Transcript Presentation

Big Data in Indian Agriculture
D. Rama Rao
Director, NAARM
Agricultural Data Sources
 Farmers portals (~100TB)
 Research data (~200TB)
 Bioinformatics (~300TB)
 Remote sensing (~100TB)
 Weather, Remote sensing (~500TB)
 Government departments (~10TB)
 Communication & Media (~200TB)
 Agribusiness (~10TB)
 Many new sources
 Features of above (~1500TB)
 Variety (numerical, text, images, etc)
 Velocity (daily, seasonal, annual etc.)
 Veracity (regional differences, highly volatile)
 Datasets are large and many cases not structured
Agricultural Big Data
Public Data Sources






Research Models and Decision Tools
Short term weather
Longer term climate
National soil database
Digital elevation models
Markets info.
Regional and national
inventories etc.,
 Integrated farm models
 Crop growth models
 Soil-water balance
models
 Machinery selection
algorithms
 Best management
practice evaluators
 Precision Agriculture
 Farm optimization
tools etc.,
Private Data and Inputs
 Machinery and labor
availability
 Crop rotations
 Marketing strategies
 Private inventories
 Remote sensing images
 Input prices
 Yield maps
 Crop input maps etc.,
 Improving Farmer Decisions
 Advancing Research Methods
 Enabling Improved Policy
Big Data
Analytics
Bioinformatics : Complex
Genetic Interactions
Environment
Genotype
Phenotype
Data processing tools getting more and more sophisticated
Repositories in KRISHI
(krishi.icar.gov.in)
Technology
Repository
Geo portal
Publication Repository
Experimental
Data
Repository
Survey Data
Repository
Observational
Data
Repository
Challenge-1: Bioinformatics
• Next-Generation Sequencing (NGS) platforms have
exponentially increased the rate of biological data
generation in the last two years
• For a large genome, DNA data can occupy many
terabytes, and completing the genome sequence
require months of computation on supercomputer
• A typical crop genetic dataset might include several
million genetic markers controlling poly genes of
various germplasm
• Data accumulating in computers and servers around
the world concerns over privacy and security
• Improvements in computational infrastructure would
permit more rapid analysis and enhance the impact
Challenge-2:
Microbes & Metagenomics
• The diverse microbial communities (microbiome) in
plant and the intestinal tracts of animals and humans
• Big Data analyses of the microbial populations provide a
view into their millions of genes and gene products, and
their mostly unknown intercommunications with host
tissues. This would lead to better diet formulations,
enhanced early life immunity, and reduced food safety
concerns, disease resistance, etc
• The explosion of metagenomic knowledge has only
begun to be explored for our soil and water resources
Drought Monitoring
A system could integrate
soil, weather, or satellite
data, and models for
parameters such as
drought with locally
collected environmental,
genetic, and phenotypic
data into a common
framework
Challenge-3:
Environmental Modelling
• Studies that measure the cause of interactions among the crop,
soil, water, weather, climate, and management differences are
complex. Impact of environmental conditions on agricultural
systems often yield conflicting results at different locations
• Vast need and scope for forewarning pest and disease
occurrences well in advance and in a dynamic mode
• Using Big Data approaches, “Modeling and the ability to combine
data from different sources, promises to revolutionize
understanding of processes affecting management of natural
resources” and thereby making Indian Agriculture CLIMATE
SMART
• With availability of Big Data, drought monitoring can help in
evolving suitable policy formulations
Challenge-4:
Market Intelligence
• For a number of commodities, online data is
available from more than 1500 markets and huge
data with high frequency is generated from Forward
and Spot Markets
• Key challenges are integration of various data from
markets and environmental conditions on
agricultural systems would pave way for location
specific assessment of risk, insurance, forecasting of
demand and supply changes, etc for timely decision
making by farmers, policy makers and other stake
holders
• Forewarning food security threats
Challenge-5:
Farm ERPs & Agri Portals
• With rising automation and access to IT, a
variety of decision models are providing
automation,
knowledge and information
services to farmers
• Big Data capability help Integration of farmers
to agri-value chains through new markets for
agricultural commodities. This help in
evolution of farmers’ based DSS and also
access to online and e-markets in food
commodities
End Note
• Big Data offers tremendous hope in Indian
agriculture. However, the implementation is likely to
be “bumpy” and sporadic and may take quite long to
realise substantial benefits
• Big Data has the potential to create the next major
technological “sea change” in agriculture
• The technology may change the “balance of power”
in the agri-food value chain
• There will be “winners” and “losers” with the new
technology
• Ownership and control of data will be of concern