Peter Wellstead SFI Research Professor Hamilton Institute NUI, Maynooth Co. Kildare Ireland

Background

Born and educated in Hertfordshire, England. Apprentice in mechanical and electrical trades with Marconi Instruments for eight years. Later becoming a student apprentice and part-time student at Hatfield College of Technology (now Hertfordshire University). Awarded the De Havilland Prize (named for the aviation pioneer Sir Geoffrey De Havilland – whose company was based in Hatfield) and graduated with BSc 1^st class Honours in Electrical Engineering. From 1967-1970 he studied at Warwick University School of Engineering Science for the degrees M.Sc (Control and Systems), PhD (Engineering Science). During his PhD research he developed methods for closed loop frequency response identification [1], characterised aliassing in frequency response estimates, [2] and a multi-sine synthesis method for pseudo random test signals as an alternative to the then standard PRBS approach [3]. Both closed frequency response estimation and multi-sine techniques are now standard in signal processing instruments and system identification. As a Technical Fellow at CERN (1970-1972) he worked with one of the most innovative computer scientists of the day, Gordon Brandon, in the computer control and digital signal processing group of the bubble chamber division. Together they developed the first real time feature recognition system for digitised bubble chamber photographs. At the time this form of real time, on-line signal processing was in its infancy , and the CERN group pioneered real time methods for this problem. In 1972 he joined the staff of the Control Systems Centre, UMIST, as a lecturer, becoming Professor of Control Engineering in 1990 and Head shortly thereafter. The Control Systems Centre was one of the first UK postgraduate centres for interdisciplinary control engineering research, control systems analysis and its applications. His first scientific work at UMIST was to develop with Francisco Carvalhal (seconded from the National Civil Engineering Laboratory (LNEC) of Portugal) real time adaptive spectral equalisation algorithms for vibration testing. This system was later implemented at LNEC as part of their earthquake testing system for civil and engineering structures, and implemented commercially by vibration control system manufacturer Solartron. Francisco is now a senior scientific administrator in Lisbon. The global stability properties of this form of adaptive system was later derived by Martin Zarrop [4]. In the area of system identification he worked on fast algorithms [5] and maximum likelihood methods [6] for parametric system identification with Allan Robins and Denis Prager and developed the instrumental product moment test structure testing method [7]. System identification application ‘firsts’ by his group are the dynamical identification an automotive gas turbine (with CAV Ltd) for multivariable control, estimation of the vertical motion dynamics of the Joint European Torus, and the wave energy device known as Salter’s Duck. The work on adaptive shaping of closed loop test spectra, combined with interest in the sensitivity issues of minimum variance control methods led to collaboration with John Edmunds and other research students to develop the theory and application of the Pole Assignment Self – Tuner, [8]. Later with research student, Denis Prager, he developed the MIMO version of the same algorithm [9]. Self-tuning application ‘firsts’ by his group include a digital self-tuning diesel engine regulator (with CAV Ltd), self – tuning super-heater control (with CEGB) and a low cost, portable self-tuning instrument. Pole assignment is now a standard textbook technique for adaptive control. He had a fruitful relationship with Lucas Automotive Research and was appointed the Lucas Professor of Automotive Control as a result. A highlight of this twenty year period of collaborative research was the development, with Peter Scotson, of a Self-Tuning Extremum Control method for gasoline engine spark angle optimisation[10]. This method was patented and used by Lucas and its derivative companies. In addition to general automotive control research, his research group also developed methods for hybrid systems found in automotive applications. These methods were used to analyze problems in rule based control of ABS brake control and thus save valuable track testing time [11]. As a result of the collaboration with Lucas Automotive, 15 PhD and MSc graduates from UMIST joined the Lucas Research Centre and formed the core of Lucas’s control and systems technology expertise. A number of these engineers remain and have influential positions in the automotive research sector with TRW Inc. His work on self-adaptive/self-tuning algorithms continued with the development of Two-Dimensional (2-D) predictor algorithms and their dual 2-D minimum variance control methods [12]. These 2-D control methods found application in the area of sheet manufacturing systems, such as paper making [13] and plastic film extrusion, and he was given a Royal Society Industry Fellowship to develop the applications of the theory in the film extrusion area. Traditionally paper machine wet end control and film extrusion has been treated as a conventional control problem. The Control Systems Centre group are credited with changing this viewpoint and introducing the 2-D paradigm to the sheet product manufacturing area. Experience with practical film extrusion led to two new bodies of work: The first concerned 2-D sensing, [14]. In particular, the signal processing of 2-D data when the 2-D surface is scanning in a sparse way – as with a scanning sensor. This work with colleagues Steve Duncan, Martin Zarrop and Jim Skelton, led to the application of Generalised Sampling Theory to the reconstruction of 2-D surfaces from sparse (scanned ) data, and new methods for reconstructing full sheet information from conventional scanning gauges [15]. Steve Duncan and his group has since generalised the 2-D sensing idea to actuation [19]. The second body of work was the result of collaboration with the eminent NIR instrument designers and scientists, Roger Edgar and Peter Hindle on a portable and robust form of the classical Michelson interferometer so that the NIR spectrum of light from a product could be measured and analysed in real time. This led to the invention, again with John Edmunds, of new active control methods for two-beam interferometers. With first round funding from Manchester Technology Fund , he and a team of UMIST colleagues developed the Active Control Interferometer technology into a compact portable NIR spectrometer. A company, Acimetrics, has since been incorporated to develop and commercialise the ACI technology. The application sectors are industrial real-time non contact measurement of organics, and in the medical sector – high throughput, non-invasive measurement and analysis of blood and tissue contents. He has writtenreports page of the systems biology group) text books on Systems Modelling [16] and Self-Tuning Control [17], directed 23 government research grants, and supervised over 90 MSc and 30 PhD research students. He co-founded and directed a Contrreports page of the systems biology group)ol Technology Transfer Network for the North of England. Over its 6 year life the Network managers, Steven Schooling and Lita Denny, assisted over 30 companies and developed the core technology for a ‘start-up’ company working on battery and corrosion cell testing. As an educationalist his passion is for learning through practical experience, in particular the use of scale models of real systems for teaching systems dynamics, modelling and control. Based on the concept that students should deal with real systems, not simulations, he and Roy Moody (Chief Technician in the Control Systems Centre), developed in the 1970’s and 1980’s a range of scale model systems from different application areas of systems dynamics, and used them to teach Masters and undergraduate students over a period of three decades. Products based on his designs now appear in control laboratories throughout the world ( see Control Systems Principles). In the community, he is a former Honorary Editor of the Institute of Electrical Engineers (IEE) Proceedings, Control Theory and Applications, and serves on a number of committees and international conference panels. He is  a regular mountain biker but his last vintage motorcycle run was the Banbury Run with Bill Geraldine  in 2003.

Project Photo Gallery Early output from the real time bubble chamber feature recognition programme (circa 1970, CERN Geneve). The first Coupled Tanks Apparatus (circa 1975, Control Systems Centre, UMIST) The Solartron 1200 Signal Processor with closed loop identification and multi-sine test signal generation (circa 1980, Solartron Instruments) Digital self-tuning diesel engine speed regulator (circa 1978, Control Systems Centre UMIST with CAV Ltd) Roy Moody bench testing a steam engine experiment (circa 1985, Control Systems Centre Teaching Laboratory)

Current position: He is Science Foundation Ireland (SFI) Research Professor of Systems Biology, based at the Hamilton Institute. For full details see http://www.systemsbiology.ie/.

Current research interests:

Systems Biology – his personal interest is to contribute to a systems understanding of Parkinson’s Disease and other Neurodegenerative diseases.

Non-Invasive Sensing – in particular the use of his ACI technology for high throughput non-invasive medical and biological measurement.

System Identification [19,20] ,

Control System Technologies in the Innovation process, [18, 21]

Selected references:

1. Reference Signals in Closed-Loop Identification, Int. J. Control, 26, 945-962, 1977

2. Signal Aliassing in System Identification, Int. J. Control, 22, 3, 1975

3. Pseudo Noise Test Signals and the Fast Fourier Transform, Electronic Letters, 11, 10, 1975

4. Self-Tuning Regulators: Non-Parametric Algorithms, Int. J. Control, 37, 4, 787, 1983

5. Fast Algorithms in Systems Identification, Int. J. Control, 33, 3, 1981

6. Interactive Maximum Likelihood Estimation, Int. J. Control, 32, 6, 1980

reports page of the systems biology group)7. An Instrumental Product-Moment Test for Model Order Testing, Automatica, 14, pp 89-91, 1978

8. Pole Assignment Self-Tuning Regulator, Int. J. Control, 30, 1, 1979

9. Multivariable Pole-Assignment Self-Tuning Regulators, Proc. IEE. 128, pt. D, 1981

10. Self-Tuning Extremum Control, Proc IEE, Vol. 137, Pt. D. 3, pp 165-175, 1990

11. Analysis and Redesign of an Antilock Brake System Controller, IEE Proc, vol. 144, 5, pp 413-426, 1997

12. Self-Tuning Prediction and Control for Two Dimensional Processes Part 1: Fixed Parameter Algorithms, Int. J. Control, 62, 1, pp 65-107, 1995

13. Two Dimensional Control Systems: Application to the CD and MD Problem, Pulp and Paper Canada, 95, 4, 1994

14. On-line Sensing of Paper Machine Wet-end Properties, IEEE Trans. on Control Systems Technology, vol35, 8, pp 232-247, 1997

15. Reconstructing 2-D Variations from Scanned Measurements, Proc. Control 2000, Cambridge, Sept 2000.

16. Introduction to Physical System Modelling, Academic Press, London, 1979

17. Self-Tuning Systems, P.E. Wellstead and M. B. Zarrop, J. Wiley, London, 1991

Selected recent references:

18. Control and systems concepts in the innovation process, IEEE Control Systems Magazine (Special Section on Control as a Systems Technology), pp 21—58, December 2003.

19. Processing data from scanning gauges on industrial web processes, Automatica, 40, pp 431—437, 2004.

20. Distortion of web profiles by scanning measurements, Pulp and Paper Canada, 104:12, pp 81—84, 2004.

21. Systems modelling in team building and innovation, IFAC Conference Advanced Control Strategies for Social and Economic Systems, Vienna, September, 2004.

22. A Frequency Domain Approach to Determining the Path Separation for Spray Coating, IEEE Trans ASE, vol 2, 3, pp 233-240, 2005.

23. Active alignment for two-beam interferometers, Review of Scientific Instruments, 77 (1), 2006.

Selected Systems Biology references:

24. Schroedinger’s Legacy, E.T.S. Walton Lecture, Royal Irish Academy, April, 2005  (see reports page of the systems biology group)

25. The Dynamic Systems Approach to Control and Regulation of Intracellular Networks, FEBS Letters, 579, pp 1846 – 1853, 2005.

26. A Systems and Signal Oriented Approach to Intracellular Dynamics, Biochemical Society Transactions, 33, 3, pp 507 – 515, 2005.

27. A Unified Framework For Unravelling The Functional Interaction Structure Of A Biomolecular Network Based On S timulus-Response Experiment Data., FEBS Letters, 579, pp 4520-4528, 2005.

28. Feedback Medicine: Control Systems Concepts in Personalised, Predictive Medicine and Combinatorial Intervention.Report (arXiv:q-bio. TO/0603032 v1), 2006.

29. A Plea for More Theory in Molecular Biology, In Systems Biology - Applications and Perspective. E.Butcher, P.Bringmann, B.Weiss (eds.), Springer-Verlag, 2006.

30. From Regulation, Control and Adaption to the Coordination of Cell Function. In Regulation. M.Laubichler, H.-J.Rheinsberger, P.Hammerstein (eds.), for publication in Vienna Series of Theoretical Biology, MIT Press, 2006.

31. On the Industrialisation of Biology, Catedra Amundson, University of Guadalajara, September 2006 (see reports page of the systems biology group)

32. Power-Law models of signal transduction pathways.  J. Vera, E. Balsa-Canto, P. Wellstead, J.R. Banga, O. Wolkenhauer. Cellular Signalling. 19(2007):1531-1541

33. Control Opportunities in Systems Biology, Plenary Address, IFAC Symposia, CAB 2007 and DYCOPS 2007 (see reports page of the systems biology group)