Intelligent Maintenance and Operations Systems (IMOS)

Our research focuses on developing intelligent algorithms to improve performance, reliability and availability of complex industrial assets and making the maintenance more cost efficient. 

Intelligent Maintenance & Operations Systems

Our research  focuses on developing intelligent algorithms to improve performance, reliability and availability of complex industrial assets and making the maintenance more cost efficient.

The amount of measured and collected condition monitoring data for complex industrial assets has been recently increasing significantly due to falling costs, improved technology, and increased reliability of sensors and data transmission.

The measured condition monitoring signals of complex industrial assets are typically high dimensional, highly redundant, have several interdependencies and prevalent non-​linear relationships. The diversity of the fault types and operating conditions makes it often impossible to extract and learn the fault patterns of all the possible fault types affecting a system and to develop model-​based approaches. Even collecting a representative dataset with all possible operating conditions can be a challenging task (depending on the variability of the operating regimes of the assets) and may delay the implementation of data-​driven fault detection systems.


This mission involves tackling the following challenges:

  • Decision support: Decision making for effective maintenance of complex systems and fleets of systems is complex and requires an integration of several sources of information with different degrees of uncertainty :
    • The current health condition of each system.
    • Predicted evolution of the system state condition including the uncertainty of the performed predictions.
    • The scheduled maintenance plan.
    • Scheduled production or operation plan.
    • Anticipated future operating conditions.
    • Costs of maintenance resources and unavailability.
    • Restrictions of resource availability.
    • The system configuration and decision alternatives.


  • Paper at CVPR:
    The paper entitled ‘DARE-GRAM: Unsupervised Domain Adaptation Regression by Aligning Inverse Gram Matrices.’ This work was a collaboration between our Ph.D. student Ismail Nejjar, Dr. Qin Wang, and Prof.Dr. Olga Fink. We proposed a novel domain adaptation for regression method that aligns the inverse Gram matrix of the features, which was motivated by its presence in the OLS solution and the Gram matrix’s ability to capture feature correlations. Our method leverages the pseudo-inverse low-rank property to align the scale and angle in a selected subspace generated by the pseudo-inverse Gram matrix of the two domains. Experimental results on several benchmark datasets demonstrate that our method outperforms state-of-the-art methods.
    The paper is available open access:
  • New Ph.D. student:
    Warmly welcomes our new Ph.D. student, Keivan Faghih!
  • Paper in Reliability Engineering & System Safety:
    Are you looking for a solution that jointly optimizes revenue, reliability, and safety in the electricity market? Our team’s paper ‘Safe multi-agent deep reinforcement learning for joint bidding and maintenance scheduling of generation units’ . This work was a collaboration between our postdoc Pegah Rokhforoz, visiting Ph.D. mina montazeri, and Prof.Dr. Olga Fink.
    Don’t miss out on this research, check out the full paper now! Direct link to the paper: