The Laboratory for Stochastic Mechanical Systems & Automation (SMSA) is an international leader on innovative methods for stochastic mechanical, structural, and aeronautical systems, with emphasis on Statistical Time Series methods for stochastic identification and fault diagnosis / Structural Health Monitoring (SHM).

waveformhealthy vs damaged

Our challenge: Extract the maximum information from random signals and data – use the information for analysis, prediction, decision making, diagnostics, control!

Our experience: 30 years of teaching and research excellence on stochastic identification, fault diagnosis / SHM, data analysis. Cooperation with major industrial companies, research institutes, and universities across Europe, the US, and the world.

 

What we can do – sample GENERIC capabilities: 

  • Non-parametric data & random signal analysis
  • Parametric data & random signal analysis (regression, ARMA models, state-space models,…)
  • Random mechanical vibration analysis
  • Statistical Time Series analysis
  • Signal Processing: Time domain signal analysis
  • Signal Processing: Frequency domain signal analysis – Spectral Estimation
  • Data based modeling of stochastic dynamical systems (System Identification)
  • Data based prediction
  • Data based diagnostics
  • Statistical Time Series methods for Structural Health Monitoring (SHM)
  • Precise damage localization
  • Non-Linear signal and system data-based modeling (identification)
  • Non-Stationary signal and system data based modeling
  • Data-based modeling of systems under various operating conditions

Sample APPLICATION areas of our work:

  • Automotive systems & diagnostics
  • Railway vibration & diagnostics / SHM
  • Experimental structural dynamics – Modal Analysis
  • Analysis of railway reliability and availability data
  • In-flight aircraft vibration analysis & diagnostics / SHM
  • Bridge non-stationary vibration & SHM – bridges with passing train
  • Structures under earthquake excitation – identification & SHM
  • Random vibration based SHM for mechanical structures (trusses, beams, wind turbines, rotating machinery, robotics,….)
  • Identification of mechanical systems & mechanisms with friction
  • Aircraft fault detection systems & avionics
  • Aircraft air traffic systems – stochastic collision avoidance – guidance systems
  • Production Quality Control (automotive assembly,….)
  • Biomedical signal analysis (neuro signals, human motion signals,…)
  • Environmental signal & data analysis

 

 

Our key innovative contributions include:

  • innovative methods for high-fidelity stochastic identification of structural and mechanical systems,
  • innovative non-stationary random vibration data-based modeling/identification, including various types of Time-dependent AutoRegressive Moving Average (TARMA) modeling,
  • innovative statistical time series methods for SHM,
  • innovative, high-fidelity methods for the identification of non-linear stochastic mechanical systems, including systems with friction, systems operating under varying pseudo-static conditions, Wiener-Hammerstein-type systems
  • design of innovative aircraft systems, including avionics, safety, and air traffic systems.

Current thrust research questions & areas include:

  • Is it possible to effectively detect damage in complex and large structures using ambient vibration signals?
  • Is it possible to effectively detect damage under varying environmental and operating conditions?
  • Is it possible to precisely localize damage using few and remote vibration sensors?
  • How many sensors are minimally required for the above?
  • Is it possible to organize an effective SHM system for a number of like (e.g. aircraft, wind turbines, etc) structures?
  • Is it possible to obtain an improved dynamical model (of systems and structures) using multiple data records?
  • Is it possible to build data-based dynamical models from data under various operating conditions?
  • Is it possible to obtain highly compact, high fidelity, data-based non-stationary vibration models? (applications in bridges, robotics, rotating machinery, etc)
  • Is it possible to organize effective diagnostic strategies for non-stationary (time-varying) structures and systems?
  • How accurately may the Remaining Useful Life (RUL) be predicted for certain components?
  • How to best organize a Health Management System (HMS) on an aircraft?
  • How to best predict undesirable situations in air traffic and other systems?

 

More …. The SMSA Laboratory builds upon a tradition of over 30 years of academic excellence in research and teaching and fertile collaboration with industry in Europe and the US. It maintains a strong publication program with over 270 technical articles published in major international journals and conference proceedings, articles published in major technical encyclopedias and books, plus a multitude of additional technical reports. The Laboratory involves faculty members, post docs, PhD candidates, other researchers, as well as numerous M.Sc. students. It features state-of-the-art computational and experimental facilities, and is also active in the organization of international short courses, technical conferences, thematic issues for major international journals. It maintains active collaborations with research and industrial partners around the globe, and its research is supported by national and international organizations, including the European Commission.

 

Interested in our work? You may find out more by contacting us.