13. !13
Satellites & Agriculture
• Crop Classification
• Crop Health Index
• Crop Water Stress Index
• Sowing Intelligence
• Sowing Progress
• Harvest Intelligence
• Harvest Progress
• Relative Yield
• Crop Risk Zoning
• Crop Cutting Experiment Optimization
• Assessment of Crop Damage due to :
• Flood
• Frost
• Disease, etc
• etc
14. Area Covered
!14
90 Districts in India
• Governments
• Banks
• Insurers
• Re-Insurers
Outside India
• Ivory Coast
• Tanzania
• Nigeria
• Bolivia
• Australia
• Japan
• UAE
• Ghana
1 Million Sq Km Weekly
500 Million Sq Km Analysed Globally
Sectors
• Agriculture
• Infrastructure
• Renewables
“… an Indian bank was able to increase
its book size by 2% and achieve a
reduction of 1.5% in nonperforming
assets during a single season.”
15. Farm Level Performance
!15!15
Crop : Mustard
Sate : Rajasthan
District : Sawaimadhopur
• Taluk : Bonli
• Village : Banholi
• Lat / Long : 26.162 / 76.407
04-Nov-2016 20-Nov-2016 06-Dec-2016 22-Dec-2016 07-Jan–2017 23-Jan–2017
17. !17
Ex-VP, Goldman Sachs
23 yrs industry exp
UK, India & U.S.
Amardeep Sibia
CEO
Scientist, ISRO
Consultant NSR
7 yrs space industry exp
India, U.S. & EU
Prateep Basu
CSO & Founder
Ex-Petrofac Risk
Manager
11 yrs Oil & Gas
GCC
Samuel John
COO
Serial Entrepreneur
12 yrs Insurance & Consulting
EU & Asia
Abhishek Raju
Director, Partnerships & Founder
CFA
17 years Finance
Managing Partner Silver
Crescent Capital
M. Gopinath
CFO
Analyst, Political
Quotient
3 yrs Political Consulting
UN Youth Ambassador
India
Naga Sravan
VP, Policy & Govt Relations
Scientist, ISRO
7 yrs astrophysics
India
Ishan Tomar
CTO & Founder
Scientist, ISRO
7 yrs satellite remote
sensing
India
Vivek Gautam
Remote Sensing & Founder
Scientist, ISRO
5 yrs GIS
India & Canada
Rashmit Singh
Global Head Product MGMT
Scientist, NCFC
6 yrs Crop Modelling
India
Pradeep Bisen
Agriculture Lead
The Team
23. Crop Risk - Yield Estimation
!23!23
• Higher accuracy of yield estimates
• Reduce number of CCEs by 70%
SMART SAMPLING OF CROP
CUTTING EXPERIMENTS
• Reducing the cost of insurance
administration
• Setting threshold crop yield with
better accuracy
SRIKAKULAM
In-Season Crop State
24. Crop Risk Zoning
!24
Risk rating of crops :
• Time Series Rainfall Data
• % irrigation
• Crop Yield
• Crop Prices
• Crop Feasibility Analysis
27. !27
Cyclone Titli Impact
Pre-Cyclone Flooding
Highlights
• 11 Oct : Cyclone Landfall
• Sentinel-1 radar satellite
data used due to clouds
Timelines
• 10 Oct : Satellite imagery
- Inundation due to leading
cyclone clouds
• 17 Oct : Satellite imagery
- Post cyclone inundation
assessment
• 18 Oct : Portal Live
• 25 Oct : Sentinel-2 Optical
data
- Final assessment based
on yield loss
Post-Cyclone FloodingCropped AreaDamage
29. Dispute Resolution
!29
Objective
1. Validate reported CCE historical data
2. Create alternate adjudication data for missing CCE data
Proposed Methodology
Use satellite based yield esitmates
Plan
1. Estimate yield by using the satellite data based yield model from
CCE Optimisation.
2. Average yield for the village can be calculated after removing
outliers. This yield can be checked against threshold yield agreed
to in the contract
3. If the yield for a particular CCE is disputed, accept reported yield
a. within the upper & lower limits after removing outliers or
b. within 3 SD of average yield calculated above
30. Srikakulam Land Classification
Kharif 2017
!30
Class Area (ha) Class %
Crop land 3,18,675.15 54.91
Deciduous Forest 89,920.89 15.49
Scrub Land 52,167.06 8.99
Plantation/ Orchard 36,995.22 6.37
Fallow land 34,085.61 5.87
Water Bodies 22,676.58 3.91
Other Waste Land 12,493.35 2.15
Built-up + Roads 11,232.18 1.94
Riverbed 1,886.76 0.33
Shrubs 202.5 0.03
Crop land
55%
Built-up
15%
Water Body
9%
Fallow land
6%
Deciduous Forest
6%
Plantation/ Orchard
4%
Other Waste Land
2%
Riverbed
2%