Dachs et al header image
Automated car assembly line in a factory.(Image: Lenny Kuhne via Unsplash) 

Leaders and laggards in the diffusion of industrial robots

Differences between data sources make it challenging to identify the leaders and laggards in the adoption of industrial robots.

By Bernhard Dachs, Xiaolan Fu and Angela Jäger

Over the last 30 years, industrial robots have made an impressive leap from the pages of science fiction books to the shop floors of factories around the world. Industrial robots are pivotal for Industry 4.0, and their use in the service sector is rapidly increasing as well.1

The importance of robots is corroborated by recent research which includes indicators of robot penetration in econometric models. Several studies investigate the impact of industrial robots on wages and employment2, on employment and productivity3 , and on international production4.

Data sources for robot adoption

The widespread use of indicators of robot penetration infers that it should be easy to identify the leaders and laggards of industrial robot diffusion. This task, however, is more difficult than it initially seems.

Several data sources on robot adoption have become available in recent years. The best-known source is the International Federation of Robotics (IFR), which publishes annual data on robot shipments for a number of countries and industries. Data on robot intensity is also provided by the European national statistical offices in cooperation with EUROSTAT, and by the European Investment Bank (EIB) in its annual Investment Report. EUROSTAT data is only available for European countries, while the EIB collects information on European countries and the U.S.

Based on these data, countries can be ranked according to their robot intensity. The country rankings of the three different data sources deliver highly divergent results, as we will see below, which raises some questions about international comparisons of robot intensity.

The rankings in comparison

According to the IFR, Singapore is the country with the highest robot density, followed by the Republic of Korea and Germany (figure below). Robot intensity of European countries mostly lags behind Japan, the U.S. and China. Only a few European countries, namely Germany, Sweden and Denmark, have a high level of robot penetration. The countries with the lowest robot intensity in Europe are Eastern European countries, including Estonia, Romania and Croatia.

Number of robots per 10,000 employees in manufacturing (2016)

Note: Hover over the bar chart to see the data for each country or grouping.

Source: IFR.

Next, we examine EUROSTAT data (figure below). Accordingly, the highest level of robot penetration is found in Spain, Denmark, Finland and Italy. By contrast, IFR data ranks Germany, Sweden and Denmark as the countries with the highest robot intensity in Europe (see figure above). EUROSTAT finds that the lowest share of robots are being used in firms in Cyprus (1 per cent), Estonia, Greece, Lithuania, Hungary and Romania (figure below), which is consistent with IFR data.

Another interesting result provided by EUROSTAT data is that the level of robot adoption among Western European Union (EU) member states and the EU-28 average are the same (two right-hand columns in figure below). This is in stark contrast to many other indicators of innovation and technology diffusion, which reveal a considerable gap between EU member states in Western and in Central and Eastern Europe. One explanation could be the strong manufacturing base in some Central and Eastern European member states. Czechia, Poland, Hungary, Slovakia and Slovenia, for example, have large automotive and metal products industries, a factor which manifests itself in a higher share of firms that uses robots.

European enterprises that use industrial or service robots (2018)

Note: Share of enterprises in different European countries that use industrial or service robots. Hover over the bar chart to see the data for each country or grouping.

Source: EUROSTAT, ICT usage in enterprises.

The EIB’s Investment Survey also covers robot intensity. According to EIB data, the diffusion of robots in EU manufacturing industries is highest in small countries (figure below), namely Slovenia, followed by Finland, Austria, Denmark and Sweden. Malta, Cyprus and Ireland lag behind.

The Investment Survey also compares the penetration of robots between the EU and the U.S. The data suggest that robot intensity in the U.S. is only slightly higher than in the EU, which clearly contradicts the results of the IFR data. The IFR finds that there are 114 robots per 10,000 employees in European manufacturing firms compared to 217 per 10,000 employees in U.S. firms.

Manufacturing firms that use advanced robots (2018)

Note: Share of manufacturing firms that use advanced robots. CEEC are Central and Eastern European Countries. Hover over the bar chart to see the data for each country or grouping.

Source: European Investment Bank, Investment Survey 2019/20.

Making sense of the available data

The data provide different answers about who the leaders and laggards in the diffusion of industrial robots are. The gap between the U.S. and Europe is huge according to one data set, and small according to another. The IFR reports that Germany, Sweden and Denmark are the European leaders in robot use, while EUROSTAT ranks Spain, Denmark and Finland, and the EIB ranks Slovenia, Finland and Austria in leading positions in the EU.

How can we make sense of the data? Considering that all three data sources generally measure the same object—the adoption of industrial robots within firms—the divergent results can be explained by differences in how robot use is measured. There is indeed a considerable difference in how data are measured: the IFR counts the number of installed robots per 10,000 manufacturing employees, while EUROSTAT and the EIB focus on the share of firms that use robots in a given country or sector.

The difference between these two approaches becomes clear when we take the extreme case of a country where one single firm has adopted a very large number of robots: this results in a very high ranking in the IFR, while that same country would rank much lower in the EUROSTAT and EIB surveys. This may be the case in the automotive industry, for example, where a few large firms in Germany, Sweden and Italy heavily invest in robots. According to IFR data, around 40 per cent of all robots are adopted by firms in the automotive industry. This could also explain the spikes in IFR data for Germany, Sweden and Italy. The IFR and EUROSTAT rankings differ most for Finland, a country which does not have an automotive industry.

The numbers reported by the EIB are considerably higher than those published by EUROSTAT, although both surveys cover the year 2018. The most likely reason for this discrepancy are differing survey methods and different questions on robot adoption. EUROSTAT asks firms whether they use industrial or service robots while the IFR data entails actual shipments of robots, which differs from robots installed in firms. Moreover, as the diffusion of robots increases with firm size, the share of small firms in the EIB survey could be lower relative to that of the EUROSTAT survey.

To compare the IFR, EUROSTAT and EIB data, we normalize the country values to the EU average. Countries with a higher diffusion rate than the EU average have a value larger than 1, and vice versa (figure above). Only four countries (Finland, the Netherlands, France and Denmark) lie above the EU average in all three surveys. These countries can be considered leaders in robot adoption in Europe. Germany and Austria rank above the EU average according to the IFR and EIB data, but below the average in the EUROSTAT survey. Poland, Romania, Hungary, Greece and Estonia lie below the EU average in all three surveys, hence it is fair to say that these countries are not among the leaders in the diffusion of robots. Data for Cyprus, Lithuania, Bulgaria and Malta are only available in the EUROSTAT and EIB surveys, and they feature quite low in the ranking. That is, these countries seem to also be lagging behind.

Finally, we compare the three rankings with the Spearman rank correlation, a method that compares two vectors with identical items based on each item’s ranking. With an index value of 0.63, the EIB and the EUROSTAT surveys are most comparable, followed by EIB-IFR (correlation of 0.62) and EUROSTAT-IFR (correlation of 0.6). However, some small EU countries are not included in the IFR data, which makes the comparison difficult.

Data on robot diffusion (2016 and 2017)

Note: Normalized values for IFR, EUROSTAT and EIB data on robot diffusion. Hover over the bar chart to see the data for each country or grouping.

Source: Authors' calculation based on IFR, EUROSTAT, EIB.


Country comparisons are an important means for innovation and technology policy because they can reveal shortcomings of national innovation systems and point to possible areas for policy intervention. It is challenging, however, to derive such conclusions from the currently available data on robot adoption, with country rankings differing considerably depending on data source. This raises some questions about cross-country studies and rankings on robot diffusion.

One main reason for these differences is the concentration of installations in some very robot-intensive firms and industries. The number of installed robots is therefore a somewhat misleading indicator for country comparisons. This fact also casts some doubts on the results of empirical analyses on robots, productivity and employment, because the average effects may be debatable when robots are highly concentrated in just a few firms of the enterprise population.

These results call for international cooperation to identify indicators that deliver unambiguous results. We can draw a parallel here to other indicators of technological activity that have undergone a long process of refinement and standardization5.

  • Bernhard Dachs is Senior Scientist at the Center for Innovation Systems and Policy of the Austrian Institute of Technology (AIT).
  • Xiaolan Fu (傅晓岚) is Founding Director of the Technology and Management Centre for Development (TMCD), Professor of Technology and International Development and Fellow of Green Templeton College at the University of Oxford, United Kingdom. 
  • Angela Jäger is Researcher at Fraunhofer Institute for Systems and Innovation Research ISI in Karlsruhe, Germany.

Disclaimer: The views expressed in this article are those of the authors based on their experience and on prior research and do not necessarily reflect the views of UNIDO (read more).

Read next