What's the BIG DEAL with Big DATA?

Originally posted in Irrigation TODAY, April 2019

Big data is something you have likely heard or read about. Maybe you have even been encouraged to incorporate it into your operation. This Q&A with Todd Janzen, attorney, writer and speaker knowledgeable in ag tech topics, gets down to the basics of what big data is and how growers can use it.

Q: What is big data?

A: Big data is a collection of aggregated data sets. Think of it like this: There are many tools out there already that collect data, whether it is our cell phones or the thermostat in our house or google — you name it. When you start to pool everybody’s collective use of those tools that collect data and analyze it, that’s when you really start to talk about big data.

A good example of big data is a phone app that shows where traffic is stuck or bottled up. Everyone’s phone is reporting how quickly they are moving on the highway, and the app collects that information and tells those using the app where people are slowing down in certain areas. You can start to draw big-picture conclusions from all of the little data sets that are going into the aggregate data set.

Q: How can big data be used in agriculture?

A: Here’s an example: Say you plant a certain type of hybrid from a certain seed grower. As a grower, you know how you raised that crop, what your yield was and all of the different attributes of the harvested crop. If everyone who planted that same variety uploaded that information to the same platform, then you can start to do some benchmarking on that seed variety to figure out optimal performance characteristics, for example, if the seed performed better in dryer climates or with certain amounts of irrigation. Collectively putting together all of those farmer data sets into one giant aggregated data set allows you to learn things about that crop that you otherwise would never know or that you would learn only through isolated trials.

Q: Why should growers care about big data?

A: The promise of big data is that it will lead to reduced inputs for crops and increased yields. It should also reduce a grower’s environmental footprint because they will be able to do more with less.

Q: For a grower, what is the first step to get into using big data?

A: When you are talking about big data, there is strength in numbers. If I was a grower, one thing I would want to know is, “How large of a data pool is out there that is the same as or similar to my farm.” If you are the first farmer in a big data set, then it is just your data. If there are multiple farms that are similar in soil type, climate, etc., then you are going to get better results.

From a data privacy standpoint, I would also want to know what steps a company takes to protect my information that is strictly related to my farm. How do they anonymize my data so that it isn’t shared more widely in the data pool?

Many companies that growers are already working with (such as equipment manufacturers) offer these types of data collection services. There are also a number of startup companies that help growers collect and analyze data that provide a different service to farmers that doesn’t include equipment promotion. I would suggest a grower sign up for a year trial with a company to see if they get results back that pay for use of the service. If not, try someone else.

Q: What do you see for the future of big data in ag?

A: I think we need to focus on two things right now — one of those is collecting good data. Because so much of the data collection that is happening right now is reliant on human input, farmers have to manually upload their information about how a crop is grown. This may not be done consistently, and you may not get good results from everyone. Something that I hear over and over is that the quality of agronomic data is still poor. If you put junk in, you get junk out.

Part two is to focus on building better analytics. Initially, there was a lot of promise that the analytics would be mindblowing, and we were going to learn things about crops that we had no idea about. I’m not sure that promise has been delivered yet. But with better quality data, I think the analytics will get a lot better. It’s just going to take some time.

Q: Final thoughts?

A: Being a good farmer is always based upon how much information you can derive from what you have done in the past and then using that information to shape your decisions going forward. To the extent that you can harness data to help with future decision-making, then using it is absolutely a good thing and should be a real benefit to farmers.

When dealing with big data, you are really only looking at what has been done in the past — what has worked and what hasn’t worked. It isn’t always going to give you suggestions about how to try something new because your datasets are all based on past decisions. So you still have to experiment or do things out of the ordinary to see if they work.

Mark Sullivan