Content uploaded by Kenneth Yigael Goldberg
Author content
All content in this area was uploaded by Kenneth Yigael Goldberg on Jan 07, 2015
Content may be subject to copyright.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING, VOL. 9, NO. 1, JANUARY 2012 1
What Is Automation?
“WHAT is quality?” asked the narrator in Robert Pirsig’s
classic Zen and the Art of Motorcycle Maintenance.
In a similar spirit, we might ask: “What is Automation?” The
question is especially relevant to members of the IEEE Robotics
and Automation Society considering where to submit their next
journal paper. Should it go to the TRANSACTIONS ON ROBOTICS
(T-RO) or to the TRANSACTIONS ON AUTOMATION SCIENCE AND
ENGINEERING (T-ASE)?
In 1984, a group of visionary researchers arranged the
marriage of two subfields to form the IEEE Robotics and
Automation Society. Twenty years later the Society bifurcated
its journal into two publications. T-ASE was astutely guided
during its first five years by Editor-in-Chief Peter Luh and the
next five by Editor-in-Chief N. Vishu Viswanadham. As we
approach the Society’s thirty-year anniversary, we might step
back to consider the respective roles and characteristics of
Robotics and Automation.
Let’s admit: Robotics is sexier, more esoteric, more alluring.
Automation has always been viewed as the workhorse, focused
on manufacturing, less glamorous, but with a larger impact on
the world economy. One reason the term “Science” was added
to T-ASE was to emphasize the rigorous and theoretical aspects
of Automation. Despite their differences, the marriage has suc-
ceeded, and over time like most married couples, the partners
have acquired many mutual interests.
Automation has expanded beyond its roots in Manufacturing
to include applications in Healthcare, Security, Transportation,
Agriculture, Construction, Energy, and many other areas. Both
Robotics and Automation explore the frontiers of automated
and semi-automated machines. Both fields are increasingly con-
cerned with the role of humans and human interfaces, and with
the potential of the Internet and Cloud Computing.
So what is the difference between Robotics and Automation?
There are many possible distinctions. Here is the summary
from our Society’s Field of Interest Statement:
“Robotics focuses on systems incorporating sensors and
actuators that operate autonomously or semi-autonomously
in cooperation with humans. Robotics research emphasizes
intelligence and adaptability to cope with unstructured environ-
ments. Automation research emphasizes efficiency, productivity,
quality, and reliability, focusing on systems that operate au-
tonomously, often in structured environments over extended
periods, and on the explicit structuring of such environments.”
This statement emphasizes how Automation emphasizes
structured versus unstructured environments, reliability versus
adaptability, and efficiency versus exploratory operations.
These are valuable distinctions and I would like to propose
another one.
In my view, research in Robotics emphasizes Feasibility. Fea-
sibility focuses on proof-of-concept, demonstrating how a new
Digital Object Identifier 10.1109/TASE.2011.2178910
functionality can be achieved. Robotics papers usually demon-
strate a new ability of a robot, for example, demonstrating how
a robot can walk, drive, fly, or perform a surgical subtask.
On the other hand, research in Automation emphasizes
Quality. I use the uppercase to indicate Quality in the tech-
nical sense, as in Quality Control, which includes efficiency,
productivity, and reliability as stated in our Field of Interest
statement. Quality can be improved with new techniques, anal-
ysis, models, and results on robustness, stability, productivity,
efficiency, completeness, optimality, convergence, performance
guarantees, time complexity, sensitivity, verification, and relia-
bility. Of course an Automation paper may present a feasibility
study for an entirely novel mechanism, model, or theory for
applications that involve repetitive operations, for example, in
manufacturing or healthcare. But an Automation paper could
also focus on making robots walk, drive, fly, or perform a
surgical subtask more efficiently, more reliably, or more cost
effectively.
Let’s dispel the myth of the excluded middle: Robotics and
Automation are not disjoint. Feasibility and Quality are closely
related. Many papers include aspects of both subfields, but em-
phasize one or the other. Viewed this way, many researchers
study both Robotics and Automation.
I should clarify the distinction between uppercase Quality and
lowercase quality. Lowercase quality is related to value and as
Pirsig noted, is a subtle characteristic related to rigor and origi-
nality. A paper that emphasizes Feasibility can be high-quality
and a paper emphasizing Quality can be of low quality.
Almost all papers include elements of both Feasibility and
Quality: the distinction will never be binary; it is a matter of
degree. Both publications emphasize “research” over “develop-
ment.” In my view, a paper with significant results that primarily
emphasize Feasibility (i.e., the focus is on proof of concept)
should be submitted to T-RO, and a paper with significant re-
sults that primarily emphasize Quality (i.e., the focus is on per-
formance) should be submitted to T-ASE. Of course there will
be many exceptions.
The IEEE Robotics and Automation Society is a successful
marriage that has grown stronger over time. I believe it is vital
for the RAS community to take a fresh look at T-ASE and ex-
pand our definition of Automation, while also welcoming a new
community of researchers who focus on Automation. As I have
argued here, an important part of this self-reflection is to expand
and clarify our definition of Automation.
It is also important to consider other changes. For example,
reproducibility is a hallmark of Science. Everyone benefits
when researchers compare their new results alongside data or
reproduced experiments from previously published methods.
To facilitate this, T-ASE encourages authors to publish data,
code, CAD models, and other media with their papers, as well
as details on experimental methods, so that others can repeat
and extend published results. To increase access and impact,
1545-5955/$26.00 © 2011 IEEE
2 IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING, VOL. 9, NO. 1, JANUARY 2012
we also encourage authors to include presentation materials
and illustrative videos. See Info for Authors, Submission of
Multimedia Material, at our website. I will write more about
this issue in a future Editorial.
I’m looking forward to the next chapter of T-ASE. Please visit
our website for updates, links to the latest issues, and informa-
tion about past and upcoming Special Issues (for example, on
Green Manufacturing), new Editors and Associate Editors, and
our newly revised list of keywords and topic areas.
I’m convinced the RAS community will grow, thrive, and in-
crease our global impact by advancing both Robotics and Au-
tomation. Maybe someone will even figure out a way for robots
to maintain motorcycles.
Ps. I owe my thanks to the colleagues I consulted with on
this Editorial, including Tim Bretl, Peter Corke, Alessandro De
Luca, Seth Hutchinson, Vijay Kumar, Peter Luh, Kevin Lynch,
Matt Mason, Bruno Siciliano, Frank van der Stappen, Dick Volz,
and many others. All blame for mistakes and omissions should
be attributed to me.
KEN GOLDBERG, Editor-in-Chief
IEEE TRANSACTIONS ON AUTOMATION SCIENCE
AND ENGINEERING
University of California at Berkeley
College of Engineering and School of Information
Berkeley, CA 94720-1758 USA
goldberg@berkeley.edu
