The introduction of innovative technologies in agriculture has been studied for decades. Formerly, several studies have delved into the topic trying to understand which are the barriers or drivers to technology diffusion [1]; latterly, others have tried to model the adoption and implementation processes themselves [2]. The main purpose of these studies was not only to encourage their diffusion, but also to foster a change in business visions and in the way production activities are organized. This effort can also be seen in the institutional support, which over time has characterized itself as a promoter of technology diffusion and facilitator to close the gap between provider supply and user demand [3]. A paradigmatic consequence of this is the funding policies implemented by global agricultural policies, such as the Farm Bill in the United States (US) and the Common Agricultural Policy (CAP) in the European Union (EU). It is precisely the latter, in its latest programming, 2023–2027, that has paid special attention to the role of innovations as a fundamental tool to achieve the ambitious goal of “producing more, polluting less”. In addition, a major chapter has been dedicated to the development of knowledge systems, or better known as Agricultural Knowledge Innovation Systems (AKIS), precisely to foster the diffusion of innovations, linked to economic support measures in rural development plans with the aim of simplifying access to technologies [4].

Among the innovations in agriculture, precision farming (PF) certainly plays an important role. PF originated in the US in the late 1980s and 1990s, then it spread globally. Conceptualized as “Precision Agriculture” or “Site Specific Management”, it is also described as “Smart Farming” and “Digital Farming” as it is based on the use of smart technologies in agriculture and digital data management [5]. PF applies principles, technologies, and strategies for differentiated management of internal plot variability, studied considering the interaction between the spatio-temporal component, the type of cultivation, and farm-specific agronomic management [6]. Farming can be compared to a dynamic system, whose qualitative-quantitative production depends on the use of techniques that allow a variable application of inputs according to the actual needs of the crops and the chemical-physical properties of the soil. Underlying the application of PF tools is a preliminary study of both spatial and temporal variability, with the aim of identifying and quantifying the intensity of one or more parameters. In this way, “homogeneous zones” are identified, thanks to which differentiated management is implemented within the plot [7]. The resulting advantages are related to a better optimization of outputs, input rationalization, cost reduction, and environmental benefits [8].

Precision farming encompasses the use of numerous technologies that have spread to different parts of the world with varying degrees of use. Studies that have attempted to measure the take-up rate of different technologies in different countries tend to show that the adoption is higher for guidance (such as Global Navigation Satellite systems) or recording technologies (such as soil and yield mapping), rather than reacting ones (such as variable rate nutrients, seeding, and pesticides) ^[9][10][9,10]. Furthermore, many studies ^[10][11][10,11] confirmed that, on average, North American farms are more likely to use VRT than European farms.

2. Theoretical Background

Variable rate application is a technology that finds application in agricultural operations, from tillage to harvesting. VRT “allows precise seeding, optimization on planting density and improved application rate efficiency of herbicides, pesticides and nutrients, resulting in cost reduction and reducing environmental impact” [12] (p. 13). Other applications are recorded in the field of weed control treatments, gypsum/lime application, and irrigation ^[13][14][15][13,14,15]. VRT represents the idea of precision agriculture well, by managing primary production based on the needs of the soil, the land, and the crops that are grown. VRT can be map- or sensor-based [15]. VRT based on the use of prescription maps (produced before the operations) varies the amount of product to be distributed according to the information in the prescription maps, which are the result of data from different acquisition systems (e.g., yield map, agronomic indices, satellite images, meteorological data, soil sampling, etc.). The other VRT methodology uses “on-the-go” sensors, which detect in real time during the operation the chemical characteristics of the soil and the phenological state of the crop. These data are sent to the reprocessing unit from which feedback is given to the actuator on the amount of input to be spread. Distribution therefore takes place by homogeneous zones, each of which corresponds to a precise dose: the on-board computer controls the actuator (hydraulic or electric), which will modulate the opening damper or the volumetric regulation system based on the different management zone [16]. Even though there are no global assessments of VRT use rates in the literature, Finger et al. [10] and Maloku [11] confirmed that, on average, North American farms are more likely to use VRT than European farms. In North America, on average the rate of use is 17% greater than in Europe, according to Nowak [17]. In the European market, Germany, Denmark, and the Netherlands were first countries interested to use this technology [18], while the Mediterranean countries have only recently begun to see the introduction of these instruments. Delving into the literature, most studies have been conducted in America. In the United States ^[19][20][19,20], VRT utilization rates hardly exceed 40%. Other studies, although residual, have been conducted in Florida, Alabama [21], and Kansas [22], where higher utilization rates are also recorded. Some studies have been carried out in the UK with utilization rates from 8% [23] to 16% [24], even based on the different variable rate application (fertilization, seeding, etc.) for different production orientations [25]. Other studies have been conducted in Australia, with variable rate technology utilization rates averaging 20% [15] or higher [26]. In Europe, rwesearchers find studies conducted in Germany, Sweden, France, the Netherlands, Belgium, and Denmark [11]. Reichardt and Jürgens [27] reported during 2001–2006 that approximately one out of five PF adopters used VRT in Germany. In Denmark, the rate of VRT use across studies ranges from 7% to 37% ^[26][27][28][26,27,28]. According to Nowak [17], the variable rate application has grown at a slower rate compared to other PF technologies. Sunding and Zilberman [29] state that there is a significant latency between the introduction of an innovative technology into the market and its widespread use by farmers, so its adoption is not immediate. Several types of barriers to the adoption and diffusion of technological innovations have been cited [30], some of which focused on VRT ^[13][14][15][13,14,15]. The first barrier is economic and can be attributed to the high initial costs and subsequent training and tool implementation costs that end-users should bear ^[31][32][31,32]. Indeed, studies identified larger operators as more willing to adopt VRT given their capacity to absorb costs ^[33][34][33,34]. However, researchers ^[1][26][35][1,26,35] also showed a positive association between farm profits and VRT adoption, also underlined possibilities to reduce costs (i.e., when adopting VRT together with soil mapping). In addition to economic barriers, there remain socio-economic, organizational, institutional, behavioral barriers ^[36][37][36,37]. Innovation can be influenced by socio-economic factors such as the user’s age, education level, gender, and degree of information ^[38][39][38,39]. Younger, better-educated, more knowledgeable about the costs and benefits of PF, and more optimistic farmers were more likely to utilize VRT [16]. Furthermore, the business organization and work intensity could not be compatible with the application of new technologies [40]. Limits to VRTs adoption have been technical issues related to equipment and software, access to services, and lack of compatibility of equipment with existing farming operations ^[15][31][15,31]. The institutional context itself can influence these choices. Literature describes physical barriers related to the agroecological context in which the new technology might operate. In fact, different VRT adoption rates can be identified based on the different agro-meteorological characteristics of an area ^[15][41][15,41], as well as on the basis of the type of cultivation [33]. Cultural barriers (i.e., habits, consumer choices, market uncertainty) ^[42][43][42,43], as well as limited institutional support [44] have been identified. Subsidies, as well as more indirect interventions such as information support can lead to increased adoption of VRT ^[1][34][1,34]. For example, Evans et al. [14] confirmed the importance of economic incentives to motivate growers to move to higher levels of variable rate irrigation adoption. Vecchio et al. ^[37][39][37,39] also describe barriers related to the cognitive sphere, emphasizing the importance of the farmer’s perception, which is now no longer linked only to risk appetite or the expected benefits of technologies. Indeed, it is with the term “perceived complexity” that these authors encapsulate the most influential barriers in the adoption process. Other authors recall how the farmer’s perception in fact produces a positive or negative attitude towards adoption [45], which is often shaped by the socio-economic characteristics of the individual [46] and the social systems in which the technology operates. In this sense, more research to understand group behavior and collective action is required ^[44][47][44,47].