Precision Agriculture Implementation Factors and Adoption Potential: The Case Study of Polish Agriculture

Abstract

Some of the current economic, social and environmental challenges could potentially be addressed by Precision Agriculture (PA) introduction. However, the pace of PA introduction is found to be slower than expected in developed, transitioning and developing countries, with the PA adoption literature is predominantly been focused on research on PA adoption in developed world. This paper addresses these shortcomings by identification and explanation of PA adoption factors and compilation of the regional ranking of PA adoption potential for 16 Polish voivodships. It contributes to the PA adoption factors’ literature by filling the gap on under-researched transitioning economies using Poland as a case-study. The key PA adoption factors were identified by Systematic Literature Review (SLR) based on the final sample of 21 papers from journals indexed in Scopus or Web of Science databases and were organized in 5 groups (socio-economic, agro-technological, financial, technological, and informational factors). These factors formed the conceptual framework for the ranking of PA adoption potential of 16 Polish voivodships based on the application of Sturgess rule. The analysis of PA adoption potential of 16 Polish voivodships shows the clear “core-periphery” divergence: i.e., well-developed metropolitan areas got the highest ranking and less developed peripheral regions were ranked low.

1. Introduction

Coming XXI century poses multiple economic, social and environmental challenges for exponentially growing population constrained by the limited resources availability and environmental depreciation. In the context of agricultural industry perspective development, Precision Agriculture (PA) is perceived as a potential solution for the above-mentioned problems by means of the cutting edge up-to-date technologies, as the application of PA disruptive technologies for agriculture (based on AI, Big Data, IoT, GNS, VR, and blockchain) would allow to mitigate the economic and social challenges for growing population and substantially improve its ecological well-being [1,2,3,4,5,6].

Despite it considerably short track-record history counting only 20 years [7,8], PA is already found to be contributing to several UN-defined Sustainable Development Goals (SDGs), like e.g., No Poverty, Zero Hunger, Decent Work and Economic Growth, Industry Innovation and Infrastructure, Responsible Consumption and Production, and Life on Land [9].

The pace of PA implementation is considered to be generally slower than expected [8,10] and uneven across the globe, been more fast in developed counties and very slow in transition and developing countries. Also, the prevalent part of academic literature is focused on analysis of PA adoption in developed countries [11,12], due to the faster pace and bigger scale of PA implementation in these countries. The combination of these two factors rise the urge of extended research of PA adoption factors and shedding some more light on PA adoption processes in transition economies.

This exploratory research is an attempt to fill in both those gaps by studying the factors significant for PA adoption and doing it in a context of the Polish transition economy.

The aim of this research is two-fold: (i) to identify and summarize the key factors of PA adoption, and then, (ii) make an empirical assessment of PA adoption potential (based on identified key factors) for 16 voivodships in Poland. The Systematic Literature Review (SLR) methodology was applied for key factors identification and then, the ranking list of PA adoption potential of Polish voivodships was compiled using the Sturgess rule.

The paper contributes to the region-specific PA adoption factors literature [13] by filling the gap on under-researched transitioning economies. The novelty of the paper is based on (1) going step further from the ordered taxonomy of PA adoption factors and expanding it to the systematized explanation on why these factors are important for PA adoption, and (2) compilation of Polish voivodships’ PA adoption ranking, which is the first ranking of Polish PA adoption potential made in a country-wide Polish context.

The paper is organized as follows. First, “Theoretical background” section provides the overview of economic, social and environmental benefits of PA (justifying the need for the faster implementation of PA), describes in more details the PA adoption literature gaps, and formulates the paper’s aims. Then, “Methodology” section outlines the methodology process, instruments and data. Next, “Results and Discussion” section describes research results and puts them in the context of existing research and potential use for stakeholders involved in PA adoption. Lastly, the “Conclusion” section finalize the paper by brief outline of work conducted, filled gaps, and future research directions.

2. Theoretical Background

Precision Agriculture (PA) is believed to provide substantial economic, environmental and social benefits [4,5,6,10,14,15]. In a very general sense, PA “allows farmers to produce more output with less inputs” [13] p. 2, leading to increase in productivity and profitability [4,5,16], which implies the significant economic impact.

PA allows reducing inputs use by applying more precise site-specific technologies, which reduce the inputs’ use per se and are also reducing the overlap use of inputs, saving money on inputs costs and increasing the net returns [13]. Site-specific application of inputs and more efficient control and operational technologies, which are able to be functional on 24/7 basis, allows for the more precision timing [17] for almost all agricultural operations, leading to increase of agricultural output. Other economic benefits of PA are gained from the improved management efficiency [18], reduced farm’s operating costs [16], and income-diversification opportunities [17].

PA contribution to reduction of environmental impact of agricultural production is mainly based on reduced use of fertilizers, pesticides and herbicides, more efficient use of water, (both of these factors contribute to slowing down the soil degradation process), and reduction of CO2 footprint of agricultural production due to the proliferation of green-based electric-powered technological solutions instead of fossil-fuel-based technologies [4,13].

The social benefits of PA adoption are numerous and the most crucial of them are related to achieving the food sovereignty, livelihood improvements, and narrowing the technological divide between the countries. The increased crop production (compare to the traditional agriculture) would help immensely in nourishing the people in developing countries and diversifying the crop production for the rest of the world [9,16].

The improvement of rural livelihoods is based on (i) inclusion in the agricultural production previously abandoned areas (e.g., remote areas, steep slopes, or soft soil areas) due to use of PA site-specific technologies for marginal areas operations [17]; (ii) increasing agricultural output with the reduced use of inputs, lesser soil degradation and water usage, and without substantial compromising the existing natural environment [9]; and (iii) creation of the new businesses and employment opportunities in rural areas due to the need of introduction and maintenance of new cutting-edge technologies, which would have a positive impact on the rural–urban migration dynamic [9].

PA introduction will also make its impact on increasing the youth employment in the rural areas, making the work in agriculture more appealing to the young people due to the reduction of harsh physical labor, repetitive tasks and introduction of up-to-date “fancy” technologies into been quite “conservative” for a long time agricultural industry, which would contribute to the reduction of the rural exodus [17].

Closing the technological divide between the developed, transitioning, and developing) economies would be possible due to the “intrinsically adaptable” nature of PA technologies [9], which allow the adoption of PA in different contexts and countries, therefore providing the possibility to leapfrog from outdated technologies (based on manual labor for subsistence farming in some countries) to commercial industrial-scale agricultural production based on contemporary PA technologies [9].

This potentially enormous impact of PA on economic, social and environmental aspects of human life, rises the urge of smooth and fast introduction of PA technologies into agricultural production, and studying the factors affecting its adoption is considered to be very important issue [13,19], thus providing the fertile ground for the different streams of PA “adoption literature [13], which this paper is contributing to.

Despite the boosted field of PA adoption literature, there is still any commonly agreed academic consensus on what PA exactly is (which technologies and processes it covers) and therefore no precise and overall agreed definition of it. Most commonly it is defined as “digital”, “site-specific”, and “smart” agriculture [11]. For the most part, these PA definitions are overlapping with each other, but sometimes they could contradict in some minor details, depending on the type and purpose of the research they used for.

For the purpose of this research, the definition of PA, as it is specified by International Association of Precision Agriculture is used [20], which states that “Precision agriculture is a management strategy that gathers, processes and analyzes temporal, spatial and individual data and combines it with other information to support management decisions according to estimated variability for improved resource use efficiency, productivity, quality, profitability and sustainability of agricultural production”. This definition of PA could be considered as the recognized one not only by the industrial, but also by the academic community by been cited at the main page of “Precision Agriculture” journal—the leading international journal in PA field [21].

Depending on the agricultural sector, the PA applications could roughly be divided into two categories: crop Precision Agriculture Technologies (crop PATs) and livestock PATs, where crop PATs could be defined as the technologies for managing spatial and temporal variability for improving production of crops production and environmental impact, while livestock PATs are defined as technologies for optimization of value contribution of each animal [22].

This research is dedicated to the crop PATs adoption, which is considered to be the less developed compare to the livestock PAT [23], due to some reluctance of farmers for crop PATs acceptance [19], and thus it provides the interesting case for sheading the light on the research gap on why some PATs are adopted quickly and widely, while others are lagging behind [11].

The very beginning of PA history dates back to 1929 [24] and commercially viable development of PA technologies started in the 1990th [7,24]. However, despite the several decades of PA development, the pace of PA adoption is still very slow [7,8,19,25], giving the rise of different streams of PA adoption factors literature, specified by region/country, particular PA technology, type of agricultural production, interest groups involved and so on.

Major part of PA adoption research is focused on developed world—mainly USA, Australia and some European countries [12,26,27], as the PA technologies started to be introduced and were commercially available in these countries and therefore its introduction is already well documented [12]. Only small number of studies are focused on transition and developing economies [11], although several studies have already proven that country and regional specificity matters, due to the agro-climatic, socioeconomic, technological and locational variability [7,8,24,28,29,30], urging to focus more research attention on these countries [26].

Therefore, the focus of this paper on the research on PA adoption in transition economies in a context of the Polish economy would allow for (i) comprehensive cross-regional and cross-countries comparisons, leading to more reliable information on which PA adoption factors are common for all countries and which are country-specific [26], and would also (ii) “complement the ubiquity” of developed countries studies providing more global perspective on PA adoption factors literature [26] p. 4.

3. Methodology

In order to identify and summarize the key factors influencing the adoption of PA technologies the Systematic Literature Review (SLR) [31] was conducted through the Google Scholar and Science Direct databases using the three key word sequences: (1) precision agriculture/precision farming, (2) adoption/use, and (3) factors/drivers. The initial search yielded 45 papers matched the formal criteria of the search, out of which only 16 were found to be directly addressing the research aims. This sample was reviewed applying the snowball approach, enabling the search of other relevant papers, and the final list of 21 papers was identified.

This sample list includes the academic papers from peer-reviewed journals indexed in Scopus or Web of Science databases. It covers ex-ante and ex-post studies, based on data collected through interviews, surveys, literature reviews, meta-analyses employing both qualitative and quantitative methodologies, which makes this final papers’ sample sufficiently credible and comprehensive for the identification and summarization of the factors influencing the PA adoption.

The final list of identified key factors comprises 9 factors, organized into 5 groups based on the classificational frameworks used by Antolini et al. (2015) [32], Maloku (2020) [33], Pierpaoli et al. (2013) [34], and Say et al. (2018) [35].

This list of identified key factors formed the conceptual framework for the compilation of the ranking of Polish voivodships’ according to their potential of PA adoption. The ranking is based on the Sturgess rule, which is quite commonly used to determine the number of groups in the sample and their limits [36,37,38,39,40]. For the purpose of this research, Sturgess rule was applied to determine the number “PA adoption clusters” for 16 Polish voivodships and the limits of each cluster. The data which were used for the calculation of this ranking were the latest available official data (2019–2020 period) from the Polish State Statistical Office (PSSO), with the exception of “Credit Availability” variable’s data taken from the consulting firm database, claimed to be based on PSSO data and “Information” variable’s data compiled from the official web-pages of Polish Ministry of Education and Polish Academy of Science.

4. Results and Discussion

The analysis of the papers selected through SLR yielded a total of 9 factors explaining the adoption of PA technologies (PATs), which were organized into 5 groups adapted from the Antolini et al. (2015) [32], Maloku (2020) [33], Pierpaoli et al. (2013) [34], and Say et al. (2018) [35] frameworks (

Table 1. PA adoption factors.

Factor	Explanation	References
Socio-economic factors
Age	Being a young age farmer increases the probability of PATs adoption	Castle et al., 2016 [13]; Daberkow and McBride, 2003 [7]; Kutter et al., 2011 [41]; Pierpaoli et al., 2013 [34]; Reichardt et al., 2008 [25]; Tamirat et al., 2018 [42]; Tey and Brindal, 2012 [12].
Education	Higher education levels positively impact the PATs adoption	Daberkow and McBride, 2003 [7]; Kutter et al., 2009 [41]; Larson et al., 2007 [29]; Michels et al., 2020 [30]; Paustian et al., 2017 [8]; Reichardt et al., 2008 [25]; Tey et al., 2012 [12]; Walton et al., 2010 [43]; Watcharaanantapong et al., 2014 [24].
Agro-technological factors
Farm size	Large farms are more likely to adopt the PA technologies	Barnes et al., 2019 [26]; Castle et al., 2016 [13]; Daberkow and McBride, 2003 [7]; Jensen et al., 2012 [44]; Kutter et al. 2011 [41]; Kernecker et al., 2020 [5]; Larson et al., 2007 [29]; Paustian et al., 2017 [8]; Watcharaanantapong et al., 2014 [24]; Pierpaoli et al., 2013 [34]; Reichardt et al., 2008 [25]; Tamirat et al., 2017 [42]; Tey et al., 2012 [12]; Walton et al., 2008 [43].
Financial factors
Farm income	Higher farm income positively impacts the PATs adoption	Barnes et al., 2019 [26]; Castle et al., 2016 [13]; Kernecker et al., 2020 [5]; Tey et al., 2012 [12]; Watcharaanantapong et al., 2014 [24].
Credit availability	Possibility of getting investment loan from bank increases PATs adoption	Aubert et al., 2012 [45]; Barnes et al., 2019 [26]; Daberkow et al., 2003 [7]; Kernecker et al., 2020; [5]; Kutter et al., 2009 [41]; Reichardt et al., 2008 [25].
Technological factors
Smartphone	Smartphone usage positively impacts PA adoption	Castle et al., 2016 [13]; Kernecker et al., 2020 [5]; Michels et al., 2020 [30]; Reichardt et al., 2008 [25].
Computer	Computer literacy positively impacts PA adoption	Daberkow and McBride, 2003 [7]; Larson et al., 2007 [29]; Pierpaoli et al., 2013 [34]; Tey and Brindal, 2012 [12]; Reichardt et al., 2008 [25]; Walton et al., 2010 [43]; Watcharaanantapong et al., 2014 [24].
Internet	Internet connectivity is the must for PATs adoption	Kernecker et al., 2020 [5]; Michels et al., 2020 [30]; Reichardt et al., 2008 [25].
Informational factors
Information about PA	Better informed about PATs farmers are more likely to adopt the PATs	Barnes et al., 2019 [26]; Busse et al., 2014 [46]; Kutter, 2009 [41]; Larson et al., 2007 [29].

).

Socio-economic group of factors refers to the human capital quality [7] or personal background of the farmer [12] and most commonly includes age and education factors.

Age is considered as an important factor for PA adoption. The younger farmers found to be more likely and more faster to adopt the PATs [47] due to the higher interest in new technologies [42], longer planning horizon, better exposure to new technologies [29] and lesser risk-aversion [48], contrary to the older farmers, which due to the factors inherited in the aging process [7] are more resistant to changes [42,49], are less exposed to PATs [49], have a shorter planning horizon [49] and thus may not envisage the long-term economic benefits of PATs [42].

Successful PATs’ adoption is highly dependent on a set of technological, informational, and analytical skills [29] and thus the role of education—especially formal university education can hardly be overestimated [12,25]. Educational factor plays an important role in PATs adoption as it increases the awareness of PATs and equips the farmers with the necessary skills and knowledge [8,25], e.g., educated farmers are better acknowledged with new technologies and more able to use and assess different informational sources [25], which allows them to understand better the potential benefits of PATs, adopt PATs and use them more efficiently later on [8,24].

Farm size as the agro-technological factor plays an important role in PATs adoption due to the economy of scale and efficiency considerations [13,45]. Farmers with a larger farms found to be more likely to adopt the PATs as the larger farms could spread the PATs costs for a higher number of hectares of land [13] and potentially generate more income from the larger-scale farm [12], thus providing the better capacity to absorb the high-costs and mitigate the risks of PATs adoption [12] and allowing for exploitation of economy of scale factor.

The efficiency consideration imply that large-scale farming requires more elaborated, sophisticated and technically advanced means to manage the large size farm, which potentially turns farm owners to PATs adoption [13]. Some attempts were even made to determine the optimal land size for achieving profitability depending on PATs crops specialization [41]. Considering the predominantly small-scale character of the Polish farming, for the purpose of this research, the farms with more than 50 ha acreage were considered as a large farms.

The last but not least reason for a positive correlation of large farm size and PATs adoption is a suggestion that PATs technologies (as well as many other tech-products) are more aggressively marketed to the large scale farmers due to the potentially higher profits and smaller operational costs for the PATs dealers [7].

Farm size is tightly bound with the farm income, which belongs to the financial group of factors. It proved to be important as the farmer with a higher income has the better financial capacity to adopt the PATs, because he/she is less concerned about the PATs high-costs and he/she can bear the long payback periods for the PATs introduction [12,13,26]. Also, he/she has the potential to sustain the risk in investment in PATs [26] if they turn out to be unprofitable.

Another important financial factor is the credit availability factor, which is considered to be the one of the most significant limitation for the PATs adoption [33], due to PATs’ high investment costs and as the public (state) funds for PA investments are not always available [5,26], the possibility of borrowing external capital (taking bank credit) found to play an important role for PATs adoption [7,26,45].

Technological factors, such as smartphone, computer, and Internet availability and usage are indicators of the overall level of farmers’ and farm technological development (as the farmers who are opened for technological advanced are more willing and prepared for PATs adoption [13,25]) and are the essential determinants of the PATs adoption and its proper functioning [12,30].

The importance of smartphone usage for PATs adoption and functioning is determined by the fact that smartphones, with the pre-installed applications and sensors could replicate the PATs to some extent and could be used as a complements (integral parts) of the PATs: e.g., smartphones can be used for the PA data collection, intermediate data storage, processing, and transferring it to the computer for a decision-making and storage [29,30].

The computer availability and usage is another factor determining PATs adoption: Pierpaoli (2013) [34] p. 64 found the “computer confidence” to be the second important factor of technology adoption, as it is the essential and integral part of PATs functioning—without it PATs would not be functional [29]. In terms of PATs pre-introduction assessment—the availability and use of computer as a part of general farm management (e.g., for e-mail correspondence, information search, records keeping) is a proxy for farmer’s technological knowledge and skills necessary for the smooth PATs introduction [12,13,29,42].

The Internet connection and its reliability is almost by default PATs adoption factor, because in the most cases Internet (as a part of wireless connection) connects all structural PATs devices in a functional PATs network and without it, it would not work [5,30].

It is almost impossible to underestimate the availability of reliable and non-commercially biased information for a farmer while he/she making a decision on adopting or non-adopting the specific PA technology. The significant part of adoption literature considers the informational awareness of PATs as the important component of its adoption either as a part of a generalist adoption research stream [5,25,26,29,46], as well as more focused on PA information availability specific studies [41]. The range of potential sources of PA information is substantial. It covers PA tech firms, Farmers’ Unions, Governmental Agencies, Educational centers, Universities, Research Institutions and Extension Services and Centers [26,29,41,46]. For the purpose of this research, the Agricultural Universities and Agricultural R&D institutes were taken into consideration as the potential sources of PA relevant information, because they can provide the considerably independent and diversified information about PATs through their educational programs, courses, seminars, extension services and other channels of PA information dissemination.

The list of above identified PA adoption factors, selected through conducted SLR is consistent with the previous research findings [5,19,34,43], where all these factors found to have an important impact on PA adoption. It also contributes to PA adoption literature by going a step further and making an attempt to explain, why these factors are important for PA adoption. The selection of these key PA adoption factors serves as a conceptual framework for the compilation of the ranking of PA adoption potential of 16 Polish voivodships.

The further compiled ranking of PA adoption potential of 16 Polish voivodships is based on the Sturgess rule, which was used at the first stage of ranking process to determine the number “PA adoption clusters” for each PA factor and their “limits”. At the next stage, the calculation of each voivodship’s PA adoption score for each factor was conducted and lastly the final ranking list based on cumulative voivodship’s score for all PA factors was compiled.

First of all, the set of variables was assigned to quantify the already identified PA adoption factors (

Table 2. PA adoption factors and their variables.

Factor	Variables
Socio-economic factors
Age	Percentage share of people aged 25–44 in rural areas (2020)
Education	Percentage share of people with tertiary education (2020)
	Percentage share of people under lifelong learning programs (2020)
Agro-technological factors
Farm size	Percentage share of farms with more than 50 ha (2020)
Financial factors
Farm income	Gross value added per employed person in agriculture, forestry and fishing in current prices, in PLN (2019)
	Agricultural output (market value) in current prices, in PLN (2019)
Credit availability	Number of bank outlets per 1 million inhabitants (2020)
Technological factors
Smartphone	Percentage of households with smartphones (in total households) (2020)
Computer	Percentage of households with computers (in total households) (2020)
Internet	Percentage of households with access to the Internet at home (in total households)(2021)
Informational factor
Information about PA	Number of Agricultural Universities and Research and Development Institutes (2022)

).

Next, the number of voivogships’ clusters was calculated using the Sturgess rule

k = 1 + log₂(N)

(1)

where, “k” is the number of clusters, and “N” is number of Polish voivodships included in the sample. The calculations show (1 + log₂(16) = 5), that 16 Polish voivodships should be divided into five PA adoption clusters (

Table 3. PA adoption clusters’ description.

Cluster 1	highest adoption potential
Cluster 2	high adoption potential
Cluster 3	medium adoption potential
Cluster 4	low adoption potential
Cluster 5	lowest adoption potential

).

Then, for the each variable, the cluster interval was calculated using the formula, which is of complementary use for Sturgess rule for clusters (groups) intervals [50] p. 63.

A = (x_max − x_min)/k

(2)

where, “a” is cluster interval, x_max and x_min are the maximum and minimum observed values, and “k” is number of clusters. After that, using the obtained “cluster interval” for each variable, all 16 voivodships were grouped in 5 clusters.

For the variable “Age” the cluster interval was found to be equal 0.424 and according to this, all 16 voivodships were classified for 5 groups depending of the their PA adoption potential for the “Age” variable. For example, Warmińsko-Mazurskie, Lubuskie, Pomorskie, Dolnośląskie, and Wielkopolskie voivodships with the highest percentage share of people aged 25–44 were grouped in cluster 1 (highest adoption potential), as the more younger rural population is more tend to adoption of PATs, while the Podlaskie and Łódzkie voivodships formed the cluster 5 (lowest adoption potential), as the share of young people was the lowest in them (

Table 4. (a) Resulting Table for 16 voivodships for variable “Age”. (b) Calculations for the “cluster interval” for variable “Age”.

(a)
	Voivodship	% Share of People Aged 25–44		Cluster Range (Limits)		Cluster
				min	max
-1-	-2-	-3-		-4-	-5-	-6-
10	Podlaskie	29.06		29.06	29.48	Cluster 5—lowest adoption potential
5	Łódzkie	29.38				Cluster 5—lowest adoption potential
7	Mazowieckie	29.68		29.48	29.91	Cluster 4—low adoption potential
3	Lubelskie	29.75				Cluster 4—low adoption potential
12	Śląskie	29.78				Cluster 4—low adoption potential
13	Świętokrzyskie	29.83				Cluster 4—low adoption potential
8	Opolskie	30.02		29.91	30.33	Cluster 3—medium adoption potential
2	Kujawsko-pomorskie	30.48		30.33	30.76	Cluster 2—high adoption potential
6	Małopolskie	30.57				Cluster 2—high adoption potential
9	Podkarpackie	30.73				Cluster 2—high adoption potential
16	Zachodniopomorskie	30.74				Cluster 2—high adoption potential
14	Warmińsko-mazurskie	30.81		30.76	31.18	Cluster 1—highest adoption potential
4	Lubuskie	30.88				Cluster 1—highest adoption potential
11	Pomorskie	30.92				Cluster 1—highest adoption potential
1	Dolnośląskie	31.00				Cluster 1—highest adoption potential
15	Wielkopolskie	31.18				Cluster 1—highest adoption potential
(b)
min “Age” (column 3- Table above)			29.06
max “Age” (column 3- Table above)			31.18
Cluster’s interval for “Age”			0.424			0.424  =  (31.18 −29.06)/5  =  0.424
			min	max
Cluster 5—lowest adoption potential			29.06	29.48		29.48  =  29.06 + 0.424
Cluster 4—low adoption potential			29.48	29.91		29.91  =  29.48 + 0.424
Cluster 3—medium adoption potential			29.91	30.33		30.33  =  29.91 + 0.424
Cluster 2—high adoption potential			30.33	30.76		30.76  =  30.33 + 0.424
Cluster 5—lowest adoption potential			30.76	31.18		31.18  =  30.76 + 0.424

).

Results of the calculations for the “clusters’ intervals” for variables “Education”, “Farm Size”, “Farm Income”, “Credit Availability”, “Smartphone”, “Computer”, “Internet”, and “Information” are presented in

Table 5. Calculations for the “cluster interval” for variable “Education”. (% share of people with tertiary education).

min “Education”	18.9
max “Education”	35.2
Cluster’s interval for “Education” (% share of people with tertiary education)	3.26		(35.2 − 18.9)/5 = 3.26
	min	max
Cluster 5—lowest adoption potential	18.90	22.16	22.16 = 18.90 + 3.26
Cluster 4—low adoption potential	22.16	25.42	25.42 = 22.16 + 3.26
Cluster 3—medium adoption potential	25.42	28.68	28.68 = 25.42 + 3.26
Cluster 2—high adoption potential	28.68	31.94	31.94 = 28.68 + 3.26
Cluster 1—highest adoption potential	31.94	35.20	35.20 = 31.94 + 3.26

,

Table 6. Calculations for the “cluster interval” for variable “Education” (% share of people under lifelong learning programs).

min “Education”	2.0
max “Education”	5.5
Cluster’s interval for “Education” (% share of people under lifelong learning programs)	0.7		(5.5 − 2.0)/5 = 0.7
	min	max
Cluster 5—lowest adoption potential	2.0	2.7	2.7 = 2.0 + 0.7
Cluster 4—low adoption potential	2.7	3.4	3.4 = 2.7 + 0.7
Cluster 3—medium adoption potential	3.4	4.1	4.1 = 3.4 + 0.7
Cluster 2—high adoption potential	4.1	4.8	4.8 = 4.1 + 0.7
Cluster 1—highest adoption potential	4.8	5.5	5.5 = 4.8 + 0.7

,

Table 7. Calculations for the “cluster interval” for variable “Farm Size”.

min “Farm Size”	0.5
max “Farm Size”	25.9
Cluster’s interval for “Farm Size”	5.08		(25.9 − 0.5)/5 = 5.08
	min	max
Cluster 5—lowest adoption potential	0.50	5.58	5.58 = 0.50 + 5.08
Cluster 4—low adoption potential	5.58	10.66	10.66 = 5.58 + 5.08
Cluster 3—medium adoption potential	10.66	15.74	15.74 = 10.66 + 5.08
Cluster 2—high adoption potential	15.74	20.82	20.82 = 15.74 + 5.08
Cluster 1—highest adoption potential	20.82	25.90	25.90 = 20.82 + 5.08

,

Table 8. Calculations for the “cluster interval” for variable “Farm Income” (Gross value added per employed person in agriculture, forestry and fishing).

min “Farm Income”	6313
max “Farm Income”	44,046
Cluster’s interval for “Farm Income”	7547		(44,046 − 6313)/5 = 7547
	min	max
Cluster 5—lowest adoption potential	6313	13,860	13,860 = 6313 + 7547
Cluster 4—low adoption potential	13,860	21,406	21,406 = 13860 + 7547
Cluster 3—medium adoption potential	21,406	28,953	28,953 = 21,406 + 7547
Cluster 2—high adoption potential	28,953	36,499	36,499 = 28,953 + 7547
Cluster 1—highest adoption potential	36,499	44,046	44,046 = 36,499 + 7547

,

Table 9. Calculations for the “cluster interval” for variable “Farm Income” (Agricultural output—Market Value).

min “Farm Income”	2710
max “Farm Income”	10,056
Cluster’s interval for “Farm Income”	1469		(10,056 − 2710)/5 = 1469
	min	max
Cluster 5—lowest adoption potential	2710	4179	4179 = 2710 + 1469
Cluster 4—low adoption potential	4179	5648	5648 = 4179 + 1469
Cluster 3—medium adoption potential	5648	7118	7118 = 5648 + 1469
Cluster 2—high adoption potential	7118	8587	8587 = 7118 + 1469
Cluster 1—highest adoption potential	8587	10,056	10,056 = 8587 + 1469

,

Table 10. Calculations for the “cluster interval” for variable “Credit Availability“.

min “Credit Availability”	285
max “Credit Availability”	382
Cluster’s interval for “Credit Availability”	19		(382 − 285)5 = 19
	min	max
Cluster 5—lowest adoption potential	285	304	304 = 285 + 19
Cluster 4—low adoption potential	304	324	324 = 304 + 19
Cluster 3—medium adoption potential	324	343	343 = 324 + 19
Cluster 2—high adoption potential	343	363	363 = 343 + 19
Cluster 1—highest adoption potential	363	382	382 = 363 + 19

,

Table 11. Calculations for the “cluster interval” for variable “Smartphone“.

min “Smartphone“	71.3
max “Smartphone“	86.3
Cluster’s interval for “Smartphone“	3		(86.3 − 71.3)/5 = 3
	min	max
Cluster 5—lowest adoption potential	71.3	74.3	74.3 = 71.3 + 3
Cluster 4—low adoption potential	74.3	77.3	77.3 = 74.3 + 3
Cluster 3—medium adoption potential	77.3	80.3	80.3 = 77.3 + 3
Cluster 2—high adoption potential	80.3	83.3	83.3 = 80.3 + 3
Cluster 1—highest adoption potential	83.3	86.3	86.3 = 83.3 + 3

,

Table 12. Calculations for the “cluster interval” for variable “Computer“.

min “Computer“	66.5
max “Computer“	82.7
Cluster’s interval for “Computer“	3.2		(82.7 − 66.5)/5 = 3.2
	min	max
Cluster 5—lowest adoption potential	66.5	69.7	69.7 = 66.5 + 3.2
Cluster 4—low adoption potential	69.7	73.0	73.0 = 69.7 + 3.2
Cluster 3—medium adoption potential	73.0	76.2	76.2 = 73.0 + 3.2
Cluster 2—high adoption potential	76.2	79.5	79.5 = 76.2 + 3.2
Cluster 1—highest adoption potential	79.5	82.7	82.7 = 79.5 + 3.2

,

Table 13. Calculations for the “cluster interval” for variable “Internet“.

min “Internet“	86.5
max “Internet“	95.3
Cluster’s interval for “Internet“	1.8		(95.3 − 86.5)/5 = 1.8
	min	max
Cluster 5—lowest adoption potential	86.5	88.3	88.3 = 86.5 + 1.8
Cluster 4—low adoption potential	88.3	90.0	90.0 = 88.3 + 1.8
Cluster 3—medium adoption potential	90.0	91.8	91.8 = 90.0 + 1.8
Cluster 2—high adoption potential	91.8	93.5	93.5 = 91.8 + 1.8
Cluster 1—highest adoption potential	93.5	95.3	95.3 = 93.5 + 1.8

and

Table 14. Calculations for the “cluster interval” for variable “Information“.

min “Information“	0.0
max “Information“	6.0
Cluster’s interval for “Information“	1.2
	min	max	(6.0 − 0)/5 = 1.2
Cluster 5—lowest adoption potential	0.0	1.2	1.2 = 0.0 + 1.2
Cluster 4—low adoption potential	1.2	2.4	2.4 = 1.2 + 1.2
Cluster 3—medium adoption potential	2.4	3.6	3.6 = 2.4 + 1.2
Cluster 2—high adoption potential	3.6	4.8	4.8 = 3.6 + 1.2
Cluster 1—highest adoption potential	4.8	6.0	6.0 = 4.8 + 1.2

accordingly.

The final result compiling all 16 voivodships grouped into 5 clusters for each variable is in

Table 15. PA adoption potential of 16 Polish voivodships ranked by 5 PA adoption clusters (1-highest adoption potential; 5-lowest adoption potential).

		Age	Education		Farm Size	Farm Income		Credit Availability	S-Phones	Computers	Internet	Information	Cumulative Voivodship Score
	Voivodship	% share of peopl aged 25-44	% share of people with tertiary education	% share of people under lifelong learning programs	% share of farms with more than 50 ha	Gross value added per employed person in agriculture, forestry and fishing	Agricultural output	Number of bank outlets per 1 mln inhabitants	% of households with smartphones (in total households)	% of households with computers (in total households)	% of households with access to the Internet at home	Number of Agricultural Universities and R&D Institutes
1	Dolnośląskie	1	3	3	4	3	5	3	1	1	1	4	29
2	Kujawsko-pomorskie	2	3	1	4	2	3	4	3	4	2	5	33
3	Lubelskie	4	5	3	5	4	4	1	4	3	3	2	38
4	Lubuskie	1	4	2	1	1	5	2	5	5	3	5	34
5	Łódzkie	5	5	5	5	3	3	4	3	3	4	4	44
6	Małopolskie	2	4	4	5	5	4	5	4	2	2	4	41
7	Mazowieckie	4	3	2	5	1	2	3	3	2	1	1	27
8	Opolskie	3	1	1	4	3	4	2	3	2	4	5	32
9	Podkarpackie	2	5	5	5	5	5	5	1	1	2	2	38
10	Podlaskie	5	4	5	5	3	3	2	3	4	5	5	44
11	Pomorskie	1	4	4	4	1	4	4	1	2	2	5	32
12	Śląskie	4	3	3	5	3	1	5	3	2	1	5	35
13	Świętokrzyskie	4	4	4	5	4	3	5	5	4	5	5	48
14	Warmińsko-mazurskie	1	4	5	4	1	4	2	2	3	3	4	33
15	Wielkopolskie	1	5	5	5	2	1	3	4	2	1	3	32
16	Zachodniopomorskie	2	4	5	3	1	5	3	3	3	2	5	36

.

After that, the Cumulative Voivodships’ Score was calculated for each voivodship as a simple sum of cluster numbers of each voivodship (see the last column “Cumulative Voivodship Score” in Table 15 above). Then, based on these scores, the voivodships’ PA adoption Ranking was compiled (

Table 16. PA adoption Cumulative Voivodships’ Score and Voivodships’ Ranking.

	Cumulative Voivodship Score	Voivodship Rank
Mazowieckie	27	1
Dolnośląskie	29	2
Opolskie	32	3
Pomorskie	32	3
Wielkopolskie	32	3
Kujawsko-Pomorskie	33	4
Warmińsko-Mazurskie	33	4
Lubuskie	34	5
Śląskie	35	6
Zachodniopomorskie	36	7
Lubelskie	38	8
Podkarpackie	38	8
Małopolskie	41	9
Łódzkie	44	10
Podlaskie	44	10
Świętokrzyskie	48	11

).

The brief overview of viovodships’ PA adoption potential (Table 15 and Table 16) shows no specific geographical patterns except for the “core-periphery” divergence, where the well-developed metropolitan areas got the highest ranking (e.g., Mazowieckie voivodship is #1 in the ranking), while the less developed peripheral regions are ranked low. This pattern is quite common for the prevalent majority of countries despite their developmental stage (i.e., despite whether they are developed, transitioning or developing countries).

The compilated PA adoption Rankings (Table 15 and Table 16) could be of potential usage by all stakeholders involved in PA adoption. Industry practitioners could use them for specifying their marketing strategies (e.g., depending whether they would target earlier adopters from viovodships with higher PA adoption potential or late adopters from viovodships with the lower PA adoption potential). It could also be used by them for PA products’ sales strategies specification (depending on “Farm Size” Ranking) and for sales, price, and payments strategies (depending on “Farm Income” and “Credit Availability” Rankings). Academic educators could use the Rankings for targeting their programs, courses, seminars and educational extensions. Public policymakers could use these Rankings while working on some specific aspects of rural policies (rural areas) development, like rural-urban migrations, farm subsidies, rural investments, research and development (R&D) directions, and agriculture relevant education.

5. Conclusions

Despite countless economic, social and environmental benefits, the pace of PA adoption is still slow in both developed and developing countries, therefore raising the need for analysis of the factors impacting PA adoption.

Most of the PA adoption literature is focused on developed countries (where the PA technologies started to be commercially available earlier, then in the rest of the world), paying less attention to the transition and developing countries. This paper fills this gap by providing a research on PA adoption factors and envisaged potential of transition countries to implement PA technologies in their regional socio-economic and agro-technological contexts, using Poland as a case-study of transition economy. The most common PA adoption factors (identified by SLR conducted) formed the conceptual framework for compilation of ranking of 16 Polish voivodships according to their PA adoption potential.

This is the first generalist exploratory research on PA adoption in a country-wide Polish context and perspective research could certainly benefit from the diversification of future studies by crops and livestock varieties, particular PA technologies specializations, earlier and late PA adopters, and inclusion of other transition and developing countries in the study sample allowing for more cross-countries and cross-regional comparative analysis.

These research findings could be of interest for multiple stakeholders from industry, research and public policy communities, as the paper provides useful insights on existing gaps and potential avenues on PA development in Poland and other transition economies from the Central and Eastern European region.