MED 2015 Program

Conference Program

Coming soon…

Keynote Lectures


Magnus Egerstedt

Magnus Egerstedt
Georgia Institute of Technology



Biology has served as a rich source of inspiration when developing coordination and control strategies for distributed multi-agent systems, with algorithms derived from schooling fish or flocking birds having found their way into formation control strategies. However, as the cognitive ability of the animals increases, the team sizes tend to go down, with more specialized, heterogeneous roles emerging among the members of the team. In this talk, we investigate how heterogeneity can be formally understood in the context of control of multi-robot systems, present some initial findings as to why heterogeneous solutions are better than homogeneous ones, as well as discuss a number of open problems.


Magnus Egerstedt is the Schlumberger Professor in the School of Electrical and Computer Engineering at the Georgia Institute of Technology, where he serves as Associate Chair for Research and External Affairs. He received the M.S. degree in Engineering Physics and the Ph.D. degree in Applied Mathematics from the Royal Institute of Technology, Stockholm, Sweden, the B.A. degree in Philosophy from Stockholm University, and was a Postdoctoral Scholar at Harvard University. Dr. Egerstedt conducts research in the areas of control theory and robotics, with particular focus on control and coordination of complex networks, such as multi-robot systems, mobile sensor networks, and cyber-physical systems. Magnus Egerstedt is the Deputy Editor-in-Chief for the IEEE Transactions on Network Control Systems, the director of the Georgia Robotics and Intelligent Systems Laboratory (GRITS Lab), a Fellow of the IEEE, and a recipient of the ECE/GT Outstanding Junior Faculty Member Award, the HKN Outstanding Teacher Award, the Alum of the Year Award from the Royal Institute of Technology, and the U.S. National Science Foundation CAREER Award.



Manuel Silva

Manuel Silva
University of Zaragoza



Discrete Event Systems (DES) theory and engineering is mainly driven by needs raised by many different human-made systems (manufacturing, communications, logistic, workflow management, traffic, etc.). With the accelerated increase in complexity and size of the new technological constructions, the so called state explosion problem in DES analysis and synthesis becomes more and more acute. Two traditional conceptual and complementary means to deal with the derived computational complexities are structural techniques (which try to apply in different forms the classical divide and conquer aphorism) and fluid relaxations (natural variables of the DES are transformed into the non-negative reals). In the second case, the expected computational gains will be at the expense of the fidelity of the relaxed model and of the analyzability of certain properties.
This lecture will mainly focus on the second line. First, fluidization will be introduced and the legitimization of the hybrid and continuous approximation (both stochastic and deterministic) will be addressed. As a key advantage, the larger the population, the better is usually the approximation while the computational costs may decrease in an exponential way. Obviously, the relationship between the untimed and timed properties of the DES model and the corresponding properties of their continuous approximations is a very important issue. Moreover, the question of expressive power of the obtained formalisms – here in the Petri nets paradigm– will be raised. Together with the simulation of Turing machines (thus, presence of undecidabilities), the possibility of modelling complex and counterintuitive behaviors, as non-monotonicities, orbits or limit cycles and bifurcations due to the loss of hyperbolicity will be briefly pointed out. Finally, a broad perspective of the life-cycle of systems will be provided, and control, observation, diagnostic and implementation issues would be briefly considered.


Industrial–Chemical Engineer (Univ. of Seville, 1974), he receive the postgraduate (1975) and Ph.D. (1978) degrees in Control Engineering (INP de Grenoble). From 1975 to 1978 he worked for the CNRS at the Laboratoire d’Automatique de Grenoble. In 1978 he started the group of Systems Engineering and Computer Science at the University of Zaragoza (UZ).
His main research interests did include modeling, validation, performance evaluation, control and implementation of distributed concurrent systems using Petri Nets. He is author of the book Las Redes de Petri en la Automática y la Informática (AC, 1985; also Thomson-AC, 2002), coauthor of Practice of Petri Nets in Manufacturing (Chapman & Hall, 1993), and coeditor of Control of Discrete-Event Systems: Automata and Petri-Net Perspectives (Springer, 2013). Member or past member of the Steering Committees of the Int. Conf. on Application and Theory of Petri Nets, of the Workshops on Discrete Event Systems, and of the IFAC Int. Conf. on Analysis and Design of Hybrid Systems, was founder member of the Asociación Española de Robótica.
Prof. Silva was dean of the Centro Politécnico Superior (UZ) from 1986 to 1992, and president of the Aragon’s Research Council and of the Research and Innovation Committee of the French–Spanish Comisión de Trabajo de los Pirineos (1993-1995).
Interested in the History of Technology, is the editor of Técnica e Ingeniería en España (in Spanish:; some 6.000 pages edited). Prof. Silva has been distinguished with a medal from the city of Lille (France) and by the Association of Telecommunication Engineers of Aragón. He is Honoris Causa Doctorate by the University of Reims-Champagne-Ardennes, member of the Royal Academy of Engineering of Spain, and member of the Royal Academy of Sciences of Zaragoza.


Tore Hagglund

Tore Hägglund
Lund University


In industry, PID controller tuning is often viewed as an art that only few engineers master. Except for the on/off controller, the PID controller is the simplest controller one can imagine. It is also by far the most common controller in industry. There are numerous methods to tune the parameters of the controller proposed and published in the literature. In spite of this, many PID controllers are badly tuned or even not tuned at all. Why is it so? This question is the topic of the presentation that also will propose remedies to improve the situation.
A major reason for the situation is that there simply are so many controllers to tune in a process control plant. The instrument engineers have just not enough time to keep all controllers well-tuned. This fact has become even more evident in recent years due to the trend of reducing the personnel in process control plants. There are, however, many other reasons. There are several aspects that should be taken into account when tuning a PID controller. The closed-loop system should behave well with respect to set point changes, load disturbances, and measurement noise. It must also be robust to process variations, since most processes are nonlinear. Very few design methods take all these aspects into account, especially not simple ones. The many aspects and the fact that the specifications vary from case to case make the PID controller tuning a trade-off problem, where the trade-off has to be made by the engineers.
Set point changes can be treated using feed forward or set point weighting. Therefore, it is suggested that set point changes are excluded from the trade off and not considered in the first phase of the design. If load disturbances are measurable, feed forward can be used also in these cases, but there are often additional load disturbances that are not measurable. Therefore, it is suggested that the controllers are tuned with the goal to optimize the performance at load disturbances, and that requirements on robustness are introduced as constraints in this optimization. The constraints may act as tuning knobs that give the engineer the possibility to influence the tradeoff between performance and robustness.
Measurement noise has only recently been taken into account in PID controller tuning. This may be one reason why the derivative part is so seldom used in practice. It is difficult to take the noise into account without knowledge of the noise characteristics, but it is suggested that the filter in the PID controller is tuned so that it is as effective as possible in reducing the control signal variations due to noise, but with a limited reduction of the control performance at load disturbances.
Finally, it should be remarked that even a good design method will not solve the problem that the engineers have very limited time to spend on the controller tuning. Therefore, the presentation ends with a discussion about automatic tuning procedures, and how these should be developed to meet the demands and be accepted and used more extensively in industry.


Tore Hägglund received his M.Sc. in engineering physics in 1978 and his Ph.D. in automatic control in 1984, both from Lund Institute of Technology, Sweden, where he is currently a professor. During the PhD studies he also developed the relay auto-tuner for automatic tuning of PID controllers together with Professor Karl Johan Åström. The method got patented and is now implemented in many industrial products. Between 1985 and 1989 he worked for Alfa Laval Automation (now ABB) on the development of industrial adaptive controllers. He is a coauthor of Advanced PID Control and Process Control in Practice.
He is author or coauthor of 10 books, 6 book contributions, and more than 115 journal and conference papers. He holds 6 patents and has received along his research career several distinctions and awards, such as the awards from “Innovation Cup” in 1985 (Automatic tuning), 1989 (Deadtime compensator), 1993 (Oscillation detection), 1997 (Detection of conservative tuning), and 1999 (Ratio controller). Furthermore, he has received also the Raymond D. Molloy Award from ISA for being a best-selling author in 2007.
His current research interests are in the areas of process control, tuning and adaptation, and supervision and detection.