Academy of Romanian Scientists  
Journal of Knowledge Dynamics  
Digital Transformation and the Working Environment:  
International Evidence on Safety and Training in  
Industry 4.0.  
Sonia GARCÍA-MORENO1  
1
University of Castilla-La Mancha (UCLM), 02071, Albacete, Spain; ORCID No. 0000-0001-7552-  
5280; (corresponding author); sonia.garcia10@alu.uclm.es  
Víctor-Raúl LÓPEZ-RUIZ2  
2
University of Castilla-La Mancha (UCLM), 02071, Albacete, Spain; ORCID No. 0000-0002-2850-  
1662; Victor.Lopez@uclm.es  
Received: February 21, 2026  
Revised: March 27, 2026  
Accepted: May 5, 2026  
Published: June 30, 2026  
Abstract: Improving the working environment is a structural dimension of Industry 4.0., where  
technological change affects occupational safety and continuous training. This paper examines  
whether Industry 4.0. transformation is associated with more robust labor conditions in the  
European Union and in Germany, Spain, and Romania. The analysis distinguishes between total  
economic activity and manufacturing to capture differences by sector in safety and workforce  
adaptation. The study uses harmonized Eurostat data covering different time spans according to  
the indicator and combines standard measures with derived indicators constructed by the  
authors. These include accident incidence, severity composition, relative manufacturing  
differentials and training dynamics. The results show a general decline in accident incidence but  
also reveal uneven trajectories in the qualitative structure of occupational risk and in the  
development of training capacity. Germany presents the most stable trajectory, with declining  
accident incidence, contained severity indicators and an established training structure in firms.  
Spain shows a more ambivalent pattern. Adult learning improves markedly, but manufacturing  
remains the clearest safety concern because incidence and severity gaps persist. Romania  
records rapid gains in adult learning and low accident incidence, although these coexist with a  
heavier severity burden and a weaker training base in enterprises. The paper does not reject the  
view that Industry 4.0. can improve safety and training. Rather, it shows that improvement  
remains partial and uneven when the analysis also considers accident severity and training  
provided by enterprises. Occupational safety and training, therefore, constitute core conditions  
for a balanced, resilient, and sustainable industrial transition.  
Keywords: Industry 4.0.; working environment; occupational safety; training systems; digital  
transformation; comparative analysis.  
Introduction  
Industry 4.0., understood as a structured paradigm of industrial transformation based on  
the integration of digital, physical, and organizational systems, has usually attracted  
attention because of its potential effects on productivity, efficiency, and technological  
upgrading at firm and sectoral levels (Lasi et al., 2014; Brynjolfsson & McAfee, 2014). This  
perspective is essential, but it remains incomplete when it leaves the working  
environment outside the core analysis of industrial change. Technological progress  
reshapes organizational structures, risk profiles, and labor interactions, influencing how  
work takes place and how firms generate and manage risks (Acemoglu & Restrepo, 2020;  
Autor, 2015).  
From a knowledge management perspective, these changes go beyond the adoption of  
innovative technologies. They also involve the way organizations create, transfer, and  
apply knowledge in increasingly complex production environments. Industry 4.0.  
How to cite  
García-Moreno, G., & López-Ruiz, V.R. (2026). Digital Transformation and the Working  
Environment: International Evidence on Safety and Training in Industry 4.0 Journal of Knowledge  
Dynamics, Vol. 3, No.1, pp. 05-20. https://doi.org/10.56082/jkd.2026.1.5 ISSN ONLINE 3061-2640  
6 | Sonia GARCÍA-MORENO, Víctor-Raúl LÓPEZ-RUIZ  
Digital Transformation and the Working Environment: International Evidence on Safety and Training  
in Industry 4.0.  
intensifies the circulation of technical, organizational, and preventive knowledge, as  
workers and firms must interpret data, codify procedures, adapt routines, and learn from  
operational experience. This is particularly relevant for the two dimensions examined in  
this paper. Occupational safety depends on the capacity to transform risk information into  
shared preventive practices, while continuous training reflects the mechanisms through  
which firms update skills and convert technological change into usable organizational  
knowledge. In this sense, safety and training are not only labor outcomes. They are also  
expressions of knowledge dynamics within industrial systems.  
Assessing the sustainability of Industry 4.0. requires more than attention to output or  
operational efficiency. A more complete evaluation must consider the conditions under  
which technological transformation unfolds, particularly where production systems make  
more intensive use of data and become more interconnected and organizationally  
complex. In such contexts, the quality of the working environment becomes a relevant  
dimension of industrial performance rather than a secondary outcome of technological  
change.  
Among the different elements that shape the working environment, occupational safety  
and continuous training are especially significant. Safety reflects whether technological  
and organizational changes lead to more secure and controllable production processes.  
Training, in turn, conditions the capacity of workers and firms to adapt to new tasks, tools,  
and requirements for decision making associated with digital transformation. Taken  
together, these two dimensions offer a useful entry point for assessing whether Industry  
4.0. is associated with more robust and sustainable forms of industrial development.  
Despite growing interest in the social and organizational implications of digital  
transformation, empirical evidence jointly examining occupational safety and training  
within the context of Industry 4.0. remains limited, particularly from a comparative  
European perspective. Much of the existing literature has focused either on productivity  
and innovation outcomes or on broader labor market changes, without fully integrating  
the dimensions of technological change that operate within the workplace (Goos et al.,  
2014; Acemoglu & Restrepo, 2020). This gap is especially relevant in manufacturing,  
where technological intensity and exposure to occupational risks are particularly  
pronounced.  
This paper addresses this gap by analyzing the evolution of occupational safety and  
training as key dimensions of the working environment in the context of Industry 4.0. The  
analysis adopts a comparative perspective focused on the European Union and three  
national cases, Germany, Spain and Romania. These countries reflect different industrial  
structures, technological trajectories and labor market conditions, allowing the  
identification of both convergent patterns and persistent structural differences. The  
empirical approach is based on European statistics and distinguishes between total  
economic activity and manufacturing to capture dynamics that are specific to each sector  
and associated with technological transformation. By examining trends in accident  
incidence, severity and participation in education and training, the paper provides a  
structured assessment of how the working environment evolves alongside digital change.  
It argues that improvements in safety and training should be understood as structural  
components of Industry 4.0. rather than as secondary outcomes, and that their joint  
analysis offers a more complete view of the quality and sustainability of industrial  
transformation.  
The paper is organized as follows. The next section reviews the literature linking Industry  
4.0., knowledge management, occupational safety and training. The methodology section  
describes the data sources, indicators and comparative approach used in the empirical  
analysis. The results and discussion section examines safety trends, severity patterns,  
training dynamics and the comparative trajectories of Germany, Spain and Romania. The  
concluding section summarizes the main findings and discusses their implications for the  
Academy of Romanian Scientists | 7  
Journal of Knowledge Dynamics  
Vol. 3 (2026) No.1, pp. 5-20  
assessment of Industry 4.0. as an industrial transition focused on people and grounded in  
knowledge management.  
Literature review.  
The literature on industrial transformation has long shown that technological investment  
alone does not guarantee sustained improvements in firm performance. Early  
contributions on high-performance work systems argued that firms combining  
technological change with organizational redesign, clearer task structures, employee  
involvement and continuous training tend to achieve better results than those relying on  
capital upgrading in isolation (Appelbaum et al., 2000Ichniowski et al., 1997; ). From this  
perspective, the working environment is not external to production performance but  
embedded within it. A safer, more stable and better organized workplace may reduce  
disruption, support learning processes and contribute to more resilient operational  
outcomes.  
This argument is particularly relevant in the context of Industry 4.0., where digital  
transformation modifies both the technical and organizational architecture of production.  
The introduction of automation, data analytics, connected systems and cyber-physical  
integration may reduce exposure to certain routine risks and improve process control.  
However, these effects are not automatic. Technological change may also generate new  
forms of complexity, intensification and cognitive demand if it is not accompanied by  
appropriate organizational adaptation and clear governance of work processes  
(Brynjolfsson & Hitt, 2000; Bloom et al., 2012). The relationship between Industry 4.0. and  
the quality of the working environment should therefore be understood as conditional  
rather than linear.  
Within this broader discussion, occupational safety occupies a central position. Safety  
outcomes reflect not only the technological characteristics of production systems but also  
the quality of work organization, risk prevention and managerial coordination. The  
literature on job design and psychosocial work environments has shown that  
performance and well-being depend heavily on how demands, control and support are  
configured within the workplace (Karasek & Theorell, 1990). In parallel, more recent  
research on safety management has stressed that increasingly complex production  
environments require more adaptive and integrated approaches to risk control,  
particularly where advanced technologies alter routines, interfaces and decision  
structures (Hollnagel, 2014; Badri et al., 2018). In this sense, digital upgrading does not  
automatically improve safety. Safety depends on how technology is embedded in  
organizational practice.  
Ergonomics and human-centered design further reinforce this interpretation. Research  
has shown that aligning workplace design, tools and interfaces with human capabilities  
contributes not only to better health and safety outcomes but also to quality, process  
stability and organizational learning (Bratianu & Anagnoste, 2011; Bratianu et al., 2011;  
Dul & Neumann, 2009; Kadir & Broberg, 2021). Similarly, the literature on safety climate  
has highlighted that the preventive performance of organizations depends on shared  
perceptions, managerial commitment and the practical integration of safety into everyday  
work (Zohar, 1980). These contributions are especially relevant for Industry 4.0., where  
technological sophistication increases the importance of coordination, interpretative  
capacity and the management of hybrid human machine environments.  
Training constitutes the second major pillar of this framework. The literature on skill-  
biased technological change has consistently shown that digitalization raises the demand  
for higher-order skills while reducing the relative importance of routine tasks (Autor et  
al., 2003; Acemoglu & Autor, 2011). In industrial contexts, this implies that the capacity to  
benefit from digital technologies depends not only on the acquisition of equipment and  
systems but also on the ability of workers to interpret data, interact with digital tools and  
adapt to reconfigured production processes. Continuous training, therefore, becomes a  
8 | Sonia GARCÍA-MORENO, Víctor-Raúl LÓPEZ-RUIZ  
Digital Transformation and the Working Environment: International Evidence on Safety and Training  
in Industry 4.0.  
central mechanism through which firms translate technological investment into effective  
organizational capability (OECD, 2019; World Economic Forum, 2020). Recent European  
policy approaches increasingly reflect this broader understanding of industrial  
transformation. Human-centered, sustainable and resilient industry is now framed as a  
strategic objective, implying that technological progress should be assessed alongside  
working conditions, learning capacity and wider social outcomes (European Commission,  
2021; Eurofound, 2021). Nevertheless, the empirical literature still offers limited  
comparative evidence integrating occupational safety and training within a common  
analytical framework. This is particularly visible in cross-country analyses of  
manufacturing, where the organizational and labor dimensions of Industry 4.0. remain  
less systematically connected than productivity or innovation indicators. On that basis,  
this paper examines safety and training jointly as complementary dimensions of the  
working environment to provide a more comprehensive interpretation of industrial  
transformation.  
Methodology.  
A descriptive and comparative approach is used to examine occupational safety and  
training as two key dimensions of the working environment in the context of Industry 4.0.  
The analysis focuses on the European Union, Germany, Spain and Romania, selected  
because they reflect different industrial structures, technological trajectories and labor  
market conditions within Europe. The empirical analysis relies on harmonized Eurostat  
statistics and distinguishes between total economic activity and manufacturing to capture  
differences by sector associated with industrial transformation.  
The occupational safety dimension uses Eurostat data on non-fatal accidents at work from  
2015 to 2023. The analysis considers accident incidence and the severity structure of  
accidents involving at least four days of absence. Fatal accidents fall outside the empirical  
scope of the paper. In addition to levels and trends, the study uses derived indicators  
including cumulative changes, annual rates of variation, averages before and after the  
pandemic, relative manufacturing differentials and changes in severity composition.  
Severity is examined through the share of accidents involving at least 21 days of absence,  
at least one month of absence, and permanent incapacity or at least 183 days of absence.  
A lost days per case index is also calculated using fixed interval weights. This index is not  
intended to provide an exact medical measure of injury burden. It offers a consistent  
comparative approximation of the temporal load associated with accidents.  
The training dimension combines two Eurostat sources. The first source measures  
participation in education and training among employed adults aged 18 to 64 from 2015  
to 2024. The second source measures the proportion of enterprises providing continuing  
vocational training or other training related to the job to their employees in the available  
survey waves for 2010, 2015, and 2020. For both dimensions, the analysis combines direct  
indicators with derived comparative measures, including period averages, gaps with  
respect to the European Union benchmark, and manufacturing differentials.  
The methodology rests on three decisions. It uses harmonized official statistics, combines  
direct and derived indicators, and treats occupational safety and training as  
complementary dimensions of the working environment. No causal relationship between  
digitalization and labor outcomes is claimed. The aim is to identify consistent comparative  
patterns that help to interpret the quality and sustainability of Industry 4.0. trajectories  
across national and sectoral settings.  
Results and discussions.  
Safety trends in total activity and manufacturing.  
The empirical analysis begins with non-fatal accident incidence, which provides the  
clearest initial view of how occupational safety evolved in total economic activity and  
Academy of Romanian Scientists | 9  
Journal of Knowledge Dynamics  
Vol. 3 (2026) No.1, pp. 5-20  
manufacturing. Figure 1 shows the incidence rate for the European Union, Germany, Spain  
and Romania between 2015 and 2023. The Romanian series is represented on a secondary  
axis because its level is markedly lower than that of the other cases.  
The figure shows a general downward trajectory in all four cases, although with  
differences in level and sectoral intensity. In total economic activity, the European Union,  
Germany, and Spain record declining incidence, while Romania remains at markedly  
lower levels throughout the period. In manufacturing, incidence also falls, but the sector  
remains systematically more exposed than the overall economy. Spain displays the  
highest manufacturing incidence in the comparison, Germany follows a more contained  
downward path, and Romania remains on a separate, lower incidence scale. Overall,  
Figure 1 shows improvement in accident incidence, but not the disappearance of the  
manufacturing safety gap.  
Total activity  
3000,00  
2500,00  
2000,00  
1500,00  
1000,00  
100,00  
80,00  
60,00  
40,00  
20,00  
0,00  
2015  
2016  
2017  
2018  
2019  
2020  
2021  
2022 2023  
EU27  
Romania  
Linear (Spain)  
Germany  
Spain  
Linear (Germany)  
Linear (EU27 )  
Linear (Romania)  
Manufacturing  
4500  
4000  
3500  
3000  
2500  
2000  
1500  
150  
130  
110  
90  
70  
50  
2015  
2016  
EU27  
2017  
2018  
2019  
2020  
2021  
2022  
2023  
Germany  
Spain  
Linear (Germany)  
Romania  
Linear (EU27 )  
Linear (Spain)  
Linear (Romania)  
Figure 1. Non-fatal accident incidence rate in total economic activity and manufacturing.  
EU(27), Germany, Spain and Romania, 2015-2023.  
(Source: Authors’ own elaboration based on Eurostat>Home>Database>Cross cutting topics>Quality  
of employment>Safety and ethics of employmen).  
Table 1 complements Figure 1 by summarizing cumulative changes, relative changes and  
annual rates of variation before and after 2019. It confirms the general reduction in  
accident incidence but also shows that the pace of improvement differs across countries  
and sectors. In total economic activity, the European Union, Germany and Spain record  
substantial absolute declines, while Romania shows a smaller absolute change because it  
starts from a much lower incidence base. In manufacturing, the contrast is stronger. The  
decline is clear in the European Union, Germany and Romania, whereas Spain records only  
a limited reduction over the whole period. The table therefore shows that incidence falls  
10 | Sonia GARCÍA-MORENO, Víctor-Raúl LÓPEZ-RUIZ  
Digital Transformation and the Working Environment: International Evidence on Safety and Training  
in Industry 4.0.  
overall, but that the improvement is uneven and particularly weak in Spanish  
manufacturing.  
Table 1. Metrics of non-fatal accident incidence: cumulative change and rates of variation in  
total economic activity and manufacturing. EU(27), Germany, Spain and Romania, 2015-2023.  
Total economic  
activity  
Absolute  
change 15-23  
-319.00  
Relative  
change 15-23  
-0.19  
CAGR  
15-19  
-0.01  
-0.04  
-0.02  
0.03  
CAGR  
15-19  
-0.02  
-0.04  
0.00  
CAGR  
19-23  
-0.04  
-0.04  
-0.02  
-0.08  
CAGR  
19-23  
-0.02  
-0.02  
-0.01  
-0.10  
EU(27)  
Germany  
Spain  
Romania  
-533.77  
-447.26  
-17.03  
-0.28  
-0.16  
-0.22  
Manufacturing  
Absolute  
change 15-23  
-295.77  
Relative  
change 15-23  
-0.15  
EU(27)  
Germany  
Spain  
-552.88  
-116.28  
-24.84  
-0.22  
-0.03  
-0.22  
Romania  
0.04  
(Source: Authors’ own elaboration based on Eurostat>Home>Database>Cross cutting topics>Quality  
of employment>Safety and ethics of employment)  
Table 2 compares average incidence before and after the pandemic and the relative  
position of each country with respect to the European benchmark. This distinction is  
relevant because a lower incidence level over time does not necessarily mean convergence  
toward the European average. The comparison shows that average incidence declines  
between the period before the pandemic and the period after the pandemic in all four  
cases. However, absolute improvement does not always imply convergence toward the  
European benchmark. Germany moves close to the EU average in total economic activity,  
while Spain remains clearly above it and even worsens its relative position by 2023. In  
manufacturing, the sectoral contrast is stronger: Spain continues to show the largest  
relative gap, Germany remains moderately above the EU average, and Romania stays well  
below it. This confirms that the Spanish manufacturing case combines improvement over  
time with persistent relative overexposure.  
Table 2. Non-fatal accident incidence: averages for 2017-2019 and 2021-2023, post-pre  
difference and relative position. EU(27), Germany, Spain and Romania.  
Total economic  
activity  
Pre-  
average  
Post-  
average  
Post-pre  
difference  
Relative position  
vs EU(27)  
(17-19)  
(21-23)  
2019  
0.00  
0.02  
0.56  
-0.95  
2023  
0.00  
0.00  
0.67  
-0.96  
EU(27)  
Germany  
Spain  
1662.35  
1730.63  
2726.52  
84.90  
1472.84  
1484.73  
2377.15  
55.74  
-189.51  
-245.90  
-349.37  
Romania  
-29.16  
Manufacturing  
Pre-  
average  
(17-19)  
Post-  
average  
(21-23)  
Post-pre  
difference  
Relative  
position vs  
EU(27)  
2019  
2023  
0.00  
0.14  
1.27  
-0.95  
EU(27)  
Germany  
Spain  
1906.03  
2194.54  
4071.03  
126.13  
1744.01  
2073.72  
3762.05  
84.89  
-162.02  
-120.82  
-308.97  
-41.24  
0.00  
0.14  
1.14  
-0.93  
Romania  
(Source: Authors’ own elaboration based on Eurostat>Home>Database>Cross cutting topics>Quality  
of employment>Safety and ethics of employment)  
Academy of Romanian Scientists | 11  
Journal of Knowledge Dynamics  
Vol. 3 (2026) No.1, pp. 5-20  
Table 3 examines whether manufacturing remains more exposed than total economic  
activity within each country. It uses the ratio between manufacturing incidence and total  
incidence, together with the excess relative to the European benchmark. The results  
confirm that manufacturing remains more exposed to accidents than total economic  
activity in all cases, with ratios above one in both 2019 and 2023. The differential widens  
in the European Union, Germany and Spain, while it declines in Romania. Spain records  
the highest manufacturing differential in both years, rising from 1.58 to 1.65, which  
confirms the persistence of sectoral overexposure. Romania reduces its differential,  
although manufacturing remains more exposed than the national economy. The table  
therefore reinforces one of the central findings of the safety analysis: lower incidence has  
not eliminated the structural risk differential of manufacturing.  
Table 3. Manufacturing incidence differential relative to total economic activity, 2019 and  
2023. Excess relative to EU(27) in percentage points.  
EU(27)  
Germany  
Spain  
1.15  
1.28  
1.58  
1.53  
1.21  
1.38  
1.65  
1.44  
15.17%  
28.49%  
58.18%  
53.47%  
21.14%  
37.51%  
65.16%  
44.21%  
0.06  
0.09  
0.07  
-0.09  
0.13  
0.43  
0.38  
0.16  
0.44  
0.23  
Romania  
(Source: Authors’ own elaboration based on Eurostat>Home>Database>Cross cutting topics>Quality  
of employment>Safety and ethics of employment)  
Overall, the incidence of non-fatal accidents evolves favorably, but not uniformly.  
Incidence declines in total economic activity and manufacturing, yet manufacturing  
remains structurally more exposed in all cases. Germany combines decline with relative  
convergence, Spain combines absolute improvement with persistent manufacturing gaps,  
and Romania improves from a markedly lower incidence base while reducing its sectoral  
differential. The safety dimension therefore requires attention not only to average decline,  
but also to sectoral distribution and comparative position.  
Severity and quality of safety improvement.  
The previous subsection showed that non-fatal accident incidence generally declined.  
However, lower frequency does not necessarily imply an equivalent improvement in the  
quality of occupational safety. A decline in accidents may coexist with a less favorable  
severity profile if the remaining cases become longer or more demanding. For this reason,  
the analysis now moves from incidence to severity.  
Table 4 examines the severity composition of accidents involving at least four days of  
absence. It considers three thresholds, accidents involving at least 21 days of absence, at  
least one month of absence, and permanent incapacity or at least 183 days of absence. The  
results indicate a shift toward longer durations in part of the comparison. Germany shows  
the most contained profile. The European Union records moderate increases. Spain  
remains above the European average and deteriorates slightly. Romania records the  
highest shares and the sharpest worsening, especially in long duration cases. These results  
show that a decline in incidence does not necessarily lead to an equivalent improvement  
in severity.  
12 | Sonia GARCÍA-MORENO, Víctor-Raúl LÓPEZ-RUIZ  
Digital Transformation and the Working Environment: International Evidence on Safety and Training  
in Industry 4.0.  
Table 4. Severity composition of accidents with at least four lost days. Total economic activity  
and manufacturing. EU(27), Germany, Spain and Romania, 2019 and 2023.  
Total  
economic  
activity  
Long  
duration  
share  
2019  
(≥21  
days)  
0.33  
Share  
≥1  
month  
2019  
Share  
183  
days+  
2019  
Long-  
duration ≥1  
share  
2023  
(≥21  
Share  
Share  
183  
days+  
2023  
month  
2023  
days)  
EU(27)  
0.23  
0.04  
0.34  
0.24  
0.05  
Germany  
Spain  
Romania  
0.22  
0.37  
0.43  
0.15  
0.25  
0.32  
0.02  
0.03  
0.04  
0.22  
0.39  
0.54  
0.16  
0.27  
0.40  
0.02  
0.04  
0.05  
Manufacturing Long  
duration  
Share  
≥1  
Share  
183  
Long-  
duration ≥1  
Share  
Share  
183  
share  
2019  
(≥21  
month  
2019  
days+  
2019  
share  
2023  
(≥21  
month  
2023  
days+  
2023  
days)  
days)  
EU(27)  
Germany  
Spain  
0.30  
0.20  
0.03  
0.30  
0.21  
0.12  
0.25  
0.44  
0.04  
0.18  
0.36  
0.48  
0.12  
0.24  
0.37  
0.02  
0.03  
0.04  
0.17  
0.37  
0.59  
0.02  
0.03  
0.05  
Romania  
(Source: Authors’ own elaboration based on Eurostat>Home>Database>Cross cutting topics>Quality  
of employment>Safety and ethics of employment)  
Table 5 reports changes in severity composition from 2019 to 2023. It shows whether  
longer absence categories gained weight within accidents involving at least four days of  
absence and whether this change was similar across countries and sectors. The results  
confirm that the change in severity composition is uneven. Germany remains broadly  
stable, while the European Union records moderate increases. Spain shows a clearer  
deterioration, especially in long-duration cases and absences of at least one month.  
Romania displays by far the strongest worsening in both total economic activity and  
manufacturing. The table therefore reinforces the previous result by showing that the  
qualitative structure of safety improvement differs substantially across countries.  
Table 5. Changes in accident severity composition, 2019-2023. Total economic activity and  
manufacturing. EU(27), Germany, Spain and Romania, percentage points.  
Total economic Δ Long duration cases Δ ≥1 month  
Δ 183 days +  
activity  
EU(27)  
Germany  
Spain  
0.94  
-0.14  
1.57  
10.74  
1.11  
0.15  
1.70  
8.67  
0.68  
0.04  
0.52  
0.92  
Romania  
Manufacturing  
EU(27)  
Δ Long duration cases Δ ≥1 month  
Δ 183 days +  
0.58  
-0.30  
1.54  
0.77  
0.06  
1.63  
6.97  
0.60  
0.31  
0.57  
0.79  
Germany  
Spain  
Romania  
10.72  
(Source: Authors’ own elaboration based on Eurostat>Home>Database>Cross cutting topics>Quality  
of employment>Safety and ethics of employment)  
Academy of Romanian Scientists | 13  
Journal of Knowledge Dynamics  
Vol. 3 (2026) No.1, pp. 5-20  
Table 6 summarizes the severity burden through a lost days per case index based on fixed  
interval weights. The index does not provide an exact medical measure of injury duration,  
but it gives a consistent comparative approximation of the temporal load associated with  
accidents. The same pattern appears in Table 6. The index rises in the European Union,  
Spain and Romania, while Germany remains almost unchanged. Romania records both the  
highest values and the strongest increase in total economic activity and manufacturing.  
This confirms that, even where accident incidence falls, the average duration associated  
with accidents may become heavier. The result is important because it connects accident  
severity with its organizational burden.  
Table 6. Lost days per case index. Total economic activity and manufacturing. EU(27),  
Germany, Spain and Romania, 2019 and 2023.  
Total economic activity  
EU(27)  
2019  
29.84  
22.84  
29.52  
34.12  
2019  
27.34  
20.48  
28.67  
36.07  
2023  
31.18  
22.92  
30.96  
39.70  
2023  
28.41  
20.84  
30.14  
40.92  
Germany  
Spain  
Romania  
Manufacturing  
EU(27)  
Germany  
Spain  
Romania  
(Source: Authors’ own elaboration based on Eurostat>Home>Database>Cross cutting topics>Quality  
of employment>Safety and ethics of employment.  
Note: Average weights by duration interval = 12, 26, 60 and 183 days)  
The severity evidence qualifies the favorable picture reflected by incidence trends.  
Accident frequency declines across the comparison, but severity does not follow the same  
path. Longer absences gain relative weight in the European Union, Spain, and Romania,  
and the lost days per case index also increases in those cases. Germany shows the most  
stable profile, Spain combines falling incidence with a less favorable severity pattern, and  
Romania records the sharpest deterioration despite maintaining low incidence levels. For  
this reason, the assessment of occupational safety must consider both frequency and  
severity.  
Training dynamics and learning capacity.  
Training is examined as the adaptive dimension of the working environment. It captures  
the capacity of workers and firms to update skills, absorb technological change and  
reorganize work in more digitalized production settings. The analysis therefore considers  
adult participation in education and training together with the training effort made by  
enterprises. Table 7 reports participation in education and training among employed  
adults between 2015 and 2024. Adult learning increases after 2020 in most cases, but the  
pattern is not uniform. The European Union records a gradual rise, Germany remains  
comparatively flat, and Spain and Romania show the strongest acceleration. In  
manufacturing, participation is generally lower at the beginning of the period, but Spain  
and Romania rise above the EU manufacturing average by 2024. This indicates that the  
strongest post-2020 intensification of adult learning is concentrated in Spain and  
Romania.  
14 | Sonia GARCÍA-MORENO, Víctor-Raúl LÓPEZ-RUIZ  
Digital Transformation and the Working Environment: International Evidence on Safety and Training  
in Industry 4.0.  
Table 7. Participation of the employed population in education and training. Total economic  
activity and manufacturing. EU(27), Germany, Spain and Romania, 2015-2024  
Total economic activity  
2015  
13.70  
12.40  
12.00  
1.80  
2016  
13.80  
12.60  
11.30  
1.50  
2017  
13.90  
12.50  
11.90  
1.40  
2018  
14.10  
12.50  
12.60  
1.20  
2019  
2020  
2021  
14.40  
12.00  
17.30  
7.50  
2022  
15.70  
12.70  
18.40  
8.10  
2023  
16.70  
12.70  
18.80  
9.90  
2024  
17.40  
13.90  
19.00  
13.10  
2024  
11.60  
9.60  
EU(27)  
14.30  
12.60  
12.30  
Germany  
12.30  
13.00  
Spain  
12.50  
Romania  
1.60  
Manufacturing  
2019  
1.10  
2015  
9.20  
2016  
9.30  
2017  
9.20  
2018  
9.10  
2020  
2021  
8.80  
2022  
10.00  
8.60  
2023  
10.90  
8.70  
EU(27)  
9.30  
7.90  
Germany  
9.10 9.40  
Spain  
8.40  
Romania  
2.20 1.60  
9.50  
9.40  
8.20  
9.20  
8.10  
8.30  
8.20  
8.00  
8.10  
8.70  
10.90  
11.70  
13.00  
13.40  
1.40  
1.40  
1.10  
0.80  
6.70  
7.40  
9.60  
13.20  
(Source: Authors’ own elaboration based on Eurostat> Home>Database>Population and social  
conditions>Education and training)  
Tables 8 and 9 clarify whether the increase in adult learning reflects a temporary  
fluctuation or a more persistent shift. The comparison between 2017-2019 and 2021-  
2023 shows moderate improvement in the European Union, almost no progress in  
Germany and strong increases in Spain and Romania. The gap with respect to the EU  
benchmark confirms this divergence. Germany moves further below the European  
average by 2024, while Spain moves above it in both total economic activity and  
manufacturing. Romania narrows its gap substantially in total economic activity and  
surpasses the EU manufacturing average by 2024. Adult learning therefore shows  
convergence in Spain and Romania, but not in Germany.  
Table 8. Pre- and post-period averages for adult learning participation among the employed  
population. Total economic activity and manufacturing. EU(27), Germany, Spain and  
Romania, 2017-2019 and 2021-2023.  
Total economic activity  
EU(27)  
Average 17-19 Average 21-23  
Post-pre difference  
14.10  
12.53  
12.33  
1.40  
15.60  
12.47  
18.17  
8.50  
1.50  
-0.07  
5.83  
7.10  
Germany  
Spain  
Romania  
Manufacturing  
EU(27)  
Average 17-19 Average 21-23  
Post-pre difference  
9.20  
8.83  
8.17  
1.37  
9.90  
8.47  
11.87  
7.90  
0.70  
-0.37  
3.70  
6.53  
Germany  
Spain  
Romania  
(Source: Authors’ own elaboration based on Eurostat> Home>Database>Population and social  
conditions>Education and training)  
Academy of Romanian Scientists | 15  
Journal of Knowledge Dynamics  
Vol. 3 (2026) No.1, pp. 5-20  
Table 9. Gap in adult learning participation relative to the EU. Total economic activity and  
manufacturing. EU(27), Germany, Spain and Romania, 2019 and 2024.  
Total economic activity  
EU(27)  
Gap 2019  
Gap 2024  
Gap 2024  
0.00  
-1.70  
-1.80  
0.00  
-3.50  
1.60  
Germany  
Spain  
Romania  
-12.70  
-4.30  
Manufacturing  
Gap 2019  
EU(27)  
Germany  
Spain  
0.00  
-0.20  
-0.90  
-7.10  
0.00  
-2.00  
1.80  
1.60  
Romania  
(Source: Authors’ own elaboration based on Eurostat> Home>Database>Population and social  
conditions>Education and training)  
Participation by workers does not exhaust the training dimension. Tables 10 and 11 shift  
the focus to enterprises providing continuing vocational training. This indicator captures  
the organizational commitment of firms to workforce development. The results show a  
polarized pattern. Germany and Spain remain above the EU average in both total economic  
activity and manufacturing, although Spain’s advantage narrows by 2020. Romania  
remains far below the European benchmark in all years and sectors. This contrast is  
important because it shows that a marked increase in adult learning does not  
automatically imply a comparable commitment by firms to provide training.  
Table 10. Enterprises that provide continuing vocational training to their employees in total  
economic activity and manufacturing. EU(27), Germany, Spain, and Romania, 2010, 2015, and  
2020.  
Total economic activity Year  
2010  
63.60  
72.80  
74.90  
24.10  
2015  
70.50  
77.30  
86.00  
26.70  
2020  
67.40  
77.20  
73.20  
17.50  
EU(27)  
Germany  
Spain  
Romania  
Manufacturing  
Year  
2010  
60.90  
72.80  
74.30  
24.40  
2015  
69.50  
79.90  
87.20  
27.30  
2020  
70.50  
78.20  
74.00  
21.70  
EU(27)  
Germany  
Spain  
Romania  
(Source: Authors’ own elaboration based on Eurostat> Home>Database>Population and social  
conditions>Education and training)  
16 | Sonia GARCÍA-MORENO, Víctor-Raúl LÓPEZ-RUIZ  
Digital Transformation and the Working Environment: International Evidence on Safety and Training  
in Industry 4.0.  
Table 11. Gap relative to the EU in enterprises providing continuing vocational training. Total  
economic activity and manufacturing. EU(27), Germany, Spain and Romania, 2010, 2015 and  
2020.  
Total economic activity  
EU(27)  
Gap 2010  
Gap 2015  
Gap 2020  
0.00  
9.20  
0.00  
6.80  
0.00  
9.80  
Germany  
Spain  
Romania  
11.30  
-39.50  
15.50  
-43.80  
5.80  
-49.90  
Manufacturing  
EU(27)  
Gap 2010  
Gap 2015  
Gap 2020  
0.00  
11.90  
13.40  
-36.50  
0.00  
10.40  
17.70  
-42.20  
0.00  
7.70  
3.50  
Germany  
Spain  
Romania  
-48.80  
(Source: Authors’ own elaboration based on Eurostat> Home>Database>Population and social  
conditions>Education and training)  
A final issue concerns the relative position of manufacturing within each country. Table  
12 examines whether manufacturing is more or less intensive in training than the overall  
economy. The pattern is mixed but informative. In the European Union, manufacturing  
moves from a negative differential to a positive one by 2020. Germany and Spain also show  
positive differentials after 2010, while Romania records the strongest positive differential  
in 2020. This does not mean that Romanian firms provide high levels of training in  
absolute terms. Instead, it indicates that, within Romania, manufacturing is relatively  
more engaged in training provision than the rest of the economy.  
Table 12. Industrial differential in enterprise training provision, manufacturing relative to  
total economic activity. EU(27), Germany, Spain and Romania, 2010, 2015 and 2020.  
2010  
-2.70  
0.00  
-0.60  
0.30  
2015  
-1.00  
2.60  
1.20  
0.60  
2020  
3.10  
1.00  
0.80  
4.20  
EU(27)  
Germany  
Spain  
Romania  
(Source: Authors’ own elaboration based on Eurostat> Home>Database>Population and social  
conditions>Education and training)  
The training evidence points to differentiated trajectories. Adult learning expands  
strongly in Spain and Romania after 2020, while Germany loses relative ground. Training  
provided by enterprises follows another pattern. Germany and Spain remain above the  
European benchmark, whereas Romania stays far below it. The training dimension  
therefore cannot be assessed using a single indicator because individual participation and  
firms’ own training effort do not move in the same way.  
Comparative interpretation of national trajectories.  
The results do not point to a single pattern linking digital transformation and the working  
environment. They reveal three distinct national trajectories, each combining safety and  
training differently. Reading these dimensions together matters because the quality of the  
transition toward Industry 4.0. depends not only on whether accident incidence falls or  
training expands, but also on how both processes interact within each national and  
sectoral context.  
These trajectories also connect with previous research on Industry 4.0., occupational  
safety, training, and knowledge management. The literature recognizes that digital  
technologies may improve monitoring, process control, and safety management, but it  
also warns that new occupational safety and health challenges arise when technological  
change is not accompanied by prevention systems, work organization, and human  
Academy of Romanian Scientists | 17  
Journal of Knowledge Dynamics  
Vol. 3 (2026) No.1, pp. 5-20  
adaptation (Badri et al., 2018; Rosen, 2022; Forcina et al., 2021; Arana-Landín et al., 2023).  
The evidence presented here is consistent with this more nuanced interpretation.  
Accident incidence declines, but severity worsens in the European Union, Spain, and  
Romania. Adult learning expands in Spain and Romania, but training provided by  
enterprises remains much more uneven. From a knowledge management perspective, this  
confirms that the quality of Industry 4.0. depends not only on technological diffusion, but  
also on the capacity to transfer preventive knowledge, update skills, and embed learning  
within firms.  
Germany presents the most stable profile in the comparison. In safety, it combines  
declining accident incidence, relative convergence toward the European benchmark and  
limited deterioration in severity indicators. Manufacturing remains more exposed than  
the total economy, but the overall pattern is comparatively contained. The training  
dimension is more mixed. Adult learning does not accelerate as strongly as in Spain or  
Romania, and Germany moves further below the European average by 2024. However,  
firms remain strongly engaged in continuing vocational training. The German case  
therefore reflects a trajectory based less on rapid expansion than on institutional stability,  
organizational continuity, and a comparatively controlled safety profile.  
Spain’s case follows a more ambivalent path. Accident incidence declines in both total  
economic activity and manufacturing, but manufacturing continues to record the highest  
incidence levels and the widest differential relative to the total economy. The severity  
indicators reinforce this weaker position, because falling incidence coexists with a less  
favorable internal structure of accidents. Training evolves more positively. Spain records  
a strong increase in adult learning and moves above the European benchmark in both total  
economic activity and manufacturing. Training provided by enterprises also remains  
above the European average, although the gap narrows by 2020. The Spanish case  
therefore combines a strengthening of adaptive capacity with persistent safety  
weaknesses in manufacturing.  
Romania presents the most dynamic and fragile national profile. Accident incidence  
remains far below the other cases throughout the comparison, but this favorable position  
is qualified by a sharper deterioration in severity indicators. Adult learning expands  
strongly after 2020. Romania narrows its gap with the European benchmark in total  
economic activity and moves above the European manufacturing average by 2024.  
However, this convergence is not reproduced in training provided by enterprises, where  
Romanian firms remain far below the European pattern. The Romanian case therefore  
combines rapid progress in individual learning participation with persistent weaknesses  
in firm training and accident severity.  
Viewed together, these trajectories show that the relationship between Industry 4.0. and  
the working environment is neither automatic nor uniform. Germany illustrates a more  
stable and institutionally grounded path. Spain reflects a mixed trajectory in which  
stronger training capacity coexists with persistent safety gaps. Romania shows a rapid but  
uneven process of convergence. What matters, then, is not only whether technology  
spreads, but whether safer production conditions and training mechanisms capable of  
sustaining adaptation over time accompany that process. Seen in this way, the working  
environment does not merely reflect industrial transformation. It also helps to shape its  
quality and durability.  
Conclusions.  
Improving the working environment belongs to the substance of Industry 4.0. rather than  
to its margins. The comparative evidence for the European Union, Germany, Spain, and  
Romania shows that industrial transformation cannot be reduced to gains in efficiency,  
productivity, or technological upgrading alone. Its quality is also shaped by the conditions  
under which work is organized, protected, and renewed. Occupational safety and training  
therefore form part of the core architecture of digital industrial change.  
18 | Sonia GARCÍA-MORENO, Víctor-Raúl LÓPEZ-RUIZ  
Digital Transformation and the Working Environment: International Evidence on Safety and Training  
in Industry 4.0.  
The results confirm a general decline in non-fatal accident incidence in both total  
economic activity and manufacturing. That pattern is consistent with an overall  
strengthening of occupational safety. However, this interpretation requires further  
qualification. Lower accident frequency does not systematically coincide with a more  
favorable severity profile. In several cases, longer absences gain relative weight and the  
burden reflected in lost days per case increases. The same unevenness appears in the  
training dimension. Participation in adult learning expands strongly in part of the sample,  
especially after 2020, whereas training provided by enterprises remains much more  
dependent on the structure and capacity of each national system. The working  
environment therefore evolves in differentiated ways rather than along a single, uniform  
path.  
The comparison across countries reinforces this point. Germany shows the most robust  
profile in the sample. It combines declining incidence, limited deterioration in severity and  
a strong base of continuing vocational training in firms. Spain records substantial  
progress, especially in adult learning, which marks an important strengthening of  
adaptive capacity. At the same time, Spanish manufacturing continues to display  
persistent weaknesses in its safety profile, with less favorable results in both incidence  
and severity. Romania presents the fastest expansion in adult learning participation and  
maintains low accident incidence levels, yet that more favorable picture is qualified by a  
sharper worsening in severity indicators and by a much weaker enterprise training base.  
These trajectories do not represent different speeds of a single transition. They reflect  
different configurations of industrial change, each with its own balance between  
protection, adaptation, and organizational support.  
The main contribution of the paper lies in reading occupational safety and training  
together within one comparative framework. By combining accident incidence, severity  
and training indicators, the paper shows that the quality of Industry 4.0. cannot be  
assessed only through technological upgrading, productivity or general labor market  
change. It also depends on whether safety improvement is sustained in qualitative terms  
and whether learning is embedded within firms. This joint approach offers a more  
complete assessment of the internal quality of digital industrial transformation.  
The working environment can therefore be understood as a practical test of the quality of  
digital transformation. A transition may appear favorable when accident incidence falls or  
adult learning expands, but that assessment remains incomplete if accident severity  
increases or if training is not embedded within firms. Industry 4.0. should therefore be  
assessed not only by the spread of digital technologies, but also by the extent to which  
those technologies are absorbed into safer and more adaptive work systems. The quality  
of industrial transformation depends on this alignment between technological change,  
preventive capacity and organizational learning.  
Given its descriptive and comparative design, the study does not estimate causal effects  
between digitalization and labor outcomes. Its contribution lies in clarifying how the  
working environment evolves alongside industrial transformation across countries and  
sectors. Within that scope, the evidence supports a central conclusion: occupational safety  
and training warrant a central place in the assessment of digital industrial change, not only  
as social concerns, but as conditions of resilience, balance and sustainable industrial  
development.  
Credit Author Statement.  
Sonia García-Moreno: Conceptualization, Methodology, Formal Analysis, Investigation,  
Data Curation, Writing - Original Draft, Visualization.  
Víctor-Raúl López-Ruiz: Validation, Supervision, Project Administration.  
Academy of Romanian Scientists | 19  
Journal of Knowledge Dynamics  
Vol. 3 (2026) No.1, pp. 5-20  
AI Declaration.  
The authors declare that no AI tools were used in any part of this study, including data  
handling, analysis, interpretation, or conclusion writing.  
References.  
Acemoglu, D., & Autor, D. (2011). Skills, tasks and technologies: Implications for  
employment and earnings. In O. Ashenfelter & D. Card (Eds.), Handbook of labor  
economics (Vol. 4B, pp. 1043-1171). Elsevier. https://doi.org/10.1016/S0169-  
7218(11)02410-5.  
Acemoglu, D., & Restrepo, P. (2020). Robots and jobs: Evidence from US labor markets.  
Journal  
https://doi.org/10.1086/705716.  
Appelbaum, E., Bailey, T., Berg, P., & Kalleberg, A. L. (2000). Manufacturing advantage: Why  
high-performance work systems pay off. Cornell University Press.  
https://doi.org/10.2307/2696195.  
of  
Political  
Economy,  
128(6),  
2188-2244.  
Arana-Landín, G., Laskurain-Iturbe, I., Iturrate, M., & Landeta-Manzano, B. (2023).  
Assessing the influence of industry 4.0 technologies on occupational health and  
safety. Heliyon, 9(3), e13720. https://doi.org/10.1016/j.heliyon.2023.e13720.  
Autor, D. H. (2015). Why are there still so many jobs? The history and future of workplace  
automation.  
Journal  
of  
Economic  
Perspectives,  
29(3),  
3-30.  
https://doi.org/10.1257/jep.29.3.3.  
Autor, D. H., Levy, F., & Murnane, R. J. (2003). The skill content of recent technological  
change: An empirical exploration. Quarterly Journal of Economics, 118(4), 1279-  
1333. https://doi.org/10.1162/003355303322552801.  
Badri, A., Boudreau-Trudel, B., & Souissi, A. S. (2018). Occupational health and safety in  
the Industry 4.0 era: A cause for major concern? Safety Science, 109, 403-411.  
https://doi.org/10.1016/j.ssci.2018.06.012.  
Bloom, N., Sadun, R., & Van Reenen, J. (2012). Americans do IT better: US multinationals  
and the productivity miracle. American Economic Review, 102(1), 167-201.  
https://doi.org/10.1257/aer.102.1.167.  
Bratianu, C., & Anagnoste, S. (2011). The role of transformational leadership in emergent  
economies. Management & Marketing. Challenges for the Knowledge Society, 6(2),  
319-326.  
Bratianu, C., Agapie, A., Orzea, I., & Agoston, S. (2011). Inter-generational learning  
dynamics in universities. Electronic Journal of Knowledge Management, 9(1), 10-  
18.  
Brunello, G., & Wruuck, P. (2020). Employer provided training in Europe: Determinants  
and obstacles. European Investment Bank Working Paper 2020/03.  
https://doi.org/10.2867/50660.  
Brynjolfsson, E., & Hitt, L. M. (2000). Beyond computation: Information technology,  
organisational transformation and business performance. Journal of Economic  
Perspectives, 14(4), 23-48. https://doi.org/10.1257/jep.14.4.23.  
Brynjolfsson, E., & McAfee, A. (2014). The second machine age: Work, progress, and  
prosperity in a time of brilliant technologies. W. W. Norton & Company.  
http://digamo.free.fr/brynmacafee.pdf.  
Cedefop. (2015). Job-related adult learning and continuing vocational training in Europe:  
A
statistical picture. Publications Office of the European Union.  
https://www.cedefop.europa.eu/en/publications/5548.  
Dul, J., & Neumann, W. P. (2009). Ergonomics contributions to company strategies. Applied  
Ergonomics, 40(4), 745-752. https://doi.org/10.1016/j.apergo.2008.07.001.  
Eurofound. (2021). The digital age: Implications of automation, digitisation and platforms  
for work and employment. Publications Office of the European Union.  
https://www.eurofound.europa.eu/en/publications/all/digital-age-  
implications-automation-digitisation-and-platforms-work-and.  
European Commission. (2021). Industry 5.0: Towards a sustainable, human-centric and  
resilient European industry. Directorate-General for Research and Innovation.  
20 | Sonia GARCÍA-MORENO, Víctor-Raúl LÓPEZ-RUIZ  
Digital Transformation and the Working Environment: International Evidence on Safety and Training  
in Industry 4.0.  
https://research-and-innovation.ec.europa.eu/knowledge-publications-tools-  
and-data/publications/all-publications/industry-50-towards-sustainable-  
human-centric-and-resilient-european-industry_en.  
García-Moreno, S., & López-Ruiz, V.-R. (2023). A review of the energy sector as a key factor  
in  
Industry  
4.0:  
The  
case  
of  
Spain.  
Energies,  
16(11),  
4446.  
https://doi.org/10.3390/en16114446.  
Goos, M., Manning, A., & Salomons, A. (2014). Explaining job polarization: Routine-biased  
technological change and offshoring. American Economic Review, 104(8), 2509-  
2526. https://doi.org/10.1257/aer.104.8.2509.  
Hollnagel, E. (2014). Safety-I and Safety-II: The past and future of safety management. CRC  
Press.  
Ichniowski, C., Shaw, K., & Prennushi, G. (1997). The effects of human resource  
management practices on productivity: A study of steel finishing lines. American  
Economic Review, 87(3), 291-313. http://www.jstor.org/stable/2951347.  
Kadir, B. A., & Broberg, O. (2021). Human-centered design of work systems in the  
transition to industry 4.0. Applied Ergonomics, 92, Article 103334.  
https://doi.org/10.1016/j.apergo.2020.103334.  
Karasek, R. A., & Theorell, T. (1990). Healthy work: Stress, productivity, and the  
reconstruction  
https://archive.org/details/healthyworkstres0000kara.  
Lasi, H., Fettke, P., Kemper, H.-G., Feld, T., & Hoffmann, M. (2014). Industry 4.0. Business &  
Information Systems Engineering, 6(4), 239-242.  
https://doi.org/10.1007/s12599-014-0334-4.  
of  
working  
life.  
Basic  
Books.  
OECD. (2019). OECD employment outlook 2019: The future of work. OECD Publishing.  
https://doi.org/10.1787/9ee00155-en.  
Rosen, P. (2022). Advanced robotics, AI, and automation. Definitions, uses, and OSH  
policies.  
European  
Agency  
for  
Safety  
and  
Health  
at  
Work.  
https://coilink.org/20.500.12592/zn874r.  
World Economic Forum. (2020). The future of jobs report 2020. World Economic Forum.  
https://www.weforum.org/publications/the-future-of-jobs-report-2020/.  
Zohar, D. (1980). Safety climate in industrial organizations: Theoretical and applied  
implications.  
Journal  
of  
Applied  
Psychology,  
65(1),  
96-102.  
https://doi.org/10.1037/0021-9010.65.1.96.