The agri-food sector is also aware of it: the growing availability of large amounts of data everywhere and in any case collected and made accessible, as well as the increasingly sophisticated production of algorithms that permit strategic and analytical use of such data, are incredible potential resources for developing applications, models and systems that help different classes of users (consumers, entrepreneurs, policy-makers) to make conscious decisions on quantitative bases.
However, the collection and transformation of an increasing number of data into strategic and usable information is not a process in itself sufficient to ensure greater rationality in economic decisions: it must be accompanied by growth in the capacities of the various economic operators in the agri-food sector (from institutions to companies) to know how to make full use of the information generated.
This is precisely the focus of “The Impacts of Digital Transformation on the Agri-food Chain” study carried out by the Digital Transformation Institute and Cisco Italia, which, in particular, aimed at understanding how digital transformation is affecting agri-food chains in a country like Italy, in which this sector is one of its main strengths, and what is the current state of the art in the field of digitization of agriculture, the agri-industry and the agri-food sector in general.
The agricultural and agri-food sectors are ideally suited to take full advantage of the potential offered by the availability of data to individual businesses and the production chains in which they are involved.
Precision Agriculture
In Precision Agriculture (PA), as clearly defined by MIPAAF Guidelines, data analysis is strategic, given that management uses information technologies to acquire those needed to improve decision-making for agricultural production.
The purpose of Precision Agriculture is to create a rational system for the use of environmental resources (soil, water, fertilizers) and to correlate them with the specific needs of each crop. This process therefore enables minimizing environmental damage while simultaneously raising the quality standards of agricultural products. Its practical application is now possible thanks to the use of advanced techniques and computer monitoring systems, through which information on crop and soil conditions and their variations can be obtained automatically, intervening in real time through specific and efficient improvements.
Since the early 1990s, PA has increased rapidly, largely favored by the availability of a technological structure divided into three levels:
- geographical positioning (GPS)
- geographical information (GIS)
- use of sensors.
GPS is a satellite radio navigation system that can provide position information in the three spatial directions (x, y, z), speed and time. GPS receivers use signals from four or more sight satellites to calculate the user’s position, speed, time, and other data needed for agricultural applications. Use of DGPS differential techniques makes it possible to calculate the correct positions (x,y,z) in real time with high precision, without having to correct the data after they have been recorded. This type of GPS is useful for performing operations such as high resolution mapping of harvests, parallel trajectory guidance, soil sampling and variable rate application (VRA) distribution of fertilizers and pesticides or automatic guidance of agricultural vehicles.
On the other hand, the geographic information system (GIS) is a software application consisting of several modules designed to capture, control, integrate, process and represent data that are spatially referred to the Earth’s surface and that spans vast territorial spheres. GIS software programs are available with a wide range of capabilities and features, but all are capable of showing geo-referenced data in graphic form. Adequate co-recording of data can be quantitatively analyzed using geo-statistics. A basic principle of GIS is that for being used together, the different layers of a map must be based on the same system of coordinates. All spatial data files in a GIS are therefore geo-referenced with a GPS system.
Finally, Precision Agriculture is basically based on detailed knowledge of the spatial variability of the main soil properties. Sensors, which can be divided into “remote” and “proximity sensing”, or even simply “aerial” and “terrestrial”, come into play for the capture of data. In the case of “remote sensing”, passive sensors are normally used; these exploit sunlight by calculating the amount reflected by vegetation with obvious limits to the use of remote monitoring (for example, differences in reading in case of differences in transmission of the signal in the atmosphere).
The creation of maps (GIS, pedological, etc.) with an adequate scale of detail can draw on more economical and efficient measures thanks to the development of technologies of different types of proximity sensors. By means of expeditious and low cost measures, these are able to provide very detailed mapping of soil variability at different depths. “Proximity” refers to those sensors which measure at a minimum distance (1-2 meters) or through direct contact with the ground. Compared with airborne or satellite sensors, they have the advantage of being more precise and easier to use. However, there are major limitations to wider development of these sensory techniques, which are mainly related to the difficulty of interpreting dati. Given that they are sensors which measure a set of environmental variables, for example humidity+clay+stoniness+salinity, it is sometimes difficult to automatically model a map of a particular pedological feature of interest. There is, for example, a need for a high level of pedological expertise to effectively interpret data measured by sensors, but that does not mean that steps in this direction are not being taken.
The support of Big Data Analysis
Thanks to Big Data Analysis techniques, it is possible to aggregate and monitor the large amount of data generated by sensors and to produce concrete information as output for both producers and final consumers.
The availability of this technological structure allows the systematic application of Precision Agriculture which envisages four implementation phases:
- monitoring of data (environmental, productive, pedological, mechanical, etc.)
- analysis
- decision/action.
The aim of these four pillars is the sustainable management of resources (fertilizers and nutrients, seeds, plant protection products, fuels, water, soil, etc.) through control of the machines that manage them.
Through the rational use of decision factors, Precision Agriculture thus helps and facilitates operators in planning their activities, time for implementation of crop interventions, repetitive tasks and intensity, reducing and eventually ruling out the possibility of error and, consequently, increasing production efficiency. The main problem of its development in the Agri-Food industry lies in the lack of homogeneity of the latter, which is felt in particular at the various stages in which the raw material is transformed into a product ready for the table.
In this context, new technologies open the way for homologation of the chain, starting from the field, from the moment of pre-processed product, up to reaching the consumer in the form of an interface that can be used via mobile or desktop devices to trace back to all processes in the chain.
Achieving this type of output requires that the enormous amount of data collected, for example, by sensors placed along the whole chain, be transformed into concrete knowledge, in order to match the material value of the Agri-Food product with the immaterial value represented by supply chain information. The data available must reach an accumulation point where they are analyzed and interpreted, to then move on to enrich the decision-making processes of large-scale retailers and consumers. In fact, it is a real “fil rouge” of data which, aggregated through the logic of Big Data, makes it possible to build “intelligent dashboards” for decision support systems. Specifically, a company with such intelligent dashboards has the opportunity to see the whole process of transforming matter, from source up to distribution. The same goes for the consumer who, through to a subset of this information, is able to distinguish the value of what he/she buys. This is not just a question of a product description but of going more into depth, perhaps through the presence of barcodes on product packages, regarding the interaction that transformed the raw material into the product on our tables.
However, it clearly emerges from the research carried out by the Digital Transformation Institute and the focus on the wine sector – although the result can be extended to the entire segment – that the companies investing in digital technologies realize the benefits and measure the return on investment also in terms of product quality improvement, while those that do not invest often do not do so because they do not perceive the need to invest in digital technology.
In fact, compared with the 52% of winemakers intending to make an investment that is more than a minimum threshold of € 5,000, 31% of small and medium-sized companies declare that they have no intention of making such investments in the future. If this is combined with the fact that there is a direct correlation between size of investment and product quality in the wine-growing sector, there is significant evidence that the companies characterized by quality production have an awareness of the role of digital which will help them produce better results and quality, while others will risk becoming less competitive in an increasingly demanding market. And this is a risk that no company in the Agri-Food industry can afford to run.
Stefania Farsagli
