Elsevier

Journal of Rural Studies

Volume 41, October 2015, Pages 109-117
Journal of Rural Studies

Robotic milking-farmer experiences and adoption rate in Jæren, Norway

https://doi.org/10.1016/j.jrurstud.2015.08.004Get rights and content

Highlights

  • We interview dairy farmers about their experiences with robotic milking (AMS).

  • We explore why the adoption rate of AMS is particularly high in one region.

  • Farmers domesticate the AMS technology to their needs.

  • Sociocultural norms and the agricultural knowledge system explain adoption rates.

  • High wage rates and relationship with the farming industry also contributes.

Abstract

Robotic milking or automatic milking systems (AMS) are becoming increasingly popular in Norway as well as in the other Nordic countries. To explore what motivates farmers to invest in AMS and what the consequences for farmers' lifestyle and management are, we (the researchers) visited and interviewed 19 dairy farmers in Southern Norway. Fourteen of the farmers are situated in a region of Norway (Jæren), where the adoption rate of AMS is significantly higher than in the rest of the country. Therefore our main interest was to explain the high adoption rate in Jæren. The findings suggest that to succeed with AMS farmers must be motivated, behave proactively and adapt the new technology to their specific needs. Saved time on milking, more interesting farming, more stable treatment of the cow and less need for relief are some of the advantages. Farmers experience to be constantly on call and information overload as the greatest disadvantages of AMS. The main reasons to invest in AMS are increased flexibility and reduced workload, and AMS has allowed a more modern lifestyle. The high adoption rate of AMS in Jæren can be explained by human and social capital, socio-cultural factors and the well-developed agricultural knowledge system in the area. Close relations with the farm machinery industry in the area, a strong belief in technology, high wage rates and difficulties of getting skilled labor are other factors which can explain the high adoption rate of AMS.

Introduction

Robotic milking machines milk cows automatically at any time, without the need for a human worker to be present. AMS is not only a new milking system, but rather a completely new management system. Manufacturers claim that AMS can potentially raise milk yields, and is beneficial in terms of animal health, welfare and for working conditions of the farmer. In 2008 AMS was in use on about 5500 milk farms worldwide (Svennersten-Sjaunja et al., 2008). More than 90% of all dairy farms using AMS are located in northwestern Europe where investments are driven by high labor costs, a continuous increase in the average herd-size and a dominance of the family farm structure (Mathijs, 2004). Robotic or automatic milking systems (AMS) are also becoming increasingly important in Norwegian dairy farming. Installing an AMS is a huge investment, approximately 1.2 million NOK, still almost all Norwegian farmers who refurbish their cowsheds install AMS. By the end of 2012 there were 1032 milking robots in Norway and 1636 in Sweden (TINE, 2013), and the majority of Norwegian dairy farmers (94%) deliver their milk to Tine cooperative dairy company. On average robots milked the cows on 11.1 percent of all Norwegian dairy farms and handled roughly 25 percent of all milk. Thus Norway is among the countries in the world with the highest frequency of AMS in dairy farming. The figures indicate that robotic farms are significantly larger than the average farm. Except from this we know very little about what Norwegian farmers' experiences with AMS are. Therefore my first research question is: What are Norwegian farmers' experiences with AMS? Exploring farmers' experiences will also contribute in answering my second research question. The statistics from Tine (TINE, 2013) show that the adoption rates of AMS differs substantially throughout Norway. In Jæren, a district in the Southwestern part of Norway, 18 percent of the farmers had an AMS, seven percent above the national average. The figure is even higher than in the Netherlands, where more than 10% of the farmers have applied the AMS technology (Steeneveld et al., 2012). Most research on diffusion of AMS is done at a national level, comparing adopters with non-adopters. Less is known about why the diffusion rate in one geographical region within a country can differ significantly from the country's average diffusion rate. For example we have limited knowledge of what particular sociocultural and human factors in a geographical area that can explain a high adoption rate of AMS. Therefore it is in my interest to study why the adoption rate in Jæren is so high, as compared to the rest of Norway. My second research question is: What can explain the high adoption rate of AMS in Jæren?

The paper is organized as follows: First I provide the readers with a general background to AMS and what farmers' reasons for adopting AMS are. Then I present relevant theory which can explain differences in adoption rates, particularly innovation diffusion theory and theory on social and human capital. Next I describe the empirical material and the methods I used. In the empirical Section 1 use the interviews to answer what Norwegian farmers' experiences with AMS are, and why the adoption rate of AMS in Jæren is particularly high. Finally I discuss the findings and conclude.

The adoption of AMS implies that farmers structure their time and their farms around the demands of the robot, and to be able to access, analyze and respond to the large amounts of data the AMS is capable of generating (Butler et al., 2012). According to Butler et al. (2012) there is a need for more training to be given to farmers in the use of AMS- generated data. AMS also affects how farmers relate to their cows, and on how they identify themselves as stockpersons (Seabrook, 1992). Thus Stuart et al. (2013) found that a production system where cows graze on pasture and choose when to be milked by AMS involved less alienation to dairy cows as compared to more industrial dairy farms. However, Stuart et al. (2013) conclude that work performed in a profit-maximizing animal agriculture system will inevitably cause alienation, exhaustion, and suffering for the animals. Similarly, Porcher and Schmitt (2012) argue that despite three decades of research and work on animal welfare, farm animals have experienced only small improvements. According to Porcher and Schmitt (2012) alienation and how farm animals perform work must be considered in order to truly improve animal welfare. Similarly, in a study of a conversion from family farms to a huge milking parlor Hansen (2014) points out that the mechanical separation of human and cow during the milking process can lead to affectively shared interspecies and inter-human alienation. The technology of the parlor can separate both from a process formerly dependent upon specialized knowledge, affective empathy, and embodied knowledge. Holloway et al., 2014a, Holloway et al., 2014b) also emphasize that introduction of robots has important effects, in terms of removing routine contact between humans and animals, and unsettling the usual ways in which farmers know and understand their cows. Robots also allow the cows to reveal themselves to the farmer in new ways through the use of information technology and behavior monitoring (Holloway et al., 2014a, Holloway et al., 2014b). Further, introduction of AMS unsettles the identities, roles and subjectivities of humans and animals and thus shifting the ethical relations (Holloway et al., 2014a, Holloway et al., 2014b). Robotic milking opens up new possibilities for managing the cows without being present in the milking parlor. Thus the stockmanship changes from looking at individual cows to looking at herd averages, and there is a concern that reliance on the robot may lead to neglect of cows (Holloway et al., 2014a, Holloway et al., 2014b). The technology transforms ways of knowing and spending time with cattle, such as reducing the amount of physical contact between humans and the cows in the milking parlor while potentially increasing the amount of time humans can spend observing their cows (Owen, 2003). In addition to having a good stockman's eye the farmer also has to be computer literate. Thus conversion to milking robot radically changes the work of the stockperson (Butler et al., 2012). This change requires a transformation of the whole management process. Thus a review of AMS studies suggests that differences in management and farm-level variables may be more important to AMS efficiency and milk production than features of the milking system itself (Jacobs and Siegford, 2012). To reap the benefits of AMS farmers need to fully incorporate the AMS into their management routines.

Although AMS provides increased flexibility, some farmers find that the AMS is actually more of a tie than they had envisioned (Butler et al., 2012). As farmers can be contacted by the robot 24–7 in case of problems, they are only a phone call away from having to go and check why the robot called them. To be constantly on call can be a problem. However, this burden can be lessened if the farmers act proactively and adapt the AMS technology to their farm conditions. For example the level when the alarms go off can be adjusted according to their importance. Such adjustment of the alarms is an example of technology domestication, or the practical as well as emotional adaptation to technologies (Lie and Sørensen, 1996).

Earlier studies on AMS have focused mainly on economic and labor issues. Farmers have different motives for investing in AMS. The economic benefits of AMS are mainly savings in labor and increased production per cow (Bijl et al., 2007). The milk yield increases about 10–15% on average (Steeneveld et al., 2012, Jacobs and Siegford, 2012). AMS have the potential to significantly reduce the production costs or indeed to change the capital-labor ratio (Steeneveld et al., 2012). In fact, by replacing conventional milking systems with AMS, the estimated saving is 20–30% (Mathijs, 2004, Bijl et al., 2007, Sauer and Zilberman, 2012) of the labor allocated to the milking activities. However, other authors found little difference in labor use but differences in task and work flexibilities (Steeneveld et al., 2012). In a review of AMS studies Jacobs and Siegford (2012) report a decrease in labor by as much as 18%. Recently, Steeneveld et al. (2012) quantified the capital cost of AMS at 12.71 € per 100 kg of milk instead of 10.10 € per 100 kg of milk for conventional milking systems. However, Hyde et al. (2007) and Heikkilä et al. (2012) stress that noneconomic factors such as lifestyle choices including avoiding labor management are at least as important as economic factors for the decision to adopt an automatic milking system. Other studies emphasize the importance of the farmer's risk perception, effects of peer-group behavior, and a positive impact of previous innovation experiences (Sauer and Zilberman, 2012). Recent research from Holland on the adoption of AMS (Floridi et al., 2014) supports the expansion diffusion theory, which is described in the section below. According to Floridi et al. (2014) the adoption and the timing of such a decision, is strongly affected by policy uncertainty and market conditions. The effect of this uncertainty is to postpone the decision to adopt the new technology until farmers have gathered enough information to reduce the negative effects of the technological lock-in (Floridi et al., 2014).

AMS takes care of the milking, so there is reason to believe that AMS will offer the farmer more flexibility and freedom. Mathijs (2004) as well as Hyde et al. (2007) stress that lifestyle choices such as avoiding labor management are important for the decision to adopt an AMS. On average, scholars report a 10% reduction in total labor demand compared to conventional milking systems with two milkings per day (Schick et al., 2000, DeKonig et al., 2003). However, there is conflicting evidence regarding possible time saving (Jacobs and Siegford, 2012). When studying AMS systems Butler et al. (2012) found that work routines changed, but farm families did not necessarily experience the expected improvement in the quality of life or an ‘easier’ lifestyle. In practice farmers found that their work routines changed rather than lessened.

Diffusion is the process whereby the innovation is spread, or disseminated. Webster (1971) gives a definition that emphasizes the social process by which an innovation spreads through a social system over time. The major point of interest in diffusion theory is how and why (or why not) some agents adopt ideas or phenomena. One of the first studies in agriculture was conducted by Griliches (1957) who explained the diffusion of innovation by means of an imitation process. Earlier works on this issue described innovation diffusion as an S-shape function (Rogers, 1962), where the new technology is firstly introduced by a group of Innovators, then followed by Earlier Adopters, then by the Early and Late Majority, and finally by the Laggards. The process of adopting AMS can be described as an expansion diffusion, i.e. the innovation is adopted by more and more farmers, so that the total number of adopters is growing over time. Expansion diffusion assumes two major forms, contagious and hierarchical. A hierarchical diffusion process is a ‘trickling down’ process from large to smaller units. The hierarchy may be defined differently. In contagious expansion diffusion the spread is smooth and continuous. Contact with earlier adopters and the quality of communication channels are important factors in this form of diffusion processes. Thus, close physical proximity influences the possibility of adoption, but is not a necessary condition for diffusion to occur.

A classical contribution within the expansion group of diffusion processes is Torsten Hägerstrand's ‘Innovation Diffusion as a spatial process’ (1967). In his work Hägerstrand created models to describe how diffusion takes place. The models were based on an elaborate set of assumptions and concluded that four stages mark the diffusion mechanism:

  • 1

    Primary stage. Innovation appears at its primary source (leaders).

  • 2

    Diffusion stage. Rapidly increasing set of adopters.

  • 3

    Condensing stage. The remaining area is penetrated.

  • 4

    Saturation stage. Marking the slowdown and ending of the diffusion process.

I argue that the diffusion of AMS in Jæren is currently in the condensing stage, while the rest of the country is in the primary or diffusion stage. Agents are often seen as risk-averse and uncertainty-avoiding. However, after some threshold level of adoptions is reached, a ‘bandwagon’ effect of acceptance may occur, which leads to an unevenness of the diffusion process over time. Characteristically the diffusion curve follows a logistic function or S- shaped function, see e.g. Bradford and Kent, 1977.

In order for diffusion of innovations to take place between agents they must be connected by some kind of relevant communication links, such as magazines, meetings or conferences. There is some evidence to suggest that a person's adoption of a new idea or practice is strongly influenced by the behavior of their social network (Valente, 1995). Thus the behavior of one's peers seems to have an important of one's behavior, but there is considerable variation in how much. Individuals have various thresholds to adoption such that some people adopt an idea when no or few others have adopted, while others wait until the majority has adopted. Another network factor shown to affect adoption is a person's position in a network. Thus opinion leaders, often measured as central members in the network, both reflect and drive the diffusion process (Valente and Davis, 1999, Valente and Pumpuang, 2007). The information field (IF) characterizes the extent of contacts that a potential adopter has made at a given point of time. Barriers in the IF, Hägerstrand realized, were real impediments to interactions and communications. Distance from the innovator is one physical barrier. Similarly, Von Hippel (1994) points out that some information is costly to acquire, to transfer and to use in a new location, so-called “sticky” information. According to Von Hippel (1994) such information stickiness affects diffusion of innovation. Resistance to change, information fields and barriers are all important factors that influence the diffusion of an innovation. These factors are changing through time.

The expansion diffusion theory is closely related to theory on social capital. Lin defines social capital (2001, p. 19): “investment in social relations with expected returns 138 in the marketplace”. This definition reflects most writings on social capital (Bourdieu, 1983; 139 Burt, 1992, Coleman, 1988, Lin, 1982, Portes, 1998). While there are many different perspectives on the nature of social capital, the fundamental principle is that economic and social transactions are promoted through the quality of the interactions within a community or network. The key role of social capital is that it can promote development-aiding in the accumulation of either economic or human capital, and it can do so without incurring great financial cost. Scholars emphasize that social capital may be instrumental and help actors both in a social and in an economic sense, which often are interwoven and hardly detachable from one other. In relation to diffusion of AMS the following two effects of social capital are of particular interest: 1) Getting information (Granovetter, 1973, Granovetter, 1983); and 2) transfer of knowledge, innovation, and diffusion of technology or practices (Ahuja, 2000, Brown and Duguid, 1991, Wenger, 1998). Thus farmers may learn about AMS from discussing their farming practices in discussion clubs, which have much in common with communities of practice (Wenger, 1998). Wenger (1998) defines communities of practice as “groups of people who share a concern or a passion for something they do and learn how to do it better as they interact regularly”. Communities of practice reflect the participants’ understanding of what is important generating a common set of thoughts, ideas, and resources to maintain and pass on the accumulated knowledge.

To take advantage of new ideas or practices through networks farmers need human capital. According to Becker (1964) human capital refers to the knowledge, information, ideas and skills of individuals. From Becker's point of view human capital is the most important of all forms of capital in modern economics, and more important than machinery, factories and financial capital (Becker, 1964). His concepts “human capital economy” and the “age of human capital” reflect the importance he attributed to human knowledge, skills and talent. Not only does Becker emphasize the importance of formal education, he also recognizes the fact that much unmeasured learning takes place in firms, and that people need to invest in learning during their lives. Human capital contributes to decision making and problem solving particularly in framing of decisions (Weick, 1995). Framing is an essential part of the problem solving process, and refers to how the problem solver defines and interprets the situation, what alternatives materialize, and how problem solving is implemented (Weick, 1995). Different levels of human and social capital may interact to produce joint effects on productivity. When social capital interacts with human capital, there is a transfer of know-ledge which creates higher productivity. Some people may have less human capital but more social capital, or vice versa. I have discussed human and social capital separately. However, there are obvious links between them. Studies have revealed a link between an individuals' education level and the level of social capital (Huang et al., 2009). The ability to learn depends on the amount of existing related knowledge in the field. New knowledge has to connect to existing knowledge so that people can interpret and put this knowledge into an existing frame of reference (Weick, 1979, Weick, 1995). Thus new agricultural knowledge is selected, adapted and turned into practice through a well-developed agricultural knowledge system, Roling and Engel, 1991, Engel, 1995. This system may be described as stable networks which support agricultural innovation and learning, comprising agricultural schools, advisory services, researchers and dense networks of progressive farmers.

There is an increasing tendency among firms to rely on external sources for innovations. An example is relational governance mechanisms which are based largely on trust and social identification, like teams, shared decision making, and joint development of solutions (Greve and Salaff, 2001, Uzzi, 1996). In adopting innovations, an individual may need much human capital to take advantage of social capital. Therefore I expect an interaction effect between human and social capital, so that much (low) human capital together with much (less) social capital increases (decreases) adoption of AMS. Thus a well developed agricultural knowledge system is important in the adoption of AMS.

Section snippets

Farming and farming culture in Jæren

In this Section I present the geographical area of my research, Jæren, and the farming culture. I think this can contribute to explain why the adoption rate of AMS is particularly high in the area.

Jæren is a narrow strip of productive farmland, about 40 km long, located mostly south of the city of Stavanger, between the shores of the North Sea in the west and the mountains to the interior in the east. Jæren consists of 8 municipalities. Distances to markets are small and the communications are

The empirical study

In this Section 1 will answer the two research questions by exploring the interview data.

Discussion

Since the sample of farmers is not random, this study does not claim to give a representative picture of all Norwegian dairy farmers with AMS. For example no farmers were interviewed who had reverted back to a conventional system; their response might have been different. Increased flexibility is the main advantage of AMS, a finding which indicates that farmers appreciate to have a lifestyle more similar to that of other workers. Thus investing in AMS implies a more modern lifestyle, clearly a

Conclusion

To succeed with AMS farmers must spend some of the time they save on monitoring the cows and the robot. AMS also requires a high motivation in dairy farming and proactive behavior. Further, farmers need a minimum of interest in technology to succeed with AMS, as well as proactive behavior to adapt the technology to their specific needs. The greatest disadvantage of AMS is to be constantly on call. Further, although more data about the herd has a potential to stimulate farmers' interest in dairy

Acknowledgments

The author acknowledges the Norwegian Research Council for funding the study, and Sven Martin Håland and Jon Kristian Sommerseth in Tine for help with the interviews. Tove Rita Øvstebø is also acknowledged for transcribing the interviews.

References (66)

  • P. Bourdieu

    Economic capital, cultural capital, social capital

    Soz. Welt Sonderh.

    (1983)
  • P. Bourdieu

    The Forms of Capital

  • M.G. Bradford et al.

    Human Geography. Theories and Their Applications

    (1977)
  • J.S. Brown et al.

    Organizational learning and communities-of-practice: toward a unified view of working, learning, and innovating

    Organ. Sci.

    (1991)
  • R.S. Burt

    Structural holes: the social structure of competition

    (1992)
  • D. Butler et al.

    The impact of technological change in dairy farming: robotic milking systems and the changing role of the stockperson

    J. R. Agric. Soc. Engl.

    (2012)
  • J.S. Coleman

    Social capital in the creation of human capital

    Am. J. Sociol.

    (1988)
  • K. DeKonig et al.

    Robotic milking and milk quality: effects on bacterial counts, somatic cell counts, freezing point and free fatty acids

    Ital. J. Anim. Sci.

    (2003)
  • P. Engel

    Facilitating innovation: an action-oriented approach and Participatory methodology to improve innovative social practice in agriculture

    (1995)
  • M. Floridi et al.

    Modelling the adoption of automatic milking systems in Noord–Holland

    Bio Based Appl. Econ.

    (2014)
  • A.J. Frøyen

    The enterprising farmers in Jæren

  • M.S. Granovetter

    The strength of weak ties

    Am. J. Sociol.

    (1973)
  • M.S. Granovetter

    The strength of weak ties: a network theory revisited

    Sociol. Theory

    (1983)
  • A. Greve et al.

    Exploring the contributions of human and social capital to performance

    Int. Rev. Sociol

    (2010)
  • A. Greve et al.

    The development of corporate social capital in complex innovation processes

  • Z. Griliches

    Hybrid corn: an exploration in the economics of technological change

    Econometrica

    (1957)
  • T. Hägerstrand

    Innovation Diffusion as a Spatial Process

    (1967)
  • B.G. Hansen

    Problem Solving in Dairy Farming

    (2013)
  • A.M. Heikkilä et al.

    Effects of Economic Factors on Adoption of Robotics and Consequences of Automation for Productivity Growth of Dairy Farms

    (2012)
  • L. Holloway et al.

    Robotic milking technologies and renegotiating situated ethical relationships on UK dairy farms

    Agric. Hum. Values

    (2014)
  • S. Høyland

    Stone and Steel. Development of the Farm Machinery Industry in Jæren, and the Use of Machinery in Agriculture, from 1950–1970

    (1998)
  • J. Hyde et al.

    Robots don't get sick or get paid overtime, but are they a profitable option for milking cows?

    Rev. Agric. Econ.

    (2007)
  • IFCN

    International Farm Comparison Network

    (2013)
  • Cited by (0)

    View full text