2. 2Solutions Showcase
Session Abstract
By 2050, the world's exploding demand will require farms to
feed upwards of 10 Billion people. Global leader in
precision agriculture, Land O’Lakes, uses a novel
application of cloud-scale compute and analytics to
revolutionize modern farming to produce more
corn on fewer acres today compared to 50 years ago.
4. 4Solutions Showcase
Unprecedented Opportunity
Source: United Nations Food and Agricultural Organization
● Diets are improving with rapid growth
of middle class
● Global food production will need to
increase 70% to meet higher demand
● World population to grow from 7.0 billion
to 9.2 billion by 2050
5. 5Solutions Showcase
Limited resources intensify the challenge
OECD-FAO, global water supply/demand model; agricultural production based on IFPRI computed general equilibrium model base case;
A Daunting Task, Prof. Robert Thompson
5
%
amount of additional land projected
to be cultivated by 2050
12% amount of additional arable land
available globally
40% amount water demand will outstrip
supply by 2030
6. 6Solutions Showcase 6
We must produce more food,
with fewer resources and less
environmental impact
7. 7Solutions Showcase
The Solution: Ag Tech Revolution
Ag Tech will
close the gap
between
genetic
potential and
real world
results
Source: USDA
NASS
Biotech
Revolution
Ag Tech
Revolution
Mechanization
Revolution
300
250
200
150
100
50
0
1900 1920 1940 1960 1980 2000 2020 2040 2060
Historical Corn Yields (bu/acre)
9. 9Solutions Showcase
WinField Data Silo™ – Share In, Share Out
Grower Owned
Data
Member Owned
Solution
Access Control In
Hands of Growers
or Delegates
Co-Branded with
Local
“Most Trusted”
Solution
10. 10Solutions Showcase
• Speed of development time
• Ability to change technologies as the
project’s needs changed
• Simple integration and ability to leverage
Google Maps
• Easy scalability to quickly deploy test
instances for remote developers
Value of Google Cloud
15. 15Solutions Showcase
Data Silo Development Timeline – Concept Phases 1 -
4
File Centric
Development
Spatial Centric
Development
Discovery Workshops
Feb-Apr May-Aug
Design Data Model
Build the Silo
Show boundaries
Tweak Data Model
Rebuild Database
Enrich the UI
Make it GIS Aware
Replatform the App
Build APIs
Security Assessment
Security Assessment
16. 16Solutions Showcase
Functional Architecture
Show Data
Available
Login
Upload Data
Initial Landing
Download Data
Security &
Permissions
Audit & Reporting
Impersonation
Spatial
Presentation
Search
OADA Compliant
APIs
Data
Transformation
Transaction
Engine
Rx
Seller Replica
Tissue Samples
Grower Replica
Field Boundaries
Account Replica GSI - Query
PresentationAccessData
19. 19Solutions Showcase
Boundary
• Boundaries - Must associate to fields
• Seed / Fert / Yield - ‘Deconstruct’ and
store as point / polygon data
– Associated Grower
• Has crop, date, fert,
product, yield qty, etc.
• May have point or polygon
data that is ‘outside’ all
defined boundaries
• Soil and Tissue data - Associated with
Grower
• Access controlled by Grower + Share
Logical Data Model
Farm 1
Seed Rx
Yield Data
Tissue Test Data (GIS
Points)
Farm 2
Grower
Field 3 Boundary x
Field 1
Field 2
Boundary 3
Boundary 4
Boundary 5
Boundary 2
Boundary 1
(GIS Poly)
Fert Rx
Soil Test Data (GIS
Points)
Associate any
boundary with
geospatial data
when ‘Querying’
20. 20Solutions Showcase
Application Data Model
user_audit
datasilo_user_id
event_id
date_time
datasilo_users
datasilo_user_id
primary_account_id
account_link
Datasilo_user_id
account_id
accounts
account_id
parent_account
sources
id
name
seasons
id
parent_grower
resources
resource_id
owner_datasilo_user_id
source.id
season
resource_access
resource_id
datasilo_user_id
permission
audit_change
resource_id
event_id
datasilo_user_id
data_time
audit_access
resource_id
event_id
datasilo_user_id
data_time
tags
name
Tables specific to
data type
V11.0 - 10/20/15
22. 22Solutions Showcase
API Designs
As Applied
Seed
Rx Seed
Fertilizer
Rx
As Applied
Fertilizer
Harvester
Yield Data
Raw
Processed
Summary
Soil and Tissue
Test Results
Boundaries
Please file
content to extra
Properties
Associate
to Field
x/y
Lat / Long
Upload
Files
Manual and
Metadata about
File
Browse for
Files
Publish Event
to Subscribers
R7 User
ADAPT
Data Silo
<<extend>>
<<extend>>
<<extend>>
<<extend>>
<<include>>
<<include>>
<<include>>
23. 23Solutions Showcase
Google Cloud Platform Architecture
Revised Requirements
• Provide View into geo spatial data
even as boundaries change
• Data Silo needs to be GIS aware –
All data has spatial reference
• Map UI is the main driver
• Include Field Experts in Design
Guidelines
• ‘Revolutionary’ Design – Separate
Data from boundaries
• Rapid Development
• Use LOL standards for OS
• Consider portability
Architectural Impact
• IaaS requires change in support
model
• Opportunity to consolidate Dev
platform and language
Postgres SQL
With GIS
24. 24Solutions Showcase
Lessons Learned
Cloud computing gives you flexibility.
Move freely between cloud services as you deploy
Understand your user base.
Don’t build for a scale you will never see
Use technologies your team is comfortable with
and that fit into the organizational dynamics.
25. 25Solutions Showcase
650%
more corn
today on
13%
fewer acres
than 1950
It can be done
Source: USDA Economic Research Service
Production(millionsbushels)
PlantedAcres(millionsofacres)
14000
12000
10000
8000
6000
4000
2000
0
200
150
100
50
0
Planted AcresCorn Production
26. 26Solutions Showcase
About CTP
Trusted advisor for enterprises moving
to the cloud.
• Experts in mission-critical application
development
• Experts in enterprise cloud and work
closely with Google, AWS, Microsoft,
OpenStack and others
• 300+ midmarket and enterprise projects,
all challenging, all successful
See how we can help: cloudtp.com
27. 27Solutions Showcase 27
Thank you for your time.
Teddy Bekele - TBekele@landolakes.com
Joey Jablonski - Joey.Jablonksi@cloudtp.com