Dr. Riccardo M.G. Ferrari

 

Assistant Professor in Fault Tolerant Control

Delft Center for Systems and Control

Delft University of Technology

 

Mail: r dot ferrari at tudelft dot nl

Phone: +31 (0)15 27 83519

Office hours: Thursday 10.30-12:30

Where: Room 34-C-2-260 (building 34, tower C, second floor)

  R.M.G. Ferrari picture

 

   
 
 
 
View Riccardo Ferrari's profile on LinkedIn

 

Vacancies

 

2 PhD positions: Fault diagnosis and fault tolerant control of offshore wind farms

I am looking for 2 talented, outstanding PhD candidates with an M.Sc. degree (or close to completion) in Systems and Control, or Applied Mathematics, Electrical or Mechanical or Aerospace Engineering, or related field, with theoretical background and interest in Fault Diagnosis, and/or Fault Tolerant Control, Wind Energy, Structural Health Monitoring, and with good command of the English language (knowledge of Dutch is not required).

 

General project description: The candidate will conduct theoretical and applied research on offshore wind farms subjected to faults and structural degradation, in particular due to corrosion. The research will develop tools for modelling the dynamic behaviour and the failure mechanisms of offshore wind turbines, for estimating their current health level and for implementing load limiting control algorithms that take into account the health status of single turbines while maximising the wind farm total power production. Such algorithms will be tested in simulation, on scale models and on actual offshore test sites.

 

The position is in the context of the research project “O&M tools integrating accurate structural health in offshore energy” (WATEREYE), funded by the European Commission under the call H2020-LC-SC3-2019-RES-TwoStages.

Biography

  • I was born in Trieste on November 21st, 1979.
  • On April 21st, 2004, I received the "Laurea" degree cum laude in Electronic Engineering from the University of Trieste. My dissertation was about the theoretical study, experimental measuring and simulation via Cell Method of the acoustoelastic effect in thin metal sheets. This thesis was supported by Danieli Automation (Udine, Italy).
  • On October 28th, 2005, I graduated in Classical Piano at "G. Tartini" Academy of Music, scoring 109 out of 110. Mi dissertation was about the pianistic touch, both from the pianist's point of view and the physicist's one, and involved the development of a FEM model of a grand piano action.
  • On April 20th, 2009, I earned a PhD in Information Engineering (January 2005 - April 2009), after defending a thesis entitled “Distributed Fault Detection and Isolation of Large-scale Nonlinear Systems: an Adaptive Approximation Approach”. My research activity was performed at the Control Systems laboratory, under the supervision of prof. Thomas Parisini, and was supported by a grant funded by Danieli Automation.
  • From January 2008 to August 2015 I was employed by Danieli Automation, first as a designer and then as an Executive Manager in charge of its R&D department.
  • From September 2015 to March 2017 I was Postodoctoral Researcher at Delft Center for Systems and Control, at Delft University of Technology.
  • Since September 2016, I am a Marie Curie fellow working on the project SURE.
  • On April 1st 2017, I was appointed Assistant Professor in Fault Tolerant Control at Delft Center for Systems and Control, at Delft University of Technology.

News

2019, July 2nd: we are happy to announce that our proposal for the project “O&M tools integrating accurate structural health in offshore energy” (WATEREYE)  has been selected by the European Commission for funding under the call H2020-LC-SC3-2019-RES-TwoStages !

2019, February 18th: we are very happy to announce that Yichao Liu, a Postdoc in our group, just received  a Marie Curie Fellowship to continue here at DCSC his work on fault diagnosis and fault tolerant control of Offshore Wind Turbines!

2019, January 11th: the list of new MSc final projects is available here.

2018, August 30th: our paper on "Differentially-Private Distributed Fault Diagnosis for Large-Scale Nonlinear Uncertain Systems" was selected as a finalist for the Paul M. Frank Award at SafeProcess 2018 in Warsaw.

2018, August 27th: this week we are presenting a paper on “Attack Detection and Estimation in Cooperative Vehicles Platoons: A Sliding Mode Observer Approach,” at NecSys 2018 in Groningen (preprint), and a paper on "Differentially-Private Distributed Fault Diagnosis for Large-Scale Nonlinear Uncertain Systems" at SafeProcess 2018 in Warsaw.

2018, June 8th: we took part to IFOT 2018 Science Festival and presented to the public some results of our upcoming paper on Detection of Attacks to Cooperative Autonomous Vehicles (preprint here). Here are some links on YouTube videos including examples of our interactive demo. Thanks to all people that visited our stand!

Slides:

Normal conditions:

No attack detection and no fault tolerance:

Attack detection and fault tolerance:

2018, April 3rd: our invited sessions and our papers at both ECC 2018 and SafeProcess 2018 have been accepted! Thanks to André, Vahab and Tamas for co-authoring! See you in Cyprus in June and in Warsaw in August to talk about Detection of Sensor Data Injection Attacks with Multiplicative Watermarking and about Differentially-Private Distributed Fault Diagnosis for Large-Scale Nonlinear Uncertain Systems.

2018, April 3rd: Two Internship Assignments for MSc students are available in the field of system identification for injection moulding and fiber sensors, in cooperation with Promolding and their daugher company Somni. Please see for details here and here: we are looking for strong and motivated students!

2017, September 13rd: An Internship Assignment for MSc students is available, in cooperation with Océ, the leading printer manufacturer. Please see for details here: we are looking for strong and motivated students!

2017, June 14th: A PhD position is available in my group on Fault and Cyber-attack Tolerance for Distributed Systems, with applications to formations of mobile robots. Please see for details here: we are looking for strong candidates!

2017, April 1st: Since today I am continuing here at DCSC as an Assistant Professor in Fault Tolerant Control, in the group of prof. Jan-Willem van Wingerden. Can't be more excited !!

2017, February 27th: Both our papers have been accepted for presentation at IFAC World Congress 2017. See you in Toulouse to talk about Automatic Synthesis of Probabilistic Fault Detectors for Smart Buildings, and on Replay Attack Detection and Identification in Cyberphysical Systems! Thanks to Simone, Henrik and André.

2017, February 16th: today there will be the first lecture of the course Control Systems Lab (SC42045) which I am teaching this quarter. Come to lectures with your laptop with Chrome installed, as we will use an active learning tool: ASQ ! Thanks to Vasileos, Alessandro and Sampit for letting me use it!

2017, January 27th: We want your paper! Deadline extended to 31/3 for the Workshop on Industrial Computational Applications (WICA2017) I am co-organizing with former colleagues from Univ. of Trieste. It will be hosted during the upcoming International Conference on Computational Science and Its Applications (ICCSA 2017) that will be held in beautiful and windy Trieste (Italy) on July 3-6, 2017

2017, January 20th: Both our papers have been accepted for presentation at American Control Conference 2017. See you in Seattle to talk about Probabilistic Thresholds for Fault Detection, and on Routing Attack Detection and Identification in Cyberphysical Systems! Thanks to Vahab and André.

2016, October 14th: I will give a short pitch on the problem of detecting faults and cyber-attacks in Critical Infrastructure at the Cybersecurity Matchmaking Event 2016 in Den Haag.

2016, September 15th: after finishing my one year in the AMBI project, I will continue here at DCSC for other two years as a Marie Curie fellow thanks to the grant no. 707546. I will work on the project "SURE: Safe Unmanned Robotic Ensembles". More details here.

2016, March 18th: a new journal paper on fault diagnosis of networked systems, addressing the problem of asynchronous and delayed measurements coming from different sensor networks, is available on IEEE Xplore! Here is the link: A Distributed Networked Approach for Fault Detection of Large-scale Systems.

2015, September 14th: my first day at Delft Center of Systems and Control! I will spend one year as a Postdoctoral Researcher in the project "Advanced Methods for Building Diagnostics and Maintenance" (AMBI, grant no. 324432), funded by Marie-Curie call FP7-PEOPLE-2012-IAPP. Researchers from both industry (Honeywell Labs in Prague) and academia (Delft Technical University and Oxford University) are participating, here is the project homepage.

Research

SURE: Safe Unmanned Robotic Ensembles

SURE is a two year Marie Curie individual fellowship funded under MSCA-IF-2015-EF (grant no. 707546).

Self-driving personal cars that coordinate themselves to eliminate road accidents while doubling the capacity of highways; swarms of autonomous flying machines that quickly map and explore critical areas during emergency response; fleets of service robots that inspect, clean or repair large infrastructures or vessels in a fraction of the time needed to do this manually. These are just a few examples of so called robotic ensembles, and that have the potential of revolutionizing our life in a short time span. But a key factor that can prevent them to deliver their expected benefit, is un-safety. What if such complex collective systems may be made unusable, or even dangerous, by failures or misbehaviours of a few, or even a single of their members? Threats may come from simple and inevitable physical damage occurring to individual units, or may even be the result of the deliberate action of compromised ones, due to malicious cyber-attacks directed at disrupting the service provided by the ensemble. Eradicating such scenarios is precisely the objective of the present proposal, that by directing the fellow research effort to the conception and analysis of an innovative multi-scale, adaptive and distributed diagnosis approach, will actively contribute to the creation of Safe Unmanned Robotic Ensembles (SURE). Delft University of Technology (TUD), combining top-level expertise in modelling and control of vehicular traffic flows, in distributed robotic applications, and in the conduction of robotic test-beds at its world-class experimental facilities, is a unique institution. The one where the researcher will be able to prove the effectiveness of the proposed safety techniques by operating a laboratory scale robotic test-bed. This will pave the way, within the Dutch Automated Vehicle Initiative, for trials on actual cooperative self-driving cars, a motivating example that is projected to be worth 71 billions of Euros in 2030.

 

 

Distributed Fault Diagnosis of Large–Scale Systems

Fault diagnosis (and accommodation) architectures are key components in the development of autonomous and intelligent systems that operate reliably in the presence of system faults, which are characterized by critical and unpredictable changes in the system dynamics. An approach that gained widespread adoption is the so–called model– based method, where a healthy analytical model of the system is used to estimate its states or outputs. Estimation errors, called residuals, can thus be computed and compared to suitable thresholds for detecting the presence of a fault. Conventionally, such diagnosis methods are implemented in a centralized architecture, and this leads to the impossibility of addressing large–scale systems, because of practical limitations on the available processing and communication capacity.

The individual contribution to this problem was the development of a novel distributed fault diagnosis architecture, based on an overlapping decomposition of the system to be monitored, and on the use of many, distributed Local Fault Diagnosers. The local diagnosers follow a divide et impera paradigm, and feature an adaptive on- line approximator of the interconnection dynamics between different subsystems, and a consensus–like filter for taking advantage of redundant measurements. This led to the first published result on distributed fault diagnosis of nonlinear, uncertain large–scale systems described by continuous or discrete–time equations.

 

Figure 1: General scheme of a centralized, decentralized and distributed approach when dealing with a large–scale, distributed system.

 

New results, recently submitted for publication, include the formulation of the Distributed Fault Diagnosis problem for the important case of networked systems. These systems are characterized by unpredictable communication delays and losses, and include the case of Wireless Sensor Networks as well. A natural extension of this work, under way, is to consider the problem of malicious attacks in Cyber Physical Systems, that is engineered systems that originate from a close integration of computation and communication capabilities with physical processes.

Numerical Modelling of Ultrasonic Waves in Solids via the Cell Method

Ultrasonic waves are commonly used in Non–Destructive Testing and Evaluation (NDT-E) procedures, during industrial production, and in residual life assessment, of structural materials such as metals, concrete or composites. Internal flaws can be detected through the presence of echoes, while physical properties such as yield can be computed through measurements of speed and/or attenuation of ultrasound. One subtle effect that can be taken advantage of in ultrasonic NDT-E, is the so–called Acoustoelastic Effect, that relates the internal stress state of a solid to changes in the speed of sound.

The individual contribution to this problem was the development of an analytical and, subsequently, of a computer numerical model of the acoustoelastic effect, based on the Cell Method, devised by Prof. Enzo Tonti. The success of the model led to the construction of a laboratory demonstrator for evaluating the feasibility of acoustoelastic measurements on hot strip mills, during the production of steel sheets. The demonstrator was based on custom self–built Electro-Magnetic Acoustic Transducers (EMAT), tuned in order to generate a single Lamb wave mode.

The MSc thesis originating from this modelling and experimental work was awarded the “Giacomini Prize” in 2005.

 

a)

b)

Figure 2: a) mesh used to model b) longitudinal ultrasonic waves with the Cell Method.

At a very recent conference, I presented an approach that is based on a Cell Method elastodynamic model, andwhich addresses the problem of Fault Diagnosis of distributed parameter systems. Thanks to this work, a trait d’union between my two research subjects was finally drawn.

Awards

Group

Current members

PostDoc

  • Dr. Yichao Liu (working on Fault Diagnosis of Offshore Floating Wind Turbines)

PhD

  • Niloofar Jahanshahi (working on distributed fault and cyber attack detection for multi-agent systems)

MSc

  • Maurits Mooren (working on Nonlinear Watermarking for Cyber Attack Detection)
  • Folkert Ritsma (working on Machine Learning Approaches to Design of Probabilistic Thresholds for Fault Diagnosis)
  • Rebecca Kossen (working on Fault Diagnosis of Autonomous Vehicles)
  • Filippo Maffioletti (working on Fault Prognosis of Inkjet Printing Heads)
  • Corrado Pezzato (working on Active Inference for Fault Diagnosis of a Robotic Arm)
  • Gabriele Tombakaite (working on Safe Platoons of RC Cars)
  • Simon van der Marel (working on Stereo Vision for Platoons of RC Cars)

Visiting

  • Dr. Ping Wu (working on Gaussian Processes for Fault Diagnosis of Wind Turbines)

Publications

You can also check my profile at Google Scholar Citations or at Researchgate

International journals

  1. F. Boem, R.M.G. Ferrari, C. Keliris, T. Parisini, and M.M. Polycarpou, “A distributed networked approach for fault detection of large-scale systems,” IEEE Transactions on Automatic Control, vol. 62, no. 1, pp. 18–33, 2017 (DOI).
  2. F. Boem, R.M.G. Ferrari, T. Parisini, and M.M. Polycarpou, “Distributed fault diagnosis for continuous- time nonlinear systems: The input–output case,” Annual Reviews in Control, vol. 37, no. 1, pp. 163 – 169, 2013 (DOI).
  3. R.M.G. Ferrari, T. Parisini, and M.M. Polycarpou, “Distributed fault detection and isolation of large-scale discrete-time nonlinear systems: An adaptive approximation approach,” IEEE Transactions on Automatic Control, vol. 57, pp. 275–290, Feb 2012 (DOI).
  4. F. Boem, R.M.G. Ferrari, and T. Parisini, “Distributed fault detection and isolation of continuous-time non-linear systems,” European Journal of Control, vol. 17, no. 5–6, pp. 603 – 620, 2011 (URL).
  5. R.M.G. Ferrari, T. Parisini, and M.M. Polycarpou, “Distributed fault diagnosis with overlapping decom- positions: An adaptive approximation approach,” IEEE Transactions on Automatic Control, vol. 54, pp. 794–799, April 2009 (DOI).

International conferences

  1. 2018 Safe Process, Warsaw, August 29-31. Differentially-Private Distributed Fault Diagnosis for Large-Scale Nonlinear Uncertain Systems. V. Rostampour, R.M.G. Ferrari, A.M.H. Teixeira, T. Keviczky (to appear)
  2. NecSys 2018, Groningen, August 27-28. Attack Detection and Estimation in Cooperative Vehicles Platoons: A Sliding Mode Observer Approach. N. Jahanshahi, R.M.G. Ferrari (preprint)
  3. 2018 European Control Conference, Cyprus, June 12-15. Detection of Sensor Data Injection Attacks with Multiplicative Watermarking. A.H. Teixeira, R.M.G. Ferrari (to appear)
  4. 2017 IFAC World Congress, Toulouse, July 9-14. A Message Passing Algorithm for Automatic Synthesis of Probabilistic Fault Detectors from Building Automation Ontologies. R.M.G. Ferrari, H. Dibowski, S. Baldi (DOI)
  5. 2017 IFAC World Congress, Toulouse, July 9-14. Detection and Isolation of Replay Attacks through Sensor Watermarking. R.M.G. Ferrari, A.H. Teixeira (DOI)
  6. 2017 American Control Conference, Seattle, May 24-26. A Set Based Probabilistic Approach to Threshold Design for Optimal Fault Detection. V. Rostampour, R.M.G. Ferrari, T. Keviczky (DOI)
  7. 2017 American Control Conference, Seattle, May 24-26. Detection and Isolation of Routing Attacks through Sensor Watermarking. R.M.G. Ferrari, A.H. Teixeira (DOI)
  8. 2015 Safe Process, Paris, September 2-4, 2015. Optimal Topology for Distributed Fault Detection of Large-Scale Systems (Invited paper). F. Boem, R.M.G. Ferrari, T. Parisini and M.M. Polycarpou (URL)
  9. 2015 IFAC Workshop on Mining, Mineral and Metal Processing, Oulu, Finland, 25-27 August 2015. An Algebraic Approach to Modeling Distributed Multiphysics Problem: The Case of a DRI Reactor. R.M.G. Ferrari, F. Boem and T. Parisini (URL)
  10. 2015 IFAC Workshop on Mining, Mineral and Metal Processing, Oulu, Finland, 25-27 August 2015. A Model for Aluminium Quenching Process Optimal Setup. A. Polo, C. Aurora, F. Cuzzola and R.M.G. Ferrari (URL)
  11. 2013 European Control Conference, Zurich, July 17-19, 2013. An Algebraic Approach for Robust Fault Detection of Input-Output Elastodynamic Distributed Parameter Systems. R.M.G. Ferrari, T. Parisini and M.M. Polycarpou (URL)
  12. 2013 American Control Conference, Washington, June 17-19, 2013. Distributed Fault Detection for Uncertain Nonlinear Systems: a Network Delay Compensation Strategy. F. Boem, R.M.G. Ferrari, T. Parisini and M.M. Polycarpou (URL)
  13. 2012 SafeProcess, Mexico City, Mexico, August 29-31, 2012. Distributed Fault Diagnosis for Input-Output Continuous-Time Nonlinear Systems. F. Boem, R.M.G. Ferrari, T. Parisini and M.M. Polycarpou (URL)
  14. 2011 Conference on Decision and Control, Orlando, USA, December 12-15, 2011. A Distributed Fault Detection Methodology for a Class of Large-scale Uncertain Input-output Discrete-Time Nonlinear Systems. F. Boem, R.M.G. Ferrari, T. Parisini and M.M. Polycarpou (DOI)
  15. 2011 IFAC World Congress, Milano, Italy, August 28th-September 2nd, 2011. Fault Detection and Isolation of the Wind Turbine Benchmark: an Estimation-based Approach. X. Zhang, Q. Zhang, S. Zhao, R.M.G. Ferrari, M.M. Polycarpou and T. Parisini (Invited paper, 2nd place in Competition on Fault Detection and Fault Tolerant Control for Wind Turbines, DOI).
  16. 2010 Conference on Decision and Control, Atlanta, USA, December 15-17, 2010. Distributed Fault Diagnosis of Large-Scale Discrete-Time Nonlinear Systems: New Results on the Isolation Problem. R.M.G. Ferrari, T. Parisini e M.M Polycarpou(DOI).
  17. 2008 American Control Conference, Seattle, USA, June 11-13, 2008. A Robust Fault Detection and Isolation Scheme for a Class of Uncertain Input-output Discrete-time Nonlinear Systems. R.M.G. Ferrari, T. Parisini e M.M Polycarpou(DOI).
  18. 2007 Conference on Decision and Control, New Orleans, USA, December 12-14, 2007. A Fault Detection and Isolation Scheme for Nonlinear Uncertain Discrete-Time Sytems. R.M.G. Ferrari, T. Parisini e M.M Polycarpou(DOI).
  19. 2007 American Control Conference, New York, USA, July 11-13, 2007. Distributed fault diagnosis with overlapping decompositions and consensus filters. R.M.G. Ferrari, T. Parisini and M.M Polycarpou(DOI).
  20. 2006 CCA/CACSD/ISIC, Munich, Germany, October 4-6, 2006. A Fault Detection Scheme for Distributed Nonlinear Uncertain Systems. R.M.G. Ferrari, T. Parisini and M.M Polycarpou (DOI).

Technical reports

  1. University of Trieste technical report, 2010. Distributed fault detection and isolation of large-scale nonlinear systems: an adaptive approximation approach. R.M.G. Ferrari (PDF).

Italian journals

  1. Italian Acoustics Journal, Vol. 29, N. 1-2, January-June 2005. On the acoustoelastic effect in pre-stressed metal sheets: theory, experimental results and simulation via the Cell Method (Winner of the Giacomini Award 2005. An English translation can be found here).

Italian conferences

  1. 32° National Conference IAA, Ancona, June 15th-17th 2005. On the acoustoelastic effect in pre-stressed metal sheets: theory, experimental results and simulation via the cell method.
  2. Femlab Italian Multiphysics Meeting 2005, Milano, October 11th 2005. The Pianistic Touch: Femlab Modelling of a Grand Piano Action (slides in MOV format).
  3. 2006 CIRA Symposium, Milano, September 18th 2006. Fault Diagnosis of Distributed and Uncertain Non-Linear Systems. R. Ferrari, T. Parisini, M. M. Polycarpou.

Teaching

  • System Identification Lab (Univ. of Trieste, 2006-2007)
  • Process Control and Digital Control Lab (Univ. of Trieste, 2004-2005)
  • Control Systems Laboratory (TU Delft, 2016-17)

Experiences

  • Parallel Computation Summer School: Cineca, Bologna (Italy), September 2004.
  • Research about advanced measuring techniques for seamless hot pipes: Danieli Automation, from June 2004 to January 2005.
  • Development of a fuzzy control system for a composting plant: Ecomaster Projects, from June 2004 to January 2005.
  • CIRA PhD school "Antonio Ruberti", University Center of Bertinoro (Italy), July 2005.
  • 1st Hycon PhD School on Hybrid Systems, Siena (Italy), July 2005.
  • 1st CASY Special Workshop on Advances in Control Theory and Applications, University Center of Bertinoro (Italy), May 2006.
  • Visiting student at the Department of Electrical and Computer Engineering of the University of Cyprus (ECE), Cyprus, October-November, 2006.
  • 2nd International Workshop on Networked Control Systems: Tolerant to Faults, University of Calabria, Italy, November 23-24, 2006.
  • SIDRA PhD school "Antonio Ruberti", University Center of Bertinoro (Italy), July 2007.

Links

Last edit: February 18th, 2019