What is data and what is the fuzz all about?
In this Agtech Letter we will dive into the matter of data. You often hear people talk about data and it is one of the corner-stones of agtech solutions. The success of digitalization, automation, AI solutions, and autonomy are all heavily dependent on the availability of high-quality and relevant data.
"Data is the new gold" (or is it, really?...)
During the lats 5-10 years, you probably have heard expressions such as “data is king”, “data is the new oil”, or “data is the new gold”, and with expressions like these some questions arise naturally: Is this true?, Why is data so important? and, maybe most importantly, What is data?
Here, we do not need to provide a scientifically correct description of what data is, but we instead aim at providing a description that give you an intuition about data and what to consider. Hence, for now it’s enough to think of data as something that contains information about something in a concrete format. Splitting differences on details of the definition doesn’t help our understanding and is not needed to create value in your business using data.
Data can exists in an analog format such as hand-written notes, hand-made graphs and old maps of for example drainage pipes. They can also exists in digital format such as measurements from grain drier automation systems, satellite images, and the amount of milk produced each day in a milking robot.
A computer always uses data in a digital format, but sometimes the digital data can be a digitalized version of some analog data source such as old maps or hand-written notes.
A computer always use data in a digital format, but the data can be a digitalization of analog data from older sources as well.
Some important properties of data you should know about
Data comes with many different properties that can be good to know about when collecting and working with data in your business operations. As usual, the technical details are not important here, but rather the big picture.
Data can be of different TYPES
As already mentioned, data can describe different types of information coming from different types of sources.
Some examples of different types of data are sensor measurements where the data can be for example temperature readings, financial information where data can be for example revenues from grain trade, and personal data about who owns a property or company.
Data can be of different QUALITY
The data available to us from different sources can be of very different quality and sometimes there are also lots of missing data as well. Different levels of data quality often means that different amount of time and effort is needed to process data and make it useful for you in your business. Sometimes this processing is done by you as a user, but often there are good tools and services for this from the companies selling software systems.
As an illustrative example, we can use a price list for different crop varieties. If the data is complete and with high quality, there is a given price for each variety in the list to give you the information about the costs for different types of seed. However, if the price list is printed using ink that is fading and the price for some varieties are missing, while the variety name is missing in other places, then we have low quality data. In the first example (left image below) the price list can be used directly for decisions on costs for seed, whereas the second example (right image below) will need processing and cleaning before it can be used properly.
Data can be of different VALUE
Depending on the type of data you have, how it has been collected, what it representes, and a lot of different other aspects, the available data is of different value to the one that is using it. This is largely due to the fact that the contained information can vary a lot between different types of data, and it is only the useful information found in data that is valuable and not the data itself.
As an extreme example to illustrate the concept of different values for different users, let’s consider an satellite image using the NDVI vegetation index to determine the crop status on a field. If the satellite image is captured for the correct field, then it contains lots of important information about the crop and can be valuable for the farmer. However, if the satellite image is captured at the wrong field the data contains no useful information and is more or less useless to the user. Hence, the same satellite image (data) can be very useful to one user and close to useless for another.
Data can be of different RESOLUTION
The resolution of the data can be though of as the distance in time or space between two different measurements. This is a very simplified notation of resolution but works perfectly to give you an intuition on resolution that is sufficient for understanding and using data in your business.
Resolution in time
The temporal resolution, or resolution in time, is simply how often a measurement is done or other data is created. It can be once a year, once a day, once a second, or any given interval in time. In the image we show an illustrative example where the temperature is measured more often (light green) and more seldom (dark green). Hence, the light green data is measured at a higher resolution in time than the dark green.
Resolution in space
The spatial resolution, or resolution in space, is similarly the distance between different measurements in the data. Depending on the data, this could be 100s of meters, meters, centimeters, or any other given distance. It all depends on what is measured and how it is done. As an example one can consider a soil sampling of a field to determine soil properties. Often, one sample per hectare is collected, but it could be both more sparse and more dense. In the image an illustrative example of a field where soil samples are taken more densely (light green) and more sparsely (dark green) is shown. The data from the light green samples have a higher resolution in space than the data containing the dark green ones.
Data can be of different FORMAT
Data can be stored in many different formats and for data that is stored digitally, which is the most interesting to us, there are some commonly used formats that can be mentioned here. The details are not important here, but rather the understanding that data can be stored in many different types and format.
Commonly used formats for storing data is as text, as images, as movies, and as numbers in a spreadsheet (such as Excel and others). Each type is suitable for different applications and it is the developer of a software system that determines what type of format that is used. Either way, it can be good for you to know the different types of data formats that are used by the systems you use in your operations.
Data fuels "everything" within agtech
At the beginning of this Agtech Letter we asked ourselves the question about Why is data so important? The short answer is that it is a necessary input for more or less all technology solutions in agtech (and other businesses as well for that matter). Solutions based on technology from automation, autonomy and artificial intelligence, to simple computations in a book-keeping software are dependent on the availability of data in the right format, having enough quality and resolution, containing the right information, and stored in the correct format.
In some sense, a good intuition is to view data as the fuel that is needed to run all agtech solutions and without the access to data it would be impossible for many modern technology solutions to function at all. If we look into the future, our conclusion is that the need and importance of data will continue to grow fast, and that the value of your data will become very valuable.
"Data is the new gold ore"
Given the discussion in this Agtech Letter, we can state that expressions like “data is the new gold” are not really true. In reality, data can be more similar to gold ore (or crude oil), where the ore contains different amounts of gold (information) and needs processing and refining in several steps to turn into gold (value).
Remember that using the right tools and processing steps, gold ore can be extremely valuable. The same holds for data!
Agtechers' Actions
Collecting data in a structure way is like building a stack of gold ore just waiting to be turned into gold. By collecting the right data about your business operations and using the right tools, you can create lots of value today – but even more so in the future.
In the next Agtech Letter we will present some common types of data used in crop farming.
