot
Forest Inventory Service
Source: Renne Tergujeff, VTT

15. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
1
5
Sentinel-2 Satellites
• Optical MultiSpectral Instrument
• 13 spectral bands
• 500 nm to 2200 nm
• Resolution 10 to 60 m
• Two polar-orbiting satellites (2A,
2B) at 786 km altitude operated by
European Space Agency
• Main uses: climate change,
land monitoring, emergency
management, and security
• 7 spectral bands used in the
DataBio Forest Inventory pilot
Copernicus data volumes (overall):
• Over 8 Petabytes
i.e. 8,000 TB = 8,000,000 GB
• New data: 10 TB per day
Source: Renne Tergujeff, VTT

16. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
1
6
Forestry TEP
One-stop shop for forestry
remote sensing services
for the academic, public and
commercial sectors
 Online service enabling quicker and
smoother value adding
 Access to satellite imagery,
computing power and value adding
services
 Platform for developing and sharing
own applications
 Worldwide marketing channel for
commercial remote sensing services
Source: Renne Tergujeff, VTT

17. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
1
7
Application
Platform
Infra & data
Processing pipeline
• Sentinel-2 Level 1C data (2017) acquired
from ESA via EO Cloud to Forestry TEP
• Model training with sample plot data
(2016) of the Finnish Forest Centre (FFC)
• Satellite data pre-processing and cloud
masking, using Envimon (VTT)
• Forest variable estimates produced, using
Probability (VTT)
• Output as WMS layers to Wuudis
Forestry TEP
EO Cloud / DIAS
Envimon, Probability
WUUDIS
ESA
FFC
Source: Renne Tergujeff, VTT

18. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
1
8
Pilot area in DataBio pilot
• Forest estimates generated for
the coverage of a Sentinel-2
data scene (100 x 100 km)
• Verification against the
Hippala forest estate (25
stands, 17 ha)
Reference data (2016) from
the Finnish Forest Centre
Source: Renne Tergujeff, VTT

19. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
1
9
Forest classification
Red = Broadleaved
Blue = Pine
Green = Spruce
• Visualising the
predominant forest class
Source: Renne Tergujeff, VTT

20. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
2
0
Forest parameter estimates
Hippala forest estate
Base map
(OpenStreetMap)
Test plotsSentinel-2 RGBHeightBasal areaDiameterStem numberStem vol total
Stem vol
broadleaved
Stem vol spruce
Stem volume
of pine
Source: Renne Tergujeff,
VTT

21. This document is part of a project that has received funding
from the European Union’s Horizon 2020 research and innovation programme
under agreement No 732064. It is the property of the DataBio consortium and shall not be distributed or
reproduced without the formal approval of the DataBio Management Committee. Find us at www.databio.eu.
2
1
Wuudis Service for Forest Owners, Industry and Operators

22. To conclude
22VTT 2018
• Big Data and AI boost the gathering of raw materials in bio-
economy in a sustainable way
• Plenty of data exist: satellites (largely for free), weather, machinery,
sensors
• By combining data sources, analyzing and visualizing data, the
user can get the support to
• do better decisions
• achieve better results
• save the environment