International Federation of

1

• International Federation of
• Automatic Control
• EMERGING AREAS PROJECT
• Rotterdam August, 2003
• The goal of the on-going Emerging Areas project
  is to identify recent significant trends that are
  likely to continue within the automatic control
  field. The CC Chairs Guests presented
  preliminary 2003 candidate suggestions during the
  Workshop phase. Presentations from the Workshop
  follow,
• Introduction to Emerging Areas Workshop Slide
  2
• CC Presentations
• 1 Systems Signals (Theory) Slide 6
• 2. Design Methods (Theory) Slide 12
• TB Vice Chair (Theory) Slide 17
• 3. Computers, Cognition Communications
  (Implementation) Slide 21
• 4. Mechatronics, Robotics Components
  (Implementation) Slide 25
• 5. Manufacturing Systems (Applications) Slide
  32
• 6. Industrial Systems (Applications) Slide 37
• 7. Transportation and Vehicle Systems
  (Applications) Slide 53
• 8. Bio Ecological Systems (Applications) Slide
  73

2
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• INTRODUCTION TO WORKSHOP
• Mike Masten

3
IFAC EMERGING AREAS

• WORKSHOP / PANEL SESSION
• AGENDA
• Workshop (Afternoon)
• Consideration Potential Significant
  Trends/Forecasts
• Selection Most Significant Trends/Forecasts
• Preparation Breakouts to Prepare for Panel
• Panel Session (Evening)
• Presentation Workshop Conclusions
• Evaluation Industrial Guest Feedback
• Discussion Audience Questions Comments

4
IFAC EMERGING AREAS

• WORKSHOP CONSIDERATION
• Brain Storming Presentation of Trends and
  Forecasts of Potential Control Automation
  Developments
• Brief ( 5 Minutes) Presentations
• (1) Identify likely Forecasts/Trends
• (2) Explain Forecasts/Trends
• (3) Brief Rationale for Why Forecasts/Trends are
  Likely to Happen
• Tabulation of Forecasts/Trends

5
IFAC EMERGING AREAS

• WORKSHOP - SELECTION
• Select Reduced Number Those Believed Most
  Significant
• Vote from Tabulation of Trends/Forecasts
• WORKSHOP - PREPARATION
• Break-Out Groups to Prepare for Panel Session
• Select Representative for Panel Session
• Discuss Forecast/Trends in More Detail
• Prepare for Panel Session

6
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• CC 1 SYSTEMS SIGNALS
• (THEORY)
• Tohru Katayama

7
Potential Significant Trends in Systems
Signals

• Tohru Katayama
• CC Chair
• Kyoto University, Kyoto, Japan

8
TC1.1 Modeling, Identification Signal
Processing

• These are currently important areas within
  control technology, and will continue to be
  important for the foreseeable future.
• SYSID has a long history Prague (1st, 2nd),
  Delft (3rd)
• Areas of significant development include
• 1) Subspace identification multivariable
  systems
• 2) Nonlinear system identification Hammerstein
  and Wiener models
• 3) Identification of closed-loop systems
• 4) Channel identification in time-varying
  environments

9
TC1.2Adaptive and Learning Systems

• Theory of adaptive control is approaching some
  maturity but a major new development will be
  made associated with adaptive controls of
  nonlinear systems.
• Increasing applications of adaptive control in
  many areas. e.g. stability analysis of neural
  networks nonlinear mechanical systems,
  mechatronics application to mobile
  communications
• But, practical solutions do not necessarily use
  the word Adaptive to describe their
  applications.
• ? Hidden Technology

10
TC1.3 Discrete Event and Hybrid Systems

• Event-driven and time-driven system theory is now
  converging toward Hybrid system.
• ? Developments are occurring in both theory
  and applications (automotive, military, etc.).
• Multi-disciplined field with contributions from
  different areas, including computer science,
  control theory, chemical engineering, etc. e.g.
  ADHS2003
• Many closed-loop systems include feedback loops
  with event information and/or event-based
  actuation signals.
• More are expected in the future.

11
TC1.4 Stochastic Systems

• New robust optimization methods are developing.
• 1) Randomized methods within robust control
  using average approach is still popular.
• 2) Solutions to worst case problems with small
  chance that constraints are not satisfied are
  emerging.
• New (non traditional) applications include
• 1) Computer network congestion (www, etc)
• 2) Hybrid systems modeling
• 3) Modeling of complex systems (air traffic,
  etc) to predict situations of interest
  (conflicts, etc)

12
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• CC 2 DESIGN METHODS
• (THEORY)
• Ruth Bars

13
FORECASTS/TRENDS  CC ON DESIGN METHODS 1.
Nonlinear Controller Design methods The best
controller would be the inverse model. But
generally the model is not invertible. Also
constraints, model uncertainties have to be
considered. Designing Suboptimal Inverse Models
is in the front of interest.
14
CC ON DESIGN METHODS Continued 2. In order to
solve the class of stochastic control problems
for which the certainty equivalence principle is
not valid a new integrated design approach has to
be developed. This objective can be reached only
by a joint effort of experts in estimation and
optimal control theories. This problem arises
e.g. in guidance problems. ( A message of
plenary lecture by J. Shinar at the CAO2003
Workshop)   3. New effective real-time optimal
algorithms are needed for 2D and 3D pattern
recognition in case of more complex sensing and
signal processing used e.g. for control of moving
objects. Analytical and computational methods
have to be used together.  
15
CC ON DESIGN METHODS Continued 4. Data handling
is also an important question. Building models
incorporating both physical knowledge and a
priori knowledge. Giving techniques transforming
raw data into useful information. Developing
improved measurement methods including
inferential estimation. Data based predictive
control.   5. Design of switching control
systems. Performance properties of complex
dynamical systems can be improved using switching
with logic-based control. E.g. discrete event
driven systems, constraints could be handled by
switching systems.
16
CC ON DESIGN METHODS Continued 6. Robust
control of large scale systems. Control of
networks, navigating packages from sources to
destinations on a very large scale heterogeneous
communication network (such as the Internet, web
applications) with minimum loss, high efficiency
and with decisions made by a large number of
users in a distributed fashion. (from Panel
Discussion Whitherto Robust Control,
ROCOND2003, Milano) 7. Other topics
Biotechnology, biosensors in control systems
Nanotechnologies Virtual laboratories via the
WEB
17
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• TB VICE CHAIR (THEORY)
• Alberto Isidori

18
Control under communication constraints
Problem to control of a distributed system
consisting of a large number components of very
different nature -- such as analog devices,
sensors, computers, decision logics -- which
exchange information through wireless
networks. While control, information theory and
communication are to some extent mature
disciplines, little effort has been put so far in
understanding how issues in information theory
and communication constraints affect the
performance of a distributed control system.
The presence of (substantial) communication
constraints may induce a change in control design
principles. Above all if a rigorous analysis of
robustness, and guaranteed robustness margins,
are sought.
19
Control under communication constraints

• The goal is to design sensors, encoders,
  communication channels and controllers (or
  estimators) so as to achieve prescribed
  performances despite of all the constraints and
  obstacles imposed by the communication channels
  and in the presence of possible uncertainties and
  disturbances.
• Constraints imposed by the communication channels
  include
• bandwidth
• delays (of variable amount)
• quantization errors
• transmission noise
• loss of information

20
Additional research issues in control theory

• Design methods for generalized tracking/pursuit
  problems the development of a notion of
  steady-state behavior for nonlinear systems, the
  development of an adaptive and non-equilibrium
  theory of internal-model-based control.
• The study of tracking/rejection problems in
  cases in which the exosystem (the
  command/disturbance generator) is a hybrid system
  including continuous-time and discrete-event
  systems the design of systems including hybrid
  internal models.

21
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• CC 3 COMPUTERS, COGNITION, AND COMMUNICATIONS
• (IMPLEMENTATION)
• Robert Babuska (for Wolfgang Halang)

22

• FORECASTS/TRENDS 
• CC ON COMPUTERS, COGNITION, COMMUNICATIONS
• Web technologies, COTS, SOUP, etc. entering
  controlOstensibly to save development costs up
  front, and to shorten the time to market, more
  and more web technologies, commercial
  off-the-shelf (COTS) components, and software of
  unknown pedigree (SOUP) are used in industrial
  control systems. Often, operating systems and
  application programs employed originate in the
  office world. As these components lack both the
  reliability and real time capabilities required
  by industrial control systems, we are witnessing
  a technological setback raising severe security
  and safety problems. Owing to the complexity and
  unreliability of these artifacts, it is doubtful
  whether in the end there will be any cost savings
  at all, because in the overall life-cycle of a
  software system about 80 of the development
  costs are related to the test phase and further
  to the maintenance of the system. Moreover, in
  safety related systems, the test phase assumes a
  critical role, since one cannot postpone
  verification and elimination of software errors
  to a later maintenance phase.

23

• CC ON COMPUTERS, COGNITION, COMMUNICATIONS
  Continued
• Computing with wordsComputing with words is
  becoming increasingly relevant especially in Web
  mining, text recognition, semantics, etc. So far,
  links and relevance to control are not visible,
  but one can expect that it will become an
  important item in the future. Possible
  applications may be in areas where collaboration
  of (semi-)autonomous controlled systems (robots)
  with humans is essential, such as collaborative
  decision making.
• Optical and self-organizing networks Wireless
  networks, machine-to-machine communication
  Optical bandwidth grows faster than processing
  power. Thus, there will be a paradigm shift from
  processor orientation to a network-centered view,
  resulting in different architectures of computer
  control systems. Application areas for wireless
  machine-to-machine communication are traffic
  guidance and toll collection systems, machine
  control, logistics, surveillance and alarm
  systems, environmental data acquisition, facility
  automation, remote control, consumer devices and
  tools for the handicapped.

24

• CC ON COMPUTERS, COGNITION, COMMUNICATIONS
  Continued
• Tele-presence (haptic control, force feedback
  control)Supported by the internet techniques
  remotely controlled systems become, especially in
  the area of robotics, more and more important. In
  order to have a "real" feeling of "being at" the
  remote place, haptic interfaces requiring force
  feedback control are necessary. Depending on
  limited bandwidth of the communication medium
  reasonable andvarying time delays appear, which
  prevent at least a "well feeling" of the human
  operator and can lead to command sequences
  resulting in unstable systems at the remote
  place. Therefore, distributed control concepts
  including control with variable delay times are
  pressing research areas.
• Learning control systemsAlthough quite an old
  issue, there is some increased interest in
  learning systems, both in academia and from
  industry. Learning includes also self-tuning and
  automatic commissioning.

25
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• CC 4 MECHATRONICS, ROBOTICS, AND COMPONENTS
• (IMPLEMENTATION)
• Anibal Ollero

26
.Mechatronics, Robotics and Components."Technolo
gies for control and control to improve
technologies"

• Aníbal Ollero

Escuela Superior de Ingenieros - Universidad de
Sevilla email aollero_at_cartuja.us.es
http//www.esi.us.es/GVR
27
     
Embedded controllers
Control System
Components and Instruments. Human Machine
Systems. Cost Oriented Automation.
Mechatronic Systems.
Robotics.
28
Forecast/trends in control technology

• Development of innovative sensors and actuators.
• Benefit from the progress of perception systems.
• MEMS developments.
• Distributed control technology communications
• Wireless communications in the loop.
• Computer developments embedded systems
• Software agents to support automation.

29
Forecast/trends in control applications

• Medical technologies surgical devices, medical
  instruments.
• Communication technologies network control,
  wireless devices.
• Autonomous systems technology (UAV, UGV, AUVs,
  multi-).
• Industrial robotics and automation technology.
• New cars, vehicles and transportation
  technologies.
• Biotechnology.
• Technologies for safety critical and hostile
  environments
• space, disaster remediation, defence,
• Technology for energy savings
• mixed energy sources, small scale energy
  systems.
• Embedded control systems for vehicles, consumer
  products, engines, instruments, MEMS.

Performance, reliability, cost, easy to use and
maintain.
30
Most significant technological trends

• Two poles
• Integration.
• Embedded control systems integrating perception
  and control functions that can be used,
  eventually, in a transparent way.
• Reasons Technological developments (Hardware
  integration, MEMS, ) and new applications
  (vehicles, autonomous systems, consumer products,
  biomedical systems, ).
• Distribution.
• Distributed control systems with wire and
  wireless connections between components and
  embedded controllers.
• Reasons Communications technology and new
  applications (Tele- applications, distributed
  manufacturing, protection of people and
  environment, home automation, .).

31
The single trend/forecast

• Integration of control and perception components
  in embedded systems that could be networked using
  wire or wireless technologies.

32
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• CC 5 MANUFACTURING SYSTEMS
• (APPLICATIONS)
• Shimon Nof

33
Emerging Trends in Automation and
Control Manufacturing and Logistics
Systems Collaboration, e-Work, and
Nano-Systems Shimon Y. Nof Purdue
university USA IFAC TB Meeting August
2003 Rotterdam, Holland
34
Collaborative Robotics and e-Work -- Trends
Overall Trend Smart robotic teams (normal,
micro, nano robots) will be able to interact even
better than human teams Trend 1. Collaborative
Coordination Control Theory Why? Optimized
coordination of team interactions is key to
mfg. effectiveness competitiveness Trend 2.
Control Methods to Manage errors, conflicts, and
interactions Why? Complexity, redundancy
among robots teams increases in mfg. and
service systems Trend 3. Control Protocols for
fault-tolerant, time-out integration Why?
Future collaborative robotics will depend on
cheaper, redundant arrays/networks (e.g.,
FTTP)
35
Collaborative e-Work / e-Mfg. -- Trends
Overall Trend Smart tools and collaboration will
enable significantly better innovation,
productivity and quality Trend 1.
Interoperability of enterprise applications
software Why? Move to integrate enterprise
modeling, architecture platform, and
ontology TC5.3 / D. Chen Trend 2. Control
Methods to Manage synchronization and
coordination of interactions Why?
Scalability, globalization of supply networks
increases in mfg. and service systems TC
5.2 Trend 3. Optimization of parallelism among
resources Why? Increasing reliance on
autonomous decision- and task-support systems,
which operate in parallel
36
Nano and MEMS Systems -- Trends
Overall Trend Embedded nano and MEMS components
will enable significantly better products and
services Trend 1. Auto-ID by embedded devices
within products Why? To ensure flow coherence
in production and logistics of high variety/low
volume products TC5.1 / G. Morel Trend 2.
Optimization of location and architecture of
sensor arrays Why? Low-cost sensors can be
deployed effectively TC 5.2 Trend 3.
proliferation of DNA-based and bio sensors Why?
We are learning to exploit them for better,
faster diagnostics, response, prevention in
health and environmental control
37
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• CC 6 INDUSTRIAL SYSTEMS
• (APPLICATIONS)
• Denis Dochain

38
Emerging areasin Industrial Systems (CC6)
Information from three TCs 1) TC on Mining,
Mineral and Metal Processing 2) TC on Power
Plants and Power Systems 3) TC on
SAFEPROCESS (due to my late nomination, the TC
chairs have been lately contacted)
39
1) TC on Mining, Mineral and Metal Processing

• Internet-based process control
• Internet technologies have the potential to apply
  the advantages of
• this way of working to the high-level control of
  process plants
• - Remote monitoring and adjustment of plants,
• Collaboration between skilled plant managers
  situated in
• geographically diverse locations
• Relocation of the physical location of plant
• Management staff easily in response to business
  needs.

40
In particular, by combining Internet computing
and diagnostics, a new diagnostic system would
be constructed the  Internet based remote
diagnosis system . The system has rather high
performance vs price ratio and many good
features such as open structure, high
intelligence, shared resource and low cost.
Internet based remote diagnosis system has
aroused significant attention of diagnostics
researchers, and many study projects have been
put forth. Reference Li Hongsheng, Shi Tielin,
Yang Shuzi, "Internet-based Remote Diagnosis
Concept, System Architecture and Prototype",
Proceedings of the 3rd World Congress on
intelligent control and automation. June, 2000,
ehfei, P.R. China
41
II. Recycling systems control ( sustainable 
systems) Because of stiff global competition
and strict environmental regulations, the newly
designed chemical processes are becoming
increasingly integrated with material and energy
recycles. The traditional "unit operation"
approach to process control though simple, does
not consider the interactions among various
processing units and is becoming increasingly
inadequate. There is a need to design control
systems from the systems view point.
42
Example In an industrial plant, there are unit
operations like reactor, feed effluent heat
exchanger, furnace process under various
conditions. For environmental purposes, the
full material life cycle within the technosystem
"from cradle to grave" is frequently modeled.
The models are analyzed using methods from
control theory. These methods allow us to
investigate essential properties of the chain
concerning its dynamical behavior, particularly
with respect to stability and controllability.
Reference Surya Kiran Chodavarapu, Alex Zheng,
" Control system design for recycle systems",
Journal of Process control, Vol 11, pp 459468,
2001.
43
2) TC on Power Plants and Power Systems
Power systems are viewed as large-scale,
multi-input/multi-output, nonlinear systems
distributed over large geographic areas. They
are complex systems because of large
interconnections and the present trend towards
de/re-regulation is bringing in additional
uncertainties.
44
Large Scale System Problems Required signals
are available locally and computed actuator
signals may need to be sent to actuators
located at a long distance. Controllers are
dispersed geographically implementation of
centralized control schemes becomes expensive
because of the excessive computation time
requirements and unavailability of dedicated,
robust, high bandwidth communication links. It
requires the transmission of signals back and
forth, and to and from the controller locations
in a fast and reliable manner.
45
In the future, these characteristics will require
the development of Decentralized controls
possibly with some co-ordination provided by
multilevel heirarchical control schemes
providing centralized real-time and on-line
coordination of decentralized controllers.
Development of transducers and communication
schemes for fast signal transmission. Some of
the schemes are starting to involve satellite
technology for the synchronization of control and
protection actions at the geographically
disperse locations.
46
New Control Requirements to Meet Demands of
Deregulation Deregulation is creating an all new
set of control objectives and requirements
compared to the traditional control objectives.
This aspect of the power system control is
still in a state of flux as the requirements are
still evolving and changing continuously as more
experience is gathered. The controls will not
only have to meet the strictly technical
requirements for efficient and relaible control
and operation of the traditional vertically
integrated power systems but also the additional
economic objectives of profit maximization. It
is a bit difficult to predict the shape of
controls in the future from the point of view of
de-regulated electricity markets, but the control
philosophies will certainly change.
47
3) TC on SAFEPROCESS

• Systematic methodologies to design systems for
  fault detection
• and isolation in uncertain nonlinear processes
  with fixed false
• and missed alarm rates and mean detection
  delay are still lacking.
• The mentioned criteria (false and missed alarm
  rates and mean
• detection delay) are only seldom accounted for
  explicitly in the
• FDI system design. Yet they are the only
  meaningful criteria
• when it comes to applications.
• Research on fault tolerant control often
  considers that the
• diagnostic system provides perfect information
  on the faults that
• have occurred. Much more work is needed to
  analyse the interplay
• between the system for fault detection and
  isolation on the one hand,
• and fault accommodation or controller
  reconfiguration on the other
• hand, in a fault tolerant control architecture.

48

• From a practical point of view, there is a need
  for a methodology
• that allows one to design a device with given
  reliability.
• The question of how best to design a fault
  tolerant control
• architecture with specified performances that
  meets the given
• reliability requirement is open to the best of
  our knowledge.
• Monitoring systems over networks is another
  challenge. It involves
• accounting for the distributed character of the
  application as well as
• the transmission delays. The next step is to
  achieve fault tolerant
• control for geographically distributed systems.
  This type of work
• requires collaboration between control and
  communication engineers
• as well as computer scientists.

49
One of the aims of systems for fault detection
and isolation is to provide the information
needed for predictive maintenance or on
condition maintenance. As opposed to preventive
maintenance which amounts to performing
regular maintenance operations according to
the manufacturers advice, predictive or on
condition maintenance consists in monitoring
continuously the state of health of a device
and performing a maintenance operation only
when it is needed. The way to account for the
information provided by a diagnostic system
(including false and missed alarms) in the
maintenance planning of an installation, and
the way to mix preventive and on condition
maintenance so as to optimize maintenance and
exploitation costs calls for further research.
50
Distributed Parameter Systems A Re-emerging Area

• Starting point a paradox
• 1) DPS studied for a long time, seems to
  disappear
• (symptomatically, the keyword has disappeared in
  the IFAC)
• 2) New life of the DPS in areas like process
  control, transmission
• lines, irrigation and canals,

51

• Diagnosis A (simplistic?) explanation
• DPS is referring to highly complex and abstract
  mathematics
• ( infinte-dimensional systems )
• 2) In many applications, automatic control has
  reached a limit
• by using finite-dimensional models
• Example a tubular reactor can be approximated
  fairly well by a
• CSTR in many applications, but to increase the
  control performance
• The PDE dimension might be useful
• 3) Approximation tools (for analysis and
  numerical simulation
• are better known today)

52

• Research activity paradigm
• Mathematicians and engineers discuss more and
  more
• to fill the gap between theory and practice, and
  come
• along with solutions that solve the challenging
  practical
• problems
• Renewed activity in IFAC?
• - in CC1 or CC2?
• - workshop? (last one in 1989 in Perpignan)
• - ???

53
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• CC 7 TRANSPORTATION AND VEHICLE SYSTEMS
• (APPLICATIONS)
• Philipp Nenninger (for Uwe Kiencke)

54
Emerging Areas in Automatic Control
CC 7 Transportation Vehicle Systems

• IFAC, Rotterdam
• August, 26 2003
• Philipp Nenninger, Dr. Uwe Kiencke

55
Overview

• Introduction
• Automotive Control Systems
• Navigation Systems
• Design Development
• Conclusion

56
Higher traffic density
Increasing number of cars on the road, example
Germany
57
Challenges for Transportation Systems and Vehicles
Safety situation still unsatisfactory
58
Advanced Driver Assistance Systems

• Often drivers cannot cope with multiple
  requirements, two different solutions
• Traffic jam avoidance using dynamic navigation
  systems
• Intelligent cars with driver assistance systems

59
TC Automotive Control
60
TC Automotive Control
Human factor sensing
Image signal processing

• Application of accepted theoretical approaches
• New theoretical methods

Reliable Systems
61
TC Automotive Control
Simplify driving More focus on critical
situations
62
TC Transportation Systems

• Truly dynamic, offboard navigation

Data center
63
TC Transportation Systems

• Communication beween data centers
• Intermodal navigation

DC2
DC1
DC3
Increasing complexity of routing algorithms
64
TC Transportation Systems
Avoid critical situations Less accidents
65
Design Development
Distributed systems present new challenges
Real time design methodology
Safety analysis
Application of standard procedures to distributed
systems not possible
66
Design Development

• Systems design must ensure
• Safety
• Real time capability
• Need for an established best practice

67
Conclusion

• Simplify driving through accepted theoretical
  approaches, new methods
• Avoid critical situations through truly dynamic
  offboard navigation
• Establish best practice for the design of safe
  distributed real time systems

68
TC Automotive Control

• Vehicle to vehicle communication
• Image and signal processing
• Collision avoidance
• Requires new way to prepare information
• New theories
• Active communication (e.g. P2P for ACC)
• Sensing driver behavoir, ability (fatigue,
  anxiety, ...)
• Simplify driving -gt more focus on critical
  situations

69
TC Automotive Control

• Active vehicle to vehicle communication
• Electronic towbar
• Other communication

70
TC Transportation Systems

• Truly dynamic navigation
• Completely offboard (technological and business
  reasons)
• Intermodal routing strategies
• Communication between data centers
• Centralized route planning getting the big
  picture
• Route planning shouldnt cause more/new traffic
  jams (feedback)
• Global optimization with several independent
  routing centers
• Complexity of routing algorithms (multipath
  routing algorithms)
• Added value services
• Map data conversion

71
TC Transportation Systems

• Communication beween data centers
• Intermodal navigation

DC2
DC1
DC3
Increasing complexity of routing algorithms
72
Design Development

• Real time design methodology for complex
  distributed systems
• Ensuring realtime by design
• Guarantee end to end response times
• Safety analysis for distributed systems
• Error propagation
• Implications for user safety
• crucial for systems like X-by-wire
• safe by design Establish a best practice
• Safe process focus on the safety of a single
  system
• Now Safety of a distributed system consisting of
  several subsystems, proof of safety, interaction
  (problems caused by communication, specs,...)

73
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• CC 8 BIO ECOLOGICAL SYSTEMS
• (APPLICATIONS)
• Keith Godfrey (for Ewart Carson)

74
TC 8.1 Control and Automation in Agriculture

• Speaking animals
• Automated measurement of the responses of
  individual animals to differing feeding and
  climate regimes so as to optimise their
  management
• Speaking plants
• Automated measurement of plant responses to
  different treatments for the purpose of automatic
  control of processes in the soil-substrate-plant
  continuum
• Automation of harvesting and handling of crops
• Developments required due to a lack of skilled,
  affordable labour
• Automation for tracking and tracing of products
  and of the production condition
• Stage of maturity, variability, treatments etc.

75
TC 8.2 Modelling and Control of Biomedical
Systems

• Drug delivery systems
• Development of smart drug delivery systems based
  on modelling the individual patient and of the
  drug dynamics
• Model-based image interpretation
• e.g.in the context of PET imaging for
  understanding the dynamics of physiological
  processes for clinical application
• Bio-informatics
• A range of model-based developments for drug
  industry application
• Smart home-based monitoring and telecare
• Home-monitoring of the status of an individuals
  health and well-being and of the home environment
  in the context of integrated telecare provision

76
TC 8.3 Modelling and Control of Environmental
Systems

• Multi-objective decision support for water
  resource management
• Role of MODSS (based on control theory and
  decision theory) in the context of Integrated
  Water Resources Management
• Water reservoir management systems
• e.g. in the context of designing new dams
• Control systems for integrated water resource
  management
• Needed on the basin scale to deal with the
  consequences (flood and drought) of the
  greenhouse effect

77
TC 8.4 Control of Biotechnological Systems

• Control developments in wastewater management
• for production of biomedically relevant
  substances and degradation of pollutants
• Hard and soft sensors for biotechnological system
  control
• The limiting factor in advancing effective
  control systems
• On-line characterisation of biomass state
• Important since micro-organisms are the key
  engine of all bioprocesses

78
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• CC 9 SOCIAL SYSTEMS
• (APPLICATIONS)
• Talha Dinibutun

79
IFAC EMERGING AREAS WORKSHOP CC 9 SOCIAL
SYSTEMS A. TALHA DINIBUTUN
80

• TC 9.1 Economic and Business Systems   
• Increased importance of robust control theory
  (economic and business systems are characterized
  by much more uncertainty than engineering
  systems robust control may be more easily
  applicable than stochastic control theory). 
• (2) Non-causal systems (systems with
  anticipations) (these are frequently used due to
  rational expectations assumed for economic agents
  and to the requirement of analyzing learning
  processes traditional control theory cannot deal
  with such systems). 
• (3) Dynamic game theory (an area where
  mathematics developed by engineers can yield
  insights into economic problems that could not be
  obtained by more traditional economic approaches
  in game theory such as repeated games). 
• (4) Artificial intelligence and agent-based
  models (rather fashionable now in many areas
  these methods allow quick insights without doing
  cumbersome empirical work before). 

81

• TC 9.3 Developing Countries 
• Forecasting and  Control of Financial Process
  Dynamics (The underlying dynamics of financial
  processes is indeed much more complex than
  traditional econometric modeling and conventional
  stochastic-process modeling can capture.  The
  so-called soft-computing techniques comprising
  fuzzy-systems, neural-networks and
  genetic-algorithms as well as combinations
  exploiting their synergism are promising new
  methodologies and tools for exploring financial
  process dynamics as well as the dynamics of all
  other human-centered and non-technical systems.) 
• TC 9.5 SWIIS 
• (1) Automation and machine - aided thinking (is
  beginning to displace not merely manual jobs, but
  also quasi-skilled positions). 

82
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• INDUSTRIAL GUEST
• HANS DRIESSEN

83
SYSID 2003 Emerging Areas

• Hans Driessen
• Thales Nederland
• Joint Radar Sensors

84
Emerging Areas

• Conjecture 1
• The impact of Particle Filtering on industrial
  applications of
• Signal/Image Processing
• Automatic Control and Identification
• will be comparable to that of the Kalman filter.

85
Emerging Areas

• Arguments
• It is amazing how little (human) effort is
  required to design, implement and let run a
  Particle Filter solution to traditionally
  breath-taking complex stochastic dynamic
  estimation/detection problems.
• It brings back the problem to the essence of
  modelling in stead of inventing tricks to let run
  a particular application.
• Among the alternatives, it is the most efficient
  solution.

86
Emerging Areas

• Conjecture 2
• Radar sensor management is an emerging
  application area for automatic control research.

87
Emerging Areas

• Arguments
• Conventional radars have only little flexibility
  heuristics are used to control them.
• Technology developments are enabling much more
  flexible radars and integrated antenna systems
  these concepts are hot in the military market.
• The associated control problems are way too
  complex to rely on heuristics.
• Fundamental solutions hold the additional promise
  of improved performance.

88
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• INDUSTRIAL GUEST
• HERMAN VAN DER AUWERAER

89
Industry TrendAdvanced Vehicle Control

• Vehicle electronics take up an increasing part of
  the vehicle design and construction
• Drivers comfort, traffic congestion, safety
• Current functionality
• Engine control (ECU)
• Driver assistance (ABS, ESP, distance detection)
• Suspension control (mainly semi-active)
• Navigation, audio, multimedia
• Diagnostics (ABS)
• Comfort (climate control, lighting, seats)
• Supported by standards such as CAN bus etc.

90
Industry TrendAdvanced Vehicle Control

• This trend will continue, even be reinforced
• Emerging functionality
• Active suspension control
• Active noise and vibration control
• X-by-wire drive-by-wire, brake-by-wire.
• Power management (Electric/Hybrid Vehicles)
• Driver assistance (parking assistance, speed
  control, dangerous manoeuvring)
• Active safety measures (collision warning,
  pedestrian detection)
• Telematics (on-board e-services, navigation and
  beyond)

91
Industry TrendAdvanced Vehicle Control

• Future, Expected, functionality
• Integrated control throughout the vehicle
• Driver overrule (collision avoidance, lane
  control)
• Collision pre-detection and active measures
• Road-Vehicle and Vehicle-Vehicle communication
• Intelligent traffic control
• Evolving to semi-autonomous driving on selected
  tracks distance and speed control
• Driver condition monitoring and take-over
• Advanced infotainment
• .

92
Industry TrendAdvanced Vehicle Control

• Technology aspects
• Advanced system concepts, including
• Sensors and actuators for mechatronic
  applications
• Image sensors
• Human factor sensing
• Distributed systems, heterogeneous systems
• Vehicle high-speed data bus/network
• Wireless communication
• Embedded systems
• Vehicle integration

93
Industry TrendAdvanced Vehicle Control

• Technology aspects
• Advanced control concepts, including
• High-frequency control (noise..)
• Reliability, fail-safe operation, redundancy ,
  robust control
• On-board diagnostics
• Fuzzy inputs, decision discrimination
• Adaptive systems (vehicle loading, traffic
  weather situation..)
• Nonlinear control, control of non-linear systems

94
Industry TrendAdvanced Vehicle Control

• Technology aspects
• Systems approach to control
• Model-based control
• Including control in the vehicle design process
• Including sensor/actuator models in the vehicle
  CAE models
• Including active control loops in the vehicle CAE
  models
• Performance optimization by virtual models
• Active/passive trade-offs
• Leading markets aerospace and avionics technology

95
Industry TrendAdvanced Vehicle Control

• Boundary constraints
• Cost
• Weight
• Power
• Legislation
• Acceptance
• Design time/cost, time-to-market
• ..

96
IFAC TECHNICAL BOARD

• AUGUST, 2003
• IFAC EMERGING AREAS
• PROJECT
• INDUSTRIAL GUEST
• ALEX VAN DELFT

97
Business Drivers for the Process Industry -
Compliance/Traceability - Low cost -
Flexibility - Time to Market Consequences for
the Need for Directions in Theory and Application
Development Based upon my story about the time
delay from theory till application in Plant
automation 10-20 years time delay for Closed
loop real time optimisation, and based upon my
story about the increased business dynamics (the
Reversed Auction story) in process industries.
98
Needs for the Process Industry 1. Business
process optimisation integrated (from cradle to
grave) need for planning scheduling tools,
hybrid control, inferential sensor2. Integration
of design and control need for simultaneous
optimisation methods for design and control
alternatives3. Quality management need for
inferential quality sensors4. Batch control
need for multivariate SPC tools applicable to
batch, NL optimisation for batch, and inferential
sensors5. Real time diagnosis need for tools
for data handling6. Unmanned planned need for
availability of remote competences Alex van
Delft