Content uploaded by Quang Cuong Pham
Author content
All content in this area was uploaded by Quang Cuong Pham on Sep 04, 2018
Content may be subject to copyright.
Ethical, lEgal, and SociEtal iSSuES
126 •
IEEE ROBOTICS & AUTOMATION MAGAZINE
•
jUNE 2018
The Impact of Robotics and Automation on
Working Conditions and Employment
By Q.-C. Pham, R. Madhavan, L. Righetti, W. Smart, and R. Chatila
s roboticists, we like to think
that the fruits of our research—
robots that are faster, more
efficient, more agile, and
more intelligent—can only benefit
humanity. While this is certainly true
for exploratory or disaster intervention
robots, the case is more controversial
for other types of robots such as those
used for military purposes, as discussed
in the previous article in the series [7].
In this article, we provide a quick over -
view of the concerns raised by the
accelerated introduction of robotics
and artificial intelligence (AI) tech -
nologies in all economic sectors and, in
particular, its effects on working con -
ditions and employment.
Robotics and Automation in the
Workplace
Robots, like any machines introduced
into the production process, have con-
trasting effects on workers. On the one
hand, they can eliminate some harsh,
unhealthy, or dangerous tasks. Consider
for instance, the welding process in car
manufacturing. Welding is certainly a
hazardous activity for workers to per-
form, with deleterious short- and long-
term effects ranging from irritations of
the eyes, nose, ears, throat, and chest to
pulmonary infections, heart diseases,
and lung and throat cancers. The robot-
based automation of welding in modern
car manufacturing lines has significant-
ly reduced health problems caused by
welding. On the other hand, precisely
because robots can automatically per-
form some tasks, they render the work-
ers who previously performed those
tasks “redundant” for production pro-
cesses. This has multiple adverse effects
for workers.
For example, workers rendered
redundant by robots face the risk of
being laid off. Since the first machines
were introduced on a large scale at the
beginning of the 19th century (the First
Industrial Revolution), the layoff of
redundant workers has been a common
practice. An early and particularly tragic
episode was the introduction of power
looms in the United Kingdom during
the first few decades of the 19th century.
Skilled weavers were suddenly put in
competition with machines that could
weave better and faster. Facing wage
reduction or replacement by machines
operated by cheaper, unskilled work-
ers, desperate weavers (later known
as Luddites) waged a campaign of
destruction targeted at the newly intro-
duced machines. The response by the
owner class was harsh: seventeen Lud-
dites were hanged, many others were
imprisoned, and the movement was
quickly dispelled.
These days, even though strict labor
regulations and strong workers’ organi-
zations in most developed countries
may offer some protection against or
procure compensation in the event of
layoffs, such technological layoffs and
their adverse effects on the lives of the
concerned workers seem inevitable.
Indeed, when the management of a
company considers introducing robots,
its chief concern is not whether the
robots are based on a fancy new tech-
nology or whether they will improve
workers’ welfare; it is profitability. In this
view, keeping redundant workers simply
does not make economic sense.
Additionally, workers who retain
their jobs alongside robots might not
always see their working conditions
improve. Consider, for instance, the
Amazon warehouses into which robots
have been introduced on a massive
scale over the past few years; because
the robots are so fast and so consistent,
their pace can be set arbitrarily and is,
in fact, imposed on the workers. A jour-
nalist working undercover in an Ama-
zon warehouse testifies:
Alone in a locked metal cage, ten
feet from my nearest colleague,
The IEEE Robotics and Automation Research and Practice Ethics Committee
(RARPEC) is intended as a platform to exchange ideas and discuss the impacts and
practice of robotics and automation (R&A) technologies in research, development,
and deployment that appear to pose ethical questions for humanity. With increased
awareness and controversies surrounding R&A, RARPEC is publishing a series of
opinion pieces that will focus on separating hype from reality by providing an
objective and balanced treatment of technological, ethical, legal, and societal
perspectives. Second in the series, this piece focuses on topics related to robots,
jobs, and workforce development. Please send your feedback and suggestions to
the chair of the committee, Raj Madhavan, at raj.madhavan@ieee.org. We look
forward to your comments!
Digital Object Identifier 10.1109/MRA.2018.2822058
Date of pub lication: 13 June 2 018
A
127
june 2018
•
Ieee ROBOTICS & AuTOMATIOn MAGAZIne
•
a robot approaches from the
shadows and thrusts a tower of
shelves toward me. I have nine sec-
onds to grab and process an item to
be sent for packing, a target of 300
items an hour, for hour after relent-
less hour. As I bend to the floor
then reach high above my head to
fulfill a never-ending stream of
orders, my body screams at me [8].
Far from the image of robots serving
humans, the reality is, in fact, the other
way round: “…(human) staff are just
cattle, there to serve robots” [8].
But would the increased work inten-
sity be compensated for by higher sala-
ries or shorter working hours? In fact, a
detailed study of the effects of robot
densification in 14 industries across 17
developed countries during the period
of 1993–2007 shows that low- and
middle-skilled workers actually suffered
salary reduction with the introduction
of robots, as illustrated in Table 7 in [5].
The same study shows that there was no
significant reduction in the number of
working hours.
Global Effects of Robotics
and Automation: Toward a
Jobless Society?
As discussed in the “Robotics and Auto-
mation in the Workplace” section, the
impact of robotics and automation on
the welfare of individual workers is far
from entirely positive, but what are its
long-term effects on all of society, par-
ticularly with regard to employment?
Interestingly, only a few decades
after the Luddite revolt, the perspective
of entirely automatic production, with-
out any human intervention, started to
be formulated. Andrew Ure, an early
business theorist, thus contemplated
“the most perfect manufacture […]
which dispenses entirely with manual
labor” [9]. That perspective has not,
however, materialized. As more tasks
became automated, an even larger num-
ber of new tasks, made necessary by
new products or entirely new economic
sectors, was created that required hu -
man lab or.
Yet, due to the rapid progress of
robotics and AI technologies in the past
few years, the perspective of a jobless
society, in which all work is performed
by robots and no jobs are left for
humans, has begun to capture consider-
able attention from the general public.
Alarmist articles about a jobless future
abound in the mainstream media, based
significantly on scholarly literature. For
instance, a widely cited report by Oxford
economists predicts that up to 47% of
total U.S. employment is at risk of being
taken away by automation [4]. In a
recent and well-documented book, tech-
nologist Martin Ford argues that, con-
trary to the development of automation
up until now, automation today, because
of its cognitive capability, carries an
actual threat of massive job destruction
over the coming decades [3]. However,
there are also studies that make much
less dramatic predictions. In fact, as
highlighted in a recent MIT Technology
Review survey, there is no consensus
among economists and technologists
about the degree and timeline of job
eliminations resulting from automation
[10]. Furthermore, the effects of robotics
and AI on the norms of work and em -
ployment, and the associated concerns
in developing economies (the so-called
global south), are even less well under-
stood because their societal acceptance
and assimilation differ significantly
between developed and developing
economies. In labor-intensive econo-
mies (for example, the BRICS countries,
i.e., Brazil, Russia, India, China, and
South Africa), the effects of automation
would be felt much more steeply in the
coming decade. While labor may still be
cheap in developing economies, auto-
mation in developed countries will offset
this advantage, thereby possibly result-
ing in significant adverse effects on
workforces in developing countries.
The number of robots in factories
has been rising quickly, and robotics
technologies have been introduced into
many sectors beyond manufacturing,
e.g., surgical or rehabilitation robots in
hospitals, service robots, self-driving
cars, and so on. However, from a roboti-
cists’ perspective, there is still a very
long way to go before robots can totally
replace humans. For example, outside of
the structured environments of factory
assembly lines, robot locomotion and
manipulation capabilities are still very
limited. During the 2015 Defense Ad -
vanced Research Projects Agency
Robotics Challenge, robots (teleoperat-
ed by humans and so not even autono-
mous!) from the best research labs
around the world had trouble perform-
ing tasks that most humans would find
trivial. Even the simple task of grasping
and manipulating a previously un -
known object in-hand is still the subject
of intense academic research. Moreover,
the robots already deployed in factories
still require an enormous amount of
reprogramming when facing a slightly
different task. They are far from being
able to automatically learn to perform
new tasks by themselves or from hu -
man demonstration.
Finally, the discussion of automation
and employment should not be cen-
tered only on the number of jobs lost; it
should also deal with the changing
nature of work because of the automat-
ability and functional description of
tasks. In the Fourth Industrial Revolu-
tion, the emphasis is on how machines
and humans can work together so that
repetitive and dangerous tasks can be
relegated to machines and automated
systems. This augmented collaborative
workforce is the wave of the future and
has enormous implications for employ-
ment in the automation age. It will rede-
fine the relations between workers, their
crafts, and their working environments.
On the one hand, workers can focus on
aspects that require creativity, social
skills, and emotional intelligence; on the
other, this could also have a dehumaniz-
ing effect if workers’ activities are subju-
gated to robots’ behaviors.
Proposed Solutions to
Address Unemployment
Caused by Automation
Although the degree and timeline of job
eliminations caused by automation are
still debated, there is a consensus that, in
the present global context of stagnant and
interdependent economies, automation
will inevitably take away a significant
number of jobs. This means that, in the
next few years and decades, many work-
ers will lose their jobs to robots, while
those keeping their jobs will experience
128 •
IEEE ROBOTICS & AUTOMATION MAGAZINE
•
jUNE 2018
increased physical and psychological
pressure and still more will face unem-
ployment due to the lack of jobs. A num-
ber of solutions have been proposed to
address these problems.
An important consideration is to
raise the level of workers’ education
(both initially and continuing) so that
they can undertake the higher-level jobs
required by automation. Training pro-
grams to develop new, requisite skill sets
available across the spectrum of the
workforce, and not just for low-skilled
workers, could be mandated. Such pro-
grams could be funded by public–pri-
vate partnerships and made available
for workers who are still employed and
those who are in between jobs.
Universal basic income (UBI) is
another concept proposed to address
technological unemployment, with all
of a country’s citizens or residents un -
conditionally receiving sufficient regu-
lar amounts of money that will enable
them to live. Additionally, there would
be no requirement for people to work
or look for work. There are many ver-
sions of UBI, differing widely in terms
of the proposed income amount and
the funding source. In any case, for such
a system to provide decent living condi-
tions for everyone in a country (and,
beyond, in every country), the amount
of funding required is likely to be very
significant. As a result, there is a signifi-
cant and complex debate about how UBI
could be funded, whether such a system
could be sustainable at all, and the effects
it would have on the economy.
The notion of robot taxes has been
proposed as another alternative to deal
with the potential unemployment creat-
ed by automation. The basic idea, as
suggested by Bill Gates [1], is to tax cor-
porations and entities deploying robots
that cause job losses. The tax income
could then be used to offset the eco-
nomic hardships experienced by laid-off
workers or retrain them so that they can
be reassimilated into the workforce. In
that vein, a motion (eventually rejected)
in the European Union Parliament in
2017 proposed “levying tax on the work
performed by a robot or a fee for using
and maintaining a robot should be
examined in the context of funding the
support and retraining of unemployed
workers whose jobs have been reduced
or eliminated” [2]. Robot taxes have cer-
tainly met criticisms from a number of
economists. For instance, Larry Sum-
mers [6] argued that there are no funda-
mental differences between robots and
any technologies that may cause job
losses (including Bill Gates’s software);
yet there are no specific taxes on such
technologies. Thus, taxing robots would
amount to another tax on capital, which
most capitalists would oppose.
More generally, socioeconomic, politi-
cal, and resource constraints should be
carefully considered when emerging
technologies are deployed because there
is a potential for unintended consequenc-
es such as tilting economic and power
structures to unduly benefit certain seg-
ments of society, resulting in new gaps
and/or exacerbating existing inequities.
There are time-sensitive challenges
regarding how developing nations, with
their potentially low-technology class-
room-centric curricula, can be provided
with the technical expertise that would
allow for the introduction and absorption
of these cutting-edge technologies.
Robotics and automation carries the
wonderful promise of liberating
humanity from toil. In an ideal society,
most of the repetitive, unhealthy, and
uninteresting work would be fulfilled by
robots, while humans would spend a
limited amount of time every day on
work (including deciding what the
robots should do) and the rest of the
time on creative activities. From a tech-
nical viewpoint, this future is certainly
possible, yet both the current situation
and the outlook pictured by many
reports are gloomy. Robots now tend to
be perceived by a portion of the general
public as a threat, instead of as a fantas-
tic liberation tool. Why is this so?
In the current economic system
where robots are owned by a minority,
the gains in productivity they permit
(e.g., higher wages and fewer working
hours) are not likely to be shared by the
working majority; rather, robots would
be seen as the reason for humans’ job
losses. Therefore, to reach the ideal
society that most robotics researchers
have in mind, the notion of who owns
the robots, the working majority or a
minority of capitalists, might just be the
decisive question.
References
[1] K. J. Dela ne y. (2017, Feb. 17). The robot
that takes your job should pay taxes says Bill
Gates. Quartz. [Online]. Available: https://qz
.com/911968/bill-gates-the-robot-that-ta kes-
your-job-should-pay-taxes/
[2] European Parliament. (2018, May 16).
Report to European Pa rliament resolution
with recommendations to the Commission
on Civil Law Rules on Robotics. [Online].
Available: https ://tiny url.com /EPre port2017
[3] M. Ford, Rise of the Robots: Technology and
the Threat of a Jobless Future. Ne w Yor k : Basic
Books, 2015.
[4] C. Frey and M. Osborne, The Future Of
Employment: How Susceptible Are Jobs To Com-
puterisation? Oxford, UK: Oxford Univ. Press,
2013.
[5] G. Graetz and G. Michaels, “Robots at
work,” Centre for Economic Policy Res., Lon-
don, UK, Tech. Rep. DP10477, 2015.
[6] S. Kessler. (2017, Mar. 6). Lawrence Sum-
mers says Bill Gates’ idea for a robot tax is
“profoundly misguided.” Quartz. [Online].
Available: https://qz.com /925412/lawrence-
summers-says-bi ll-gates-idea-for-a-robot-ta x-
is-profoundly-misguided/
[7] L. R ighet ti, Q.-C. Pham, R. Madhava n, and
R. Chati la, “Letha l autonomous weapon sys-
tems,” IEEE Robot. Autom. Mag., vol. 25, no. 1,
pp. 123–126, 2018.
[8] A. Selby. (2 017, Nov. 27). Undercover at
Amazon: Exhausted huma ns are inefficient
so robots are taking over. Mirror Online.
[Online]. Available: https://w ww.mi rr or.co.uk/
news/uk-news/undercover-amazon-exhausted-
humans-inefficient-11593145
[9] A. Ure, The Philosophy of Manufactures, or
an Exposition of the Scientific, Moral and
Commercial Economy of the Factory System of
Great Britain. London: Charles Knight, 1835.
[10] E. Wi nick. (2018, Jan. 25). Every study we
could find on what automation will do to
jobs, in one chart. MIT Technology Review.
Available: https://w ww.tec hnology review
.com/s/610005/every-study-we-could-find-
on-what-automation-will-do-to-jobs-in-one-
chart/
