Big Data & Big Opportunities

Big Data & Big Opportunities Dr. Matt Darr, Iowa State University For a copy of this slide deck please send an email request to darr@iastate.edu

Presentation Guiding Principles No intent to be critical of, or endorse, specific products/services to the exclusion of others that fulfill similar functions. Mention of product/service name is for information purposes only. This discussion is a snapshot of what is available today, and is intended to generate positive momentum around the ag data space. We must recognize that everyone along the chain must derive income from AG BIG DATA to be commercially viable.

What is Big Data? Big Data is data whose scale, diversity, and complexity require new architecture, techniques, algorithms, and analytics to manage it and extract value and hidden knowledge from it. Digital Agriculture is the new industry which is combining large data sources with advanced crop and environment models to provide actionable on-farm decisions.

Is Big Data New? Big Data in 1910s

Why the New Emphasis on Data Today? Autosteering and swath control technology have driven strong ROI which has led to a proliferation of GPS technology on farm. This leads to free machine data. Typical ROI in Iowa: 3.3% Planting Overlap Error, $7.89/ac 7% Tillage Overlap Error, $0.96/ac

Segments of the Digital Agriculture Industry Which ones are you in? Data Generation & Capture Yield Maps, Soil Fertility, Aerial Imagery, UAVs Wireless Data Transfer Data Warehouse Cloud Data Storage Prescription Agriculture VRA, Multi-hybrid Planting Probabilistic Decision Management Nitrogen Modeling Weather & Soil Suitability Modeling

Data Generation: What Role Will High Resolution Imagery Play? Satellite Delivered: 5m Resolution Timing can be limiting but more options are becoming available Contracted Flight: 1m Resolution Typically can schedule images within a +/- 3 day window around target date Unmanned Aerial Systems (suas) ~3 10 cm Resolution If weather permits scheduling can be within a few hours of target time

High Resolution Imagery in Agriculture Red-Green-Blue (RGB) Imagery

High Resolution Imagery in Agriculture Near Infrared (NIR) Imagery

High Resolution Imagery in Agriculture Normalized Difference Vegetation Index (NDVI) Imagery

High Resolution Imagery in Agriculture

Can you rank the image resolution of these three fields?

Eliminate Status Quo Crop Production

Data Warehouse Crop Consultant Seed/Fert Supplier Machinery Supplier Insurance Agent Grower Driven Entity Landlord Internal Mng Team Pooled Analysis

Data Analytics: Field Example 170 Acre Field, Continuous Corn

Data Analytics: Field Example Hybrid A Hybrid B 170 Acre Field, Continuous Corn

Data Analytics: Field Example 200 Hybrid A Yield Comparison of Two Hybrids in a Side-by-Side Test 176 Hybrid B Grain Yield (bu/ac) 150 100 50 131 0 Hybrid A Hybrid B 170 Acre Field, Continuous Corn

Data Analytics: Field Example

Compaction from previous machine operations Variety A Variety B 24

Highly Big productive Data Field zone Example

Highly Big productive Data Field zone Example 195 bu/ac

Big Data Field Example Healthy Plants in Compacted Area

Big Healthy Data Plants Field in Example Compacted Area 160 bu/ac

Big Data Field Example Weak Plants in Compacted Area

Weak Big Plants Data in Compacted Field Example Area 145 bu/ac

Producer Value: Quantify the impact of production practices. 120 acres x 70% compacted x 15 bu/ac yield loss = 1,260 bu yield loss = $5,000+ Cost of imagery = $240 33

Data Analytics: 1995 High Definition

Data Analytics: 2014 High Definition

Data Analytics: Aggregated Data Iowa Soybean Association On-Farm Network Over 2,500 on farm trials since 2007. Conducted in cooperation with grower partners. Data is available in a nonidentifiable form through the On-Farm network website. Increasing scale of the dataset allows for strong assessment of performance trends.

Data Analytics: Value of Aggregated Data Hundreds of data points for comparison across a broad range of geographic and crop production boundaries.

Data Analytics: Value of Aggregated Data What if every pass across the field with a machine was an On-Farm trial? How fast could we progress agriculture if this level of data was collected and shared broadly within Hundreds of data points for comparison across a broad range of geographic and crop production boundaries. grower cooperatives?

Probabilistic Decision Management Incorporate probability of events occurring, mainly weather related. Utilizes extensive historical data and weather forecasting data to drive model predictions. As the season progresses real time weather data in integrated into the model to improve robustness. Deterministic Model: Outcome is a single value with no randomness, i.e. soil sample based fertility recommendations. Probabilistic Model: Outcome is a range of potential values that represent environmental variability and can be used to manage risk.

Probabilistic Nitrogen Management

Spatially Specific Probabilistic Modeling

Big Data Policy Issues Receiving Major National Attention Farmer ownership of data Farmer control of data Disclosure of data usage Farmer choice for use of data Portability of data Security from misuse No vulnerability to FOIA Compatibility of systems Protection of GPS Regulation of UAVs Use of aggregated data Consistency of agreements Simple language Transparency and consistency Farmers express concern about all of these issues.

Producer Surveys on Big Data Skeptical of the New Technology 65% The biggest concern is misuse of farm data by: The ATPs Activist groups Grain traders The government Computer hackers Fear that it favors the large farmers. Prescriptions will recommend only some products, i.e., are biased. It doesn t work. Agriculture is too complex. Neutral or Nuanced in Attitudes 19% It has potential, but must be implemented carefully. Embracing the New Technology 16% The technology is here to stay. Let s embrace it and make it work for us. No one that is highly profitable today is doing it with only their own ideas and crop data.

Questions to Ask Your Data Services Provider Will my data be pooled with any other producers for large scale data analytics? If so, who will have access to this data? Am I compensated for the value of my production intellectual property in the development of improved services? How do you manage data ownership? The majority of bank providers will stress that producers own there data. Focus on the fine print around licensing. Many banks stipulate a perpetual royalty free license. This means they can use your data for free even if you close your account. Will copies of my data be retained within the bank if I close my account? Who has access to my data? Can I redirect my data to third parties of my choice? What type of security protocols are in place to protect against cyber hacking and data piracy? How will the bank avoid a Target incident? Is the bank independent or are they also trying to provide agronomic or business management services?

Digital Agriculture Opportunities for Providing Data Services Additional service opportunities and closer connections with clients specifically around product selection and input timing. New business opportunities in data warehousing, production benchmarking, profit/loss analysis, and data analysis. Opportunity to expand trust relationship with customers. This industry represents new market in agriculture. Solutions that off the best blend of value and simplicity will win. Ag retail already has the largest ag dataset. This is an opportunity to extend the value of this data and lead this new industry.

The Risk of the Status Quo The new technology has the potential to change the way farming is conducted and the way agronomic advice is provided. The balance of power for agronomic advice may shift to the seed / biotech companies from local ag retailers. New service opportunities will become available to those ag service providers who are far-sighted. Growers who adopt and integrated advanced technology will have increasing advantages for growth and profitability. Technologies that prove valuable may be become required risk tools in order to access capital or insurance programs. Adapted from Big Data Project Report. 2014. The Hale Group, Ltd.

Big Data & Big Opportunities Dr. Matt Darr, Iowa State University The over-all point is that new technology will not necessarily replace old technology, but it will date it. By definition. Eventually, it will replace it. It's like people who had black-and-white TVs when color came out. They eventually decided whether or not the new technology was worth the investment. ~ Steve Jobs, former CEO of Apple Inc.