IEEE Robotics & Automation Magazine - December 2018 - 117

Ethical, Legal, and Societal Issues

Robotics
Novel
Techniques
in Servicetoand
Solve
Space-Exploration
the
Responsibility to
Problems
Society
By Ruzena
Name name,
Bajcsyaffliation;
Name name, affliation;

I

n the autumn of 2000, I came to the
University of California, Berkeley,
(UC Berkeley) to become the founding director of the Center for Information Technology Research in the
Interest of Society (CITRIS). I adopted
the goals defined by many colleagues at
UC Berkeley before me and made them
very personal: on the one hand, these
goals focused on connecting with the
non-information technology (IT) users
in academia; on the other hand, they
were to facilitate connections between
industry and academia so that the
needs of industry could stimulate
research and academic research could
find its way into industry and society at
large. I am proud to say that these goals
have been accomplished and continue
to be furthered.
For example, at UC Berkeley, we
began the Digital Humanities and Digital Archeology working groups, supported the Digital Laboratory in the
Department of Economics, and participated in many more modest activities
on campus. We strengthened previously established connections with industry leaders such as Hewlett-Packard,
Intel, and Microsoft and we also forged
new connections.
Now, in 2018, I see a parallel need
for robotics at the intersection of cyberphysical systems and IT, i.e., the need to
discover opportunities and challenges
to be met for this field to serve society.
Just as with the initial CITRIS goals, two
aspects must be considered in terms of
Digital Object Identifier 10.1109/MRA.2018.2873029
Date of publication: 12 December 2018

the connection between robotics and
society. Both are suffused with ethical
issues: 1) robotics in service to the academic community and its responsibility
to society and 2) robotics in service to
industry and its responsibility to society.
Robotics in Service to the
Academic Community
Within society, academia has two profound functions: 1) teaching the next
generation and 2) pursuing research,
both fundamental and applied. As we
have seen during the last ten years,
many specialized robotics programs at
the undergraduate, graduate, and doctorate levels have been established.
When I began my academic career
in 1967 at Stanford, robotics was not
taught explicitly; there were courses relevant to robotics in the mechanical
engineering (solid mechanics), electrical
engineering (control theory, signal processing), mathematics (topology, combinatorics, graph theory), computer
science (programming, logic, automata
theory, algorithms, numerical methods),
and other departments. Research in
robotics was innovative, creative, and
basic. At the experimental level, however, what we worked with was rather
primitive compared with today's laboratory equipment.
Students and faculty had to build
their own devices and interfaces and
perform their own programming, often
in assembly language, to have access to
the raw signal coming from either the
sensor/digitizer or the input to the actuator. The advantage of this mode of working was that students and researchers

understood the limitations and boundaries of the devices they were using. The
facts about signal-to-noise ratio and the
limits of spatiotemporal sampling were
explicitly taken into account in the
control algorithms. There was a clear
transparency in what to expect during
a demonstration.
To d ay, t h e
positive effect of When I began my
having modules
academic career in
and interfaces
from the sensor, 1967 at Stanford,
processor, and
robotics was not
controller to the
actuator hides taught explicitly.
very-low-level
signals, allowing
many users, including students, to enter
the field without an initial struggle to
understand low-level properties. However, we have to answer social service
and responsibility questions about how
much and from whom we should conceal and to whom we need to reveal
low-level properties, including limits on
performance.
We need to ask two questions about
robotics education:
1) How much basic knowledge of physics, math/statistics, chemistry, and
the like should we be teaching?
2) How much new knowledge should
we be teaching based on new technologies (such as analytics in big
data) or what the current job market
is asking for?
In an article [1] discussing what we
should teach our students so they can
survive the next 20 years, I argue that the
basic laws of nature, such as Newtonian

december 2018

*

IEEE ROBOTICS & AUTOMATION MAGAZINE

*

117
IEEE Robotics & Automation Magazine - December 2018

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - December 2018
