SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

13
CGIAR Open Data and Big Data Efforts Medha Devare GODAN Workshop 30 September 2016

Transcript of SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Page 1: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

CGIAR Open Data and Big Data Efforts

Medha Devare

GODAN Workshop30 September 2016

Page 2: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Grassini et al., 2013

Devkota et. al; Field Crops Res 179:81-94

Intensification requires

layered options based on

multi- dimensional

analysis, user

knowledge, preferences

Page 3: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Challenges

http://guides.library.queensu.ca/infoneeds

Page 4: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Get from this… To this:

Requires access to data sets and harmonization on…

Challenges

interoperability (standards, ontologies…)

tools/platformsincentives/culture

Page 5: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

http://www.breidt.net/scripts/pics/hochb.jpg

Page 6: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Global discovery

Federated search across centers

Categorized content type

Faceted results

Contents referenced via standard geo-coordination (ISO)

Machine-readable

Human + machine readable content

Collection of tools

Toolkit for analytics

Improved access, reuse

High precision, integration via controlled vocabularies, ontologies

Learning from other domains: The aspiration

Credit: Soonho Kim (IFPRI) and NCBI

Page 7: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Learning from other domains… http://ncbi.nlm.nih.gov

Page 8: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Bricks and rivets…

Center+ repositories, dbs

Interoperability

Ontologies, vocabularies

Harmonized data/info

Data management

Agronomy-breeding mgmt. systems

CGIAR technology catalog

Analytics/tools

Technology mapping

Research discovery

Decision support, visualization

M+E

Infrastructure (LOD enabled)

CGIAR plans, budgets, approaches aligned

across units[leadership, legal, HR,

Proj Mgmt, IT, KM-DM]

Aligned CG-donor policies,

guidelines/DMPs…

OA-OD capacity, support, visibility

Metadata, SOPs

phase I OA/OD, build in phase II

phase II OA/OD / Big Data

Apps – links to telcos(to/from farmer)

Data quality, workflows

Genebanks

Other interoperable

platforms

Genetic Gains

Page 9: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Hey Cigi, when should I plant my maize? How should I manage my crop?

Real-time decision support system for farmers

Easy natural language as an interface

Smart artificial intelligence trained by CGIAR and partners

Leveraging open, harmonized and interoperable multiple databases

Credit: Jawoo Koo (IFPRI)

The aspiration…

Page 11: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Bricks and rivets: CG core metadata schema

Page 12: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Bricks and rivets: Agronomy Management System

Plot size: 5 x 5 sq. m (or as close to it as possible) = 8 rows in line-sown plots EXTRA PLOTS: T9 and T10

T1 T2 T3 T4 T9

OPV = Arun 2 OPV = Arun 2 OPV = Arun 2 OPV = Arun 2 OPV = Rampur Composite

broadcast line sown line sown broadcast line sown

farmer NPK farmer NPK 120:60:60 NPK 120:60:60 NPK farmer NPK

T5 T6 T7 T8 T10

hybrid hybrid hybrid hybrid OPV = Rampur Composite

broadcast line sown line sown broadcast line sown

farmer NPK farmer NPK 120:60:60 NPK 120:60:60 NPK 120:60:60 NPK

Stage of trial Protocol

Pre-planting * Perform germination tests with hybrid and OPV; estimate germination percent

Tillage * Note what farmers did to prepare each field/rep; try to get a realistic estimate of time and

labor required

Crop establishment * Record sowing date for each plot

Seed rate * 40 kg/ha; calculate the rate for your plots, be sure to adjust for germination percent!

Sowing * Note dates, time taken, labor required for planting each plot.

Line-sown

* Calculate number of rows per plot, and amount of seed required per row, and prepare

separate containers/bags per row containing seed

* Rows should be spaced 60 cm apart (i.e. 8 rows in line-sown plots)

* Use jab planter to sow line-sown plots, making sure that seed are deep enough to find

moisture and germinate. Note average time taken for line-sowing plot.

Data harmonization via AgrO-based Agronomy Management System + Fieldbook (similar to the BMS)

Adaptive Research Data 2011 Wheat

ART Name: Intercropping in Sugarcane

Yield Data of Rajma in sugarcane intercropping

REP TRT Seeding Date/Variety Flowering daysMaturity days PHT(cm)F pod/Plant UF pod/Plant Grain/pod TGW(g) YLD kg Area(m²) Seed Yield(kg/m2)

S1 S2 S3

1 1 25-Nov..

1 2 1-Dec. 124 90 1.177 1.481 1.352

1 3 1-Dec. 47 110 31.24 10 3 2.7 346 17.0 180 0.128 0.148 0.157

1 4 1-Dec. 44 65 40 6 1 6 283.5 6.5 180 0.083 0.079 0.054

Page 13: SC2 Workshop 2: CGIAR Open Data and Big Data Efforts

Farmer field trial

Experiment station trial

Greenhouse trial

Farmer field demonstration

Experiment station

demonstration

Technology adoption study

Other

Trial, demo type

Farmer field trial Farmer managed

Researcher managed

Other

Farmer field trial

Farmer managed

Personnel

PI

Field supervisor

Data collector

Site

Country

State

District

Sub-district

Town

Village

Municipality

Ecozone Altitude

Latitude

Longitude

Goal: Pre-loaded,

authoritative lists