Implications of Collecting Data on Farm

“Without data, you’re just another person with an opinion.”
W. Edwards Deming

There’s a famous case study in the data analytics world about a well-known American department store chain that learnt to predict when its female shoppers were pregnant by their buying habits. The store discovered that pregnancy signalled a significant shift in female spending habits, so by targeting pregnant customers they could get their attention at an impressionable time in their lives.

This chain of stores got so good at doing this that they sent pregnancy related advertising material to a teenage girl who had been shopping at their stores. The father of the girl was incensed by the material and told the store so, only to have to apologies days later after learning the store had not in fact, made a mistake (Hill, 2012).

As farmers we deal with something that is every bit as complex, dynamic and intricate as human buying habits. We deal with nature and the natural world. Farmers have never before had tools to measure and respond to the natural variability that comes with interacting with the natural world. But the rise of data collection and technology allows us to do just that. Our farms and our soils are not homogenous, but for too long we’ve treated them as such.

What will the information revolution mean for agriculture? What new approaches, what new ways of thinking do those of us on agricultures front line need in order to adapt our mechanised industrial agriculture into the new reality of the information age?

A new revolution requires new ways of thinking and new approaches to some old problems in order to prosper as a farmer on a data-driven farm. What new approaches, what new ways of thinking do those of us on agricultures front line need in order to adapt our mechanised industrial agriculture into the new reality of the information age?

The rise of a myriad of cheap sensors is combining with the GPS and the promise of near ubiquitous internet access to allow farmers to ask questions about their farms that haven’t been feasible to ask in the past. Rather than treat their farms and soils as homogenous farmers can become flexible and adaptable to the natural variations that exist in their environments. Never before have farmers had tools to measure, quantify and respond to the natural variability that exists on their farms like they do today.

Collecting data on farm has transformed from an expensive and laborious process that few farmers could be bothered with, to one that is relatively cheap and increasingly easy. Rather than being a one-off process, collecting, analysing and continually reviewing data can become a system for ongoing improvement on a farm.

The four data revolution steps encompass the challenge of precision agriculture are:

Collecting
It is possible to collect farm data on virtually any area of farm performance. If there’s an area of a farm that a farmer/ farm manager believes can be managed better, then data can be collected to aid in assessing and quantifying an issue. The spectrum of ways to collect data is as wide as the number of issues are on farm to collect data about. From very simple data logging via a smartphone or laptop, through to networks of remote monitoring sensor networks, collecting reliable data is no longer the challenge it once was.

Processing
Once collected, data must be processed into a format that is useful for farmers and advisors to use. Typically, this involves converting large datasets of data into a visual medium, better understood by humans. Examples of this may include graphing weather data, or laying a large series of plant health data points onto a map to give a visual representation of plant growth. Traditionally requiring some expert knowledge to operate, data processing tools are becoming increasingly automated and easier to use, leading to lower barriers to farmer adoption.

Interpreting
With data collection becoming increasingly commonplace on farm and automated processing of that data becoming increasingly automated, it is the interpretation of these pretty maps and well laid out graphs that is the challenge for farmers. Providing an appropriate context to what data points are being analysed and how that is impacting farm performance is crucial to making the most of data collected on farm. Often farmers engage trusted advisors or other outside professional help for this step of the process.

Application
This is the step that makes the previous steps worthwhile. With a newfound understanding of what is happening on farm and some quantification of the issue being investigated, a farmer now has the confidence to change management decisions based on the interpreted data. Making better and more informed decisions on farm is what justifies the expense and time required to capture better farm data. But has the implemented management change had the desired effect? There is now a need to collect data relating to the different farming or management technique to verify that improvements are definitely being made. So the cycle of collecting, processing, interpreting and applying data based decisions is an ongoing one. Each time this cycle is attempted, a farmer learns something more about their operations and how to improve what they do.

Rather than being a one-off process, collecting and appropriately using farm data can become a system for implementing continual improvement on a farm.

Conclusion
Once accurate data is being collected at the farm level, such data can be aggregated and compared across different businesses, regions, and countries. Farmers can use this aggregated data to analyse farm business performance. The promise of this is the potential for real-time business bench marking.

Third parties, including well known agribusiness multinationals are becoming interested in farm data at this aggregated stage, because it gives insights into how farmers are using various products. This leads to a strange phenomenon where a company’s clients are also doing their product research.

There are many people who believe there is much value to be extracted from this data as evidenced by the venture capital flowing into new companies attempting to make use of it. This may be concerning to farmers who may not understand the motivations behind a company wanting to access farmer data.

There are other longer term implications of data technology in agriculture. Fears about commodity market manipulation may be overstated but concerns about control of data access are valid. Like all technologies there are potential benefits to farmers as well. More open supply chain data may allow for cheaper inputs and potentially even a new revenue stream for some farmers. It will certainly lead to better genetics and machines for farmers to use.