Research at Queen's is concerned with both principles and applications, the ethos being to link theoretical and technological advances with practical requirements for intelligent systems and advanced control.

The creation of the £4.5M Virtual Engineering Centre with Mechanical, Aeronautical and Civil Engineering, pioneered by Intelligent Systems and Control, has enabled important infrastructure investment in a state-of-the-art Virtual Reality theatre, an advanced graphics computing cluster and new experimental facilities. Importantly, a number of new interdisciplinary research collaborations have emerged including soft sensors for automotive applications, condition monitoring of internal combustion engines, advanced visualization for laparoscopic surgery, augmented reality for 3D urban information, with links to ECIT, the School’s Institute of Electronics Communications and Information Technology in multi-modal internet access and distributed virtual environments.

 Important research achievements made in the recent past include:

• Original contributions to Multivariate Statistical Process Control (MSPC) theory for nonlinear, dynamical process monitoring (subspace identification, local statistics, nonlinearity measures, nonlinear fault isolation and adaptive monitoring). This has underpinned pioneering results on automotive fault diagnostics for Volkswagen and Nissan engines.
• New algorithms and tools for on-line identification, construction and training of neural and semi-physical nonlinear dynamical models (subset selection avoiding OLS matrix decomposition to promote efficient forward/backward methods, significantly reduced memory usage and computation in RBF network optimisation, velocity-based multiple-model networks.
• Recognised (IEEE Control Systems Magazine) commercial MSPC monitoring software tools emerging from longstanding collaboration with DuPont (now Invista) and Annex6, a spin-off company from the cluster.
• Proposal of the ‘eng-genes’ neural network paradigm for grey-box modelling of nonlinear dynamical systems.
• Design, implementation and testing of new neural model-based grinding controllers for aluminium substrates in collaboration with Seagate; a new semi-physical modelling approach for predicting NOx emissions from coal-fired generation plants in collaboration with BCURA, and a novel autonomous marine vehicle's controller.

• New algorithms, along with associated software and hardware tools, for advanced visualization, low-cost virtual reality and haptic training aids for laparoscopic surgery (www.ee.qub.ac.uk/graphics).

• Robust, efficient feature-matching algorithms for real time assessment of the performance of airport lighting patterns (in collaboration with Flight Precision) and for road lighting (NI Department of Environment).
• Original research results on ad-hoc wireless networks for feedback control using variable rate sampling, along with the only verified co-simulation tool for the IEEE 801.11 protocol available in Simulink for control design.
• EPSRC funded UK-China Science Bridge strategic alliance between Queen’s University Belfast and leading universities and industrial partners in both China and the UK.

Our immediate research plans for fundamental research in Intelligent Systems and Control are as follows:

• Research on QoS-based sampling for wireless networked control has raised significant theoretical design challenges, notably in respect of stability. A new EPSRC collaboration with Sheffield and Airbus will further this work.
• Future focus on fault detection and diagnosis for internal combustion engines will concentrate on tailpipe emissions, with General Motors, USA providing precisely pre-aged catalytic converters.
• Investigating the recent statistical ridge method for nonlinear system modelling, overcoming some disadvantages of suboptimal subset selection methods. The ‘eng-genes’ concept will be extended to enhance its on-line data handling capabilities, prior to investigating control strategies.
• New research in emerging areas including Systems Biology (gene therapy) and Bioinformatics (microarray data processing, with Medicine), Environmental (pollution sampling and modelling - with Environment Agency). International collaborations, through sabbatical visits from China and the EPSRC Control Network (Queen’s, Sheffield, Imperial, Loughborough, Liverpool), will contribute expertise.
• Research with the Royal Belfast Hospital on a new haptic surgical simulator for medical training will continue and with University of Rennes on low cost virtual reality. In lighting quality assessment/computer vision, the focus will be on the correspondence problem.
• Research on guidance and control of group of unmanned aerial vehicles.
• Investigation of automatic collision avoidance strategies for unmanned aerial and marine vehicles based on rules defined by international organisations for manned craft.