This project aims at improving the operational efficiency of large-scale dynamic plants as found in petrochemical, chemical and other process industries. Costs related with industrial implementation of current model-based operation support systems, like Model Predictive Control, Real-Time Optimization and soft-sensors for these complex processes are currently very high. Moreover it is widely recognized that life-time performance of these systems is rather limited, particularly due to the fact that the underlying dynamic models need to be adapted/calibrated regularly, requiring expensive dedicated measurement campaigns. Given the importance of increasing demands on economic and sustainable process operation, there is a strong need to reduce the costs and increase the performance of model-based operation support systems.Therefore it is proposed to develop model-based operation support technology that enables control and model calibration at a considerable higher level of autonomy than currently possible. The technology enables autonomous maintenance by automated surveillance, continuously monitoring the process' and operation support system performance. If performance degradation is anticipated then proactive adaptation of the model-based operation support system is initiated. The technology will operate on the basis of the least costly principle. The influence of the invasive testing on process' operation and economics will be minimised, to the extent possible given the necessary accuracy of the resulting identified model. Moreover all decisions are based on an economic trade-off between process operation costs and benefits. This system should be able to optimise plant performance under varying operational conditions and adapting to changing circumstances. The consortium is well suited for this task. Industry is specifically involved to guide research in an industrial relevant direction and to ensure industrial validation of the technology to be developed.