Automatic Control Time Magazine

The Hidden Technology

K. J. Åström
1. Introduction
2. A brief history
3. Where are we now?
4. Where is our future?
5. Conclusions
ECC August 31, 1999 Picture of Bosch System

Text The Hidden Technology

Electronic Stability Program (ESP) is a new • Pervasive
safety system which guides cars through wet • Very successful
or icy bends with more safety. ...
• Seldom talked about
The key is a yaw-rate sensor, which detects
vehicle movement around its vertical axis, Except when there is an accident!
and software which recognizes critical driv- Rare occasions!
ing conditions and responds accordingly. In • Why?
an instant, instructions are sent to the en- Easier to discuss devices than ideas
gine, transmission and brakes, thereby en- (feedback)
countering a skid at its onset. ...
We have not done our job!
&
c K. J. Åström ECC August 31, 1999 1
 Natural Science and
Engineering Science 1. Introduction
Many similarities but also many differences
2. A brief history
Natural PhenomenaTechnical Systems
3. Where are we now?
Analysis Synthesis
Isolation Interaction 4. Where is our future?
Fundamental Laws System Principles
5. Conclusions
Feedback is an a good system principle!

A Brief History Process Control

• Closely tied to emerging technologies Windmills 1787
(steam, power, electricity, telephone, ...)
Steam engines 1788
Process control
Governors 1890
Vehicle control
Water turbines 1893
Communication
Tolle Die Regelung der
• The servomechanism theory
Kraftmaschinen 1905
• Consequences One of our icons!
PID controllers 1930
• The second wave
• Maturity?
&
c K. J. Åström ECC August 31, 1999 2
 Wilbur Wright 1901
Men know how to construct airplanes.
Men also know how to build engines.
Inability to balance and steer still confronts
students of the flying problem.
When this one feature has been worked out,
the age of flying will have arrived, for
all other difficulties are of minor importance.

Flight Control The Robert E. Lee

• September 23, 1947 C-54
• Wright Brothers 1903
• Sperry A-12 autopilot
• Sperry’s 1912
• Bendix automatic throttle control
• V1 and V2 (A4)
• IBM punch card equipment for course
• Robert E Lee 1947
• No human touched the controls from start
• Sputnik 1957
until landing
• Apollo 1969
• Selection of radio station, course, speed,
• Mars Pathfinder 1997 flap setting, landing gear, final application
of wheel brakes.
&
c K. J. Åström ECC August 31, 1999 3
 Telecommunications Mervin Kelly on Black 1957
The repeater problem It is no exaggeration to say that without
Blacks invention 1927 Black’s invention (of the feedback amplifier),
Nyquist 1932 the present long-distance telephone and
Bode 1940 television networks which cover our entire
country and the transoceanic telephone
cables would not exist.

Who was Kelley?

Use of Feedback A Discipline Emerges

War pressures brought experiences in wide
• Patented and used in many areas fields together in
• Often revolutionary results
The Radiation Laboratory
• Control companies
The Instrumentation Laboratory
Woodward Governors
The Servomechanism Laboratory
Honeywell
and similar organizations in other countries.
Johnson Controls
Similarities were observed and a The Ser-
• Glimpses of generality vomechanism Theory was developed.

&
c K. J. Åström ECC August 31, 1999 4
Servomechanism Theory James Nichols Philips
• Foundations • Analysis simulation
Block diagram Stability
Complex variables Analog
Laplace transforms • Design
• System Concepts Loop shaping
Feedback Graphical
Feedforward • Implementation
Loop shaping Analog
Front page of JNP

Consequences The Second Wave

Driving forces Optimal Control
• New challenges Nonlinear Control
• New applications Stochastic Control
• Mathematics Computer Control
• Computers Robust Control
A new paradigm
• Education • Organization • State Space
System Identification

• Applications • Journals Rapid expansion Adaptive Control

• Subspecialities CACE
• Industrialization • Conferences

&
c K. J. Åström ECC August 31, 1999 5
 Current Status
1. Introduction • A well developed young field
Ideas and concepts
2. A brief history
Theory
3. Where are we now? Design methods
4. Where is our future? • Still developing rapidly
• Very wide application areas
5. Conclusions
• Some problems

Theory and Methodology System Theory

A very strong development with new ideas A large body of results with connections to
and concepts many branches of mathematics
• System Theory • Linear • Sampled
• Modeling and Identification • Nonlinear • Distributed
• Design • Stochastic • Discrete Event
• Simulation • Discrete • Hybrid
• Computer Aids Compaction badly needed to bring this
wealth of material to students!

&
c K. J. Åström ECC August 31, 1999 6
Modeling and Identification Design
• An essential feature of automatic control • Design problems are very rich
• Strong connection to physics Criteria
• White-, grey-, and black box modeling Disturbances
• A strong body of results with many con- Plant uncertainty
nections to statistics and data analysis • Difficult to formalize
• Modeling for control • A large number of specialized problems
• Linear and nonlinear models solved
• On line modeling and adaptation • New algorithms (LMI, IQC)

Autonomy and Learning Applications

• Adapt adjust to a specific use or situation
• Energy • Materials
• Autonomy being self governed
• Manufacturing • Instruments
• Learn to gain knowledge, comprehension
• Process control • Entertainment
or mastery through experience of study.
• Transportation • Biology
• Essential in many future systems
• Communication • Medicine
• Some progress in adaptive control and
learning theory but much remains • Structures • Economics

&
c K. J. Åström ECC August 31, 1999 7
 Paper Mill Steel Mill

Picture of paper mill Picture of steel mill

Robotics Automotive

Picture of car manufacturing Picture of car with systems

&
c K. J. Åström ECC August 31, 1999 8
 Aeroplane Space

Picture of aeroplane Picture of Challenger

Video Camera Scientific Instruments

• Feedback has had a profound impact
• Spectrometer Nier 1935
– Early use of feedback
– The 1984 Nobel Prize (van der Meer
student of C. J. D. M. Verhagen Delft)
• Adaptive optics

Picture of Video Camera

&
c K. J. Åström ECC August 31, 1999 9
 Adaptive Optics Economics
• Complex dynamical system with much
manual control
• H. Simon 1956:
Certainty Equivalence Control
• Program Trading - Relay feedback!
• Merton-Scholes
• Wall Street and The City
• The Financial Crisis

Biology Biology
Hoagland and Dodson: The Way Life Works A kind of feedback loop has shaped our evo-
Feedback is a central feature of life: lution. We change our environment, and it
All organisms share the ability to sense how becomes intellectually more challenging as
they are doing and to make changes in "mid- a result. This new environment selects peo-
flight" if necessary. ple who are best able to meet the challenges.
The process of feedback governs how we Some of these people make the environment
grow, respond to stress and challenge, and even more challenging, and over time this
regulate factors such as body temperature, feedback loop continues and accelerates.
blood pressure and cholesterol level. Christopher Wills, Professor of Biology at UCSD. Financial Times September 12, 1999: Throwing away a part of
our universe - In Kansas, all references to evolution has been wiped out form the school curriculum, Chrisopher
Wills rues the decision.

Feedback allow us to stand upright.

&
c K. J. Åström ECC August 31, 1999 10
 A Dilemma A Soul But No Body
Automatic control is a collection of ideas, + Generality
concepts and theories with a very wide ap- + Student attraction
plication area. Some problems are
+ Technology transfer
• Coupling to hardware
- No home court
• Coupling to industries
- No base industry
• Process knowledge
- Academic positioning
• Ownership of technology
• Academic positioning

A Perspective
1. Introduction Real World Automatic Control

2. A brief history Modeling

Processes Analysis
3. Where are we now? Systems Control Design
4. Where is our future? Implementation
5. Conclusions Commissioning
Operation

&
c K. J. Åström ECC August 31, 1999 11
 Changes in Technology Stability and "Controllability"
• Shift in fundamental limitations • Wright Brothers rejected the dogma that
From physical to organizational the aircraft should be inherently stable.
From performance to complexity • Minorsky 1922: It is an old adage that a
How fast can we make it? stable ship is difficult to steer.
Can we maintain it? • Integrated process and control design.
Time to market • The cardinal sin of control.
• The role of abstractions
• Tremendous opportunities for control

The Mercedes A Class Educational Challenges

• Theory and applications expanding
• Compaction of knowledge?
• The engineering aspect
• The field has changed a lot
• The introductory courses have not
• Use of feedback is often revolutionary

&
c K. J. Åström ECC August 31, 1999 12
 Recipe for Success
• Demanding problems 1. Introduction
• Ideas 2. A brief history
• Strong theoretical development
3. Where are we now?
• Good engineering
• Examples 4. Where is our future?
Servomechanisms 5. Conclusions
Feedback amplifiers vs process control
Geometric control theory

Conclusions Body and Soul

• An exciting field • Intellectual challenges (the soul)
• Use of feedback often revolutionary Basics that it generalizes easily
• Rapid growth of applications Give the general picture
• Many unsolved problems Make the introductory course more
• Streamline knowledge relevant and fun!

• Intellectual drivers • The engineering aspect (the body)

• Education Educate students broadly enough so

that they can take full systems respon-
sibility
&
c K. J. Åström ECC August 31, 1999 13