Collect data in fruit areas
All new technologies in the agricultural sector are based on the analysis of data. But there is a well-known rule under analyses: "Garbage in, garbage out".
Then artificial intelligence came along: machine learning, Big Data analytics, unstructured data. The promise was that the algorithms would figure everything out on their own; you just throw the data in, and they spit out the findings again.
But here, too, the old truth prevailed: "Garbage in, garbage out".
Sorry to be the bearer of bad news, but unless you manage your data properly, you have no chance of successfully using any of the smart technologies in precision agriculture.
Data is difficult to handle
The problem of data management is not unique to agriculture, but can be found in all industries:
Data is stored in silos: spraying records are in a database provided by the cooperative; crop records are in a spreadsheet and recorded in the wholesaler’s receipts – and the two sources don’t match; pictures are stuck in the family album; notes are on a memo pad; findings and recommendations from the agricultural engineer are noted in a printed form.
Data formats are not harmonized. Even simple things like dates get mixed up because they are logged differently in different computer programs. Units for areas, lengths, application rates don’t match up.
Data is not meticulously collected (since it won’t be used anyway). You forgot to document some field work, and you don’t remember exactly when the first flowering was last year. You’re pretty sure last week’s spraying is noted on the note that was on one of the tractors when you last saw it.
Make an effort to improve your data management
Solving the data problems is not that difficult. However, it requires a bit of determination and structure:
Data must be recorded. Therefore, do not forget to document things – what is not documented did not happen. And if it didn’t happen, you can’t learn from it.
Make sure that the formats are consistent. If you divide your farm into plots and you name them, stick to it and stick to it. If the name of the plot is "green orchard" in one system, but you call it "green orchards" elsewhere, the first problems already arise. Computers don’t handle inconsistencies well.
Store everything in a central location. (This is not a must, it just makes your life a little easier). Make sure it’s one that helps you share that data with those who need it to provide future services. For example, if you’re using a drone in the field, how do you tell it where to fly?? You cannot talk to her. But if all your relevant acreage data is already in the right place and in the right form, then it acts as a translator for you between you and the drone.
It all sounds easy in theory, but it’s hard in real life. My own fruit farm, which I run with two partners, is now in its eighth season, and we still don’t quite have our data management down pat. We farm 50 acres, divided into 32 plots, with five different crops (apples, plums, pears, strawberries, raspberries) and different varieties. While we have estimates and approximations, we still can’t conclusively say which culture mix is the more lucrative business model for us to use. In addition, we struggle to allocate the hours of seasonal workers to these areas and varieties. And when our accounting system reports certain costs for pesticides, we don’t quite manage to allocate them to fields and crops.
The plant timeline
In the field of data analytics, timestamp is one of the most important attributes. When something happened? If you want to save the first bloom as an observation for later and take a photo of it, it’s not much use if you don’t know when it happened, right? The rest of the world is happy to use the Gregorian calendar for this; we farmers are not.
Plants don’t really care much about dates. They care about temperature. The plant’s calendar is measured in temperature days. Therefore, we need to store all records with two time stamps: one for us humans, so we know when something happened, and the other for the plants, so they know when something happened. We call this the phenological development stages. Saving this data for all data points collected on the farm makes a huge difference in future analyses.
Precision agriculture is a journey, and data management needs to be at the forefront of the first leg of the journey. Even if it may seem boring, the reward of the effort will pay off. From good data come many new opportunities. The first one we will examine in more detail in the next blog post.