info
close
RUSSIAN FEDERATION Robots feeding cows.
©ANDREY-SHA74/Shutterstock.com

The State of Food and Agriculture 2022

Chapter 2 UNDERSTANDING THE PAST AND LOOKING TOWARDS THE FUTURE OF AGRICULTURAL AUTOMATION

KEY MESSAGES
  • Motorized mechanization is an important form of automation in agricultural production and a fundamental component of agricultural transformation worldwide, although its adoption has been uneven and particularly limited in sub-Saharan Africa.
  • Improving access to sustainable mechanization options for small-scale agricultural producers – including women, youth and other marginalized groups – requires technological and institutional innovations, such as mechanization service markets facilitated by digital platforms.
  • The increasing use and variety of digital technologies has the potential to transform agriculture even in low- and middle-income countries, particularly as these technologies become more accessible.
  • Drivers of adoption vary by technology and context. For example, adoption of milking robots is mostly driven by increased flexibility of work schedules and better quality of life; for crop automation technologies, adoption is mostly driven by higher profitability; while for forestry, safer working conditions play an important role.
  • An array of technological solutions are already available for countries at different development stages – and more are in the pipeline. Through appropriate policies and legislation, governments can promote solutions that are suitable for the specific context and needs of different producers.
  • In particular, small-scale agricultural producers need access to affordable and appropriate digital automation technologies to allow them to adopt these technologies and reap their benefits.

In the past, spanning several centuries, human muscle and animals were the main source of power in agriculture. Until recently, automation in agriculture was largely about replacing draught animals and human labour with motorized equipment in a multitude of agricultural operations including land preparation, weeding, harvesting, irrigation, animal milking and feeding, and on-farm handling operations, such as threshing and milling.

Recently, digital automation technologies (see Figure 2) have found their way into agriculture through various applications – sometimes embodied in existing agricultural machinery, sometimes separately. In both cases, these technologies have the potential to improve the diagnosis and decision-making of agricultural producers. When embodied in agricultural machines, agricultural operations can be performed with greater precision, leading to further improvements in efficiency and productivity.

Therefore, these technologies have the capacity to transform rural livelihoods and the associated agricultural landscape, including crop and livestock production, aquaculture, and forestry. In crop production, they can enhance the productivity of inputs such as seeds, fertilizers and water. In livestock and aquaculture production, they can reduce the drudgery and increase the timeliness of operations, and enhance the efficiency of inputs such as feed. In all sectors of agriculture, especially in forestry, machinery can improve working conditions and provide a safer environment for workers.

This chapter reviews the trends in automation technologies across the world, analysing how they differ across countries and regions and what has driven these differences. Due to scarcity of data, the narrative relies heavily on case studies from the literature and on two background papers prepared for this report.1, 2 (See Annex 1 for a comprehensive description of the 27 commissioned case studies.) A historical perspective is followed, from the introduction of motorized mechanization and its dissemination among high-income countries to its subsequent transfer to some low- and middle-income countries. The chapter discusses the drivers of and barriers to adoption and how these explain the divergence in uptake across regions. It also sheds light on some of the trade-offs generated by automation, including the possible negative environmental impacts of motorized machinery. It analyses how digital technologies are transforming the use of agricultural machinery and examines the potential of digital solutions for non-mechanized agriculture. Finally, the chapter describes the state of digital automation technologies across the world and their potential to supersede traditional motorized mechanization and reverse some of its negative impacts.

Trends and drivers of motorized mechanization around the world

Adoption rates vary significantly across regions

Motorized mechanization has increased substantially worldwide. Evidence shows that wide-scale adoption started in the United States of America, where tractors rose to become the main source of farm power, replacing about 24 million draught animals between 1910 and 1960.3 With the exception of the United Kingdom of Great Britain and Northern Ireland, where tractors were first adopted in the 1930s, the transformation of agriculture in Japan and some European countries (Denmark, France, Germany, Spain and former Yugoslavia) was delayed until about 1955, after which motorized mechanization happened very quickly, totally replacing animal traction.4 The use of tractors as farm power became one of the most influential modernizations of the twentieth century, as it allowed, and even triggered, innovations in other agricultural machinery and equipment, such as threshers, harvesters and a wide range of associated implements.5 This significantly eased the drudgery associated with agriculture and allowed farmers to perform tasks in a more timely manner. At a later stage, many Asian and Latin American countries witnessed considerable progress in the adoption of motorized machinery.6 Sub-Saharan Africa, on the other hand, is the only region where progress towards motorized mechanization has stalled over the past decades,7 despite more rapid adoption in some African countries.

When analysing trends in agricultural machinery adoption, paucity of data is a well-recognized constraint. The great diversity of machinery and associated equipment used in agricultural mechanization is an important challenge in terms of data collection (see Box 1 for how FAO is planning to overcome this challenge). The machinery can be generally classified into two groups: (i) engine-based machinery, such as tractors, water pumps and harvesters; and (ii) accessory machinery without an engine, but which combines with an engine-based machine (e.g. tractor implements such as ploughs and seeders, and irrigation schemes). Data are generally collected for engine-based machinery, although even for this category they are scarce due to the high variation in agroecological and agrarian conditions across countries. Different agroclimatic zones, soil conditions, topography and production orientation require the use of different types of machinery and equipment. For example, tractors can have different sizes and attributes (e.g. four vs two wheels). Also different livestock and aquaculture production systems may require very different types of machinery, for example, from feeding systems to milking machines in the case of livestock production.

Box 1Overcoming data challenges in reporting use of agricultural machinery

Until 2009, FAOSTAT reported regularly on the use and trade (volumes and values) in agricultural machinery and equipment. Statistical series, starting in 1961, were published on a relatively small number of items, including total agricultural tractors, harvesters and threshers, milking machines, soil machines and agricultural machinery.

The main source of the dataset was an annual questionnaire sent to national counterparts, covering both use and trade. Some data collected through questionnaires were sourced from national agricultural censuses – normally undertaken every ten years – and updated where possible with yearbooks and other ministerial sources and data portals in the period between censuses. Most countries reported trade data, without specifying the units of machinery in use; this raised concerns about the data and the need to improve both the quality and the detail level of the dataset.

In the early 2010s, FAO revised the questionnaire to include a request to countries for more detailed information, especially in terms of type of machinery. This was complemented with traded quantities and values obtained from the UN Comtrade database; any remaining data gaps were filled using a range of secondary sources, including country case studies.

However, the revised questionnaire did not yield the expected response rate. Only a few countries were able to provide additional details, and the reliability of the overall external information proved limited. As a consequence, administration of the revised questionnaire ceased and reported data are currently only available to 2009 (collected in 2011). The result is that very little is known about the evolution of the adoption of agricultural machinery and equipment in the last ten years. This is a major gap in our understanding of how agricultural systems are evolving.

The Statistics Division of FAO has begun the process of updating the database on machinery by combining different data sources. The methodology is still under development and, compared with the past, is more reliant on survey data, together with agricultural censuses. In the coming years, survey data are likely to be collected in the framework of a range of projects in which FAO is involved, including the Agricultural Integrated Survey Programme (AGRISurvey) and the 50x2030 Initiative to Close the Agricultural Data Gap. These projects are geared to providing technical assistance and promoting data collection in agriculture on a range of topics, touching on socioeconomic and environmental variables, following a parsimonious modular approach that covers the inter-census periods. One module among those proposed is data on machinery availability and use.

Moreover, microdata from agricultural censuses are increasingly published in a more systematic manner. For the inter-census periods, data on machinery use and stocks are available from a number of surveys, such as the household survey promoted by the World Bank – the Living Standards Measurement Study (LSMS) – and similar national surveys. A range of harmonized indicators and microdata from such surveys are gathered in the FAO Rural Livelihoods Information System (RuLIS) database, providing another source of data on machinery use.

The updated dataset will include the quantity of machinery and equipment in use and produced, and the volume of imported and exported machinery (and relative trade values).

FAO plans to evaluate all possible reliable sources by collecting, processing and developing a standardized dataset by 2023. In the longer term, the machinery data domain will be updated with data collected from the revised questionnaire for distribution to countries.

Building on the most recent available data, and acknowledging that these are patchy and outdated, Figure 4 illustrates the progress of mechanization across world regions between 1961 and 2009. It should be noted that the indicator (number of tractors in use per 1 000 ha of arable land) takes into account neither tractor size nor other types of equipment. However, the use of this indicator as a proxy for overall mechanization can be justified, in part by the unavailability of other data, and also by the fact that tractors are currently the main power source for numerous agricultural operations such as land preparation, seeding, fertilizing and chemical spraying. In addition to transportation, tractors can also provide power for pumping water for irrigation as well as for milking machines.

FIGURE 4 Tractors in use per 1 000 HECTARES of arable land

SOURCE: FAO, 2021.
NOTES: Tractors refer to total wheel, crawler and track-laying type tractors used in agriculture. A fourth type of tractor (pedestrian tractor) was considered for a subset of countries as of 2000. Only countries that provided data consistently between 1961 and 2009 were considered (total of 108 countries). Central Asia was omitted due to missing data. See Annex 2 for the complete set of countries, including the 33 countries for which the fourth type (pedestrian tractor) was considered as of 2000.
SOURCE: FAO, 2021.9

The available statistics on the number of tractors per 1 000 ha of arable land (see Figure 4) highlight the unequal regional progress towards mechanization. While high-income countries (Northern America, Europe and Oceania) were already highly mechanized in the 1960s, regions dominated by low- and middle-income countries were less mechanized. Europe witnessed a decline in tractor use between the 1990s and 2000s, with the Russian Federation experiencing the greatest decrease (over 50 percent), probably due to the political and economic transition in the country during that period. However, other countries – for example, Albania, Denmark, Germany, Ireland and the Netherlands – also underwent a significant decrease, although the underlying reasons are not clear. Possibly, as tractors evolve and farms and farmland become more concentrated, the number of hectares (ha) serviced by a single machine rises.

Asia and Northern Africa witnessed rapid mechanization after the 1960s. For example, in Eastern and South-eastern Asia and Southern Asia, the number of tractors per 1 000 ha increased by 56 and 36 times, respectively – from a combined total of 2.7 million tractors in the 1960s to 20.3 million units in the 2000s. However, part of the exponential increase observed in Eastern and South-eastern Asia in the 2000s can be explained by the addition of a fourth type of tractor (pedestrian tractor) to the measurement analysis; for countries like China, Myanmar and the Philippines, this addition increased significantly the total number of tractors. In Northern Africa and Western Asia in the same period, the increase was tenfold (from 3 to 33 units per 1 000 ha). Latin America and the Caribbean also experienced significant growth, with the number of tractors per 1 000 ha of arable land almost tripling, from 5 in the 1960s to 14 in the 2000s. Sub-Saharan Africa was the only region that did not witness noticeable progress in agricultural mechanization. In this region, the number of tractors in use increased very slowly, reaching only 2.1 million in the 1980s (or 2.8 tractors per 1 000 ha of arable land), before declining to 700 000 (or 1.3 per 1 000 ha) in the 2000s. The low level of mechanization in the region is confirmed by a recent study that examined agricultural mechanization in 11 countries and found that light hand-held tools are the main type of equipment used. The study shows that only 18 percent of sampled households have access to tractor-powered machinery, while the remaining households use either simple hand-held tools (48 percent) or animal-powered equipment (33 percent).8

For Asia and Northern Africa, evidence indicates that the already widespread use of animal traction in the 1960s facilitated the subsequent advance towards motorized mechanization. The process was further consolidated by the agricultural intensification of the green revolution, and then by rising rural wages due to industrialization and structural transformation.6 Similar patterns were evident in Latin America and the Caribbean, where it was largely private actors who drove agricultural mechanization. Governments, however, also played a key role, creating an enabling environment for mechanization, for example, through public programmes in Argentina, Costa Rica, Ecuador and Peru that gave access to credit at low interest rates and provided tax exemptions.10, 11 Moreover, several countries exempted agricultural machinery from import duties (e.g. Peru).10

The emergence of robust agricultural machinery manufacturing sectors in some countries in Asia (China and India) and Latin America and the Caribbean (Brazil, Mexico and, to some degree, Argentina) has led to diversification with machinery exported globally.11 This resulted in lower purchase costs of both small-scale equipment, such as two-wheel tractors (especially in Asia) and four-wheel tractors, and other machinery such as shallow tube-well pumps, threshers and grain mills.12, 13, 14 There is also evidence that the rise of rental machinery markets has helped spread agricultural mechanization by allowing small-scale agricultural producers to access agricultural machinery at an affordable cost.6

In sub-Saharan Africa in the 1960s and 1970s, there were many efforts to promote mechanization by providing subsidized machinery to farmers, running state and block farms, and setting up public hire centres, often with support from donors.15, 16 Such efforts proved costly and mostly failed due to poor infrastructure, inadequate investments in knowledge and skills development, poor maintenance capacities, lack of access to fuel and spare parts, absence of a real demand for mechanization, and governance challenges such as rent-seeking and corruption.6, 16 In sub-Saharan Africa, and other regions where mechanization remains limited, there appears to be a lack of public sector support for creating an enabling environment through the promotion of, among others, knowledge and skills development, access to finance, and rural infrastructure.11 Establishing commercially sustainable hire services should be a major priority in any strategy for sustainable agricultural mechanization in the region (see Box 2).

Box 2Understanding mechanization in sub-Saharan Africa

Agriculture reliant on human and animal power continues to dominate in sub-Saharan Africa, limiting productivity. The tractor is one of the most disseminated types of agricultural machinery (with varying degrees of success) of the past seven decades.15 However, tractors remain expensive and unaffordable for most farmers. Therefore, sustainable rental mechanisms are key for allowing farmers – in particular small-scale producers – to access mechanization. Tractor hire services operate in the region, involving both the traditional (four-wheel) tractor and – more recently and to a lesser degree – the power tiller (i.e. two-wheel tractor). In contrast to the negative image of government-operated tractor hire services, there are thousands of individuals across the region who own tractors and can provide tractor hire services to farmers. TROTRO Tractor in Ghana is a case in point (see Box 3).

The figure provides a snapshot of current use – through ownership or rental – of four-wheel (left) and two-wheel (right) tractors in selected sub-Saharan African countries for which data are available from the Living Standards Measurement Study - Integrated Surveys on Agriculture (LSMS-ISA) project.

Tractor ownership at the household level remains very low even for two-wheel tractors, which are usually less expensive. The availability of tractor rental services only slightly increases access to four-wheel tractors. The low uptake of two-wheel tractors, together with an almost non-existent rental market, highlights how suppliers are yet to establish fully operational and sustainable local franchises for the supply chains of these machines and spare parts.15 Establishing commercially sustainable hire services (through private or cooperative ownership) is a high priority in any strategy for sustainable agricultural mechanization in the region.

FIGURE Share of agricultural households with access to tractors, in selected countries

SOURCE: World Bank, 2022.
SOURCE: World Bank, 2022.17

Data on non-tractor-powered mechanization are even more limited, but evidence indicates that even in sub-Saharan Africa some stationary activities have been mechanized for a long time, such as mechanical mills for power-intensive milling.16 Across the world, mechanization continues to be limited for a range of operations including harvesting and weeding. Furthermore, although combine harvesters and stationary threshers are on the rise in various countries, they can only be used for harvesting cereals. With very few exceptions, fruit and vegetable production is scarcely mechanized across the globe.6

Regional averages mask important intraregional, and even national, differences

Although the average uptake of tractors has been higher in some regions than in others, there can also be significant variability within a region itself due to disparities in structural and agricultural transformation and technological change. For example, while Japan witnessed the rapid adoption of tractors in the 1960s, other countries in the region (e.g. Thailand) did not undergo a similar development until the 1990s–2000s.9 In China, on the other hand, the spread of tractor use began in the 1970s and 1980s, while in Bangladesh, India, Myanmar and Sri Lanka, it was recently estimated that up to 90 percent of farmland (mostly used for rice production) is prepared using motorized machinery.18, 19, 20, 21 Topographic conditions have also limited mechanization, or made its adoption uneven in some Asian countries.6, 14 For example, in Nepal, only 23 percent of agricultural producers use tractors and power tillers in the mountainous parts of the country, while this share reaches 46 percent in the flatter Terai zone. In Latin America and the Caribbean, there is significant variability between large- and small-scale farms; large-scale farms are much more mechanized than small-scale ones due to the latter being located, at least in part, in remote and hilly areas.10, 11, 22, 23

Even in the least mechanized subregions of sub-Saharan Africa, adoption levels are uneven across and within countries. For example, in 2000, tractors per 1 000 ha of arable land in Botswana and South Africa numbered 8 and 5, respectively, while in countries such as Madagascar, Mali and Senegal, they did not exceed 0.4. In Ghana, it is estimated that on average one-third of farm households use tractors (mainly for tillage), but the share differs from just 2 percent in forest zones to 88 percent in the savannah.6 In the United Republic of Tanzania, mechanization levels are highest in regions with commercial farming.24 In Nigeria, while 7 percent of producers use tractors, another 25 percent use their own or hired animal traction for land preparation.25 In Ethiopia, only around 1 percent of farm plots are mechanized using tractors, mainly in easy-to-mechanize wheat–barley systems, which are also dominated by large producers and have witnessed the emergence of service markets for wheat combining.

What the (limited) available data tell us about livestock and aquaculture mechanization

Data on the adoption of machinery for livestock and aquaculture production are either very scarce, very patchy or non-existent. The same applies to data for forestry. Analysis of the limited data show that livestock machinery (e.g. milking machines) is concentrated in high-income countries. On the other hand, in low- and middle-income countries, although present, such equipment is more likely to be used in large-scale production units. However, given the paucity and inconsistency of the data, it is difficult to appreciate the precise scenario in various contexts. Moreover, it is not clear what exactly constitutes a milking machine, nor how many cows are serviced by each machine. As the technology evolves, the number of cows milked by a machine goes up, and the number of machines may therefore go down. Denmark is a case in point: a major milk-producing country with a declining uptake of milking machines, there may have been a technology replacement towards more advanced methods not covered by the statistics.9 However, anecdotal evidence from one case study (Lely) points to the consolidation of dairy farms in Northern Europe as the underlying cause of the falling numbers of milking machines resulting from technology replacement and greater economies of scale.2

back to top TOP