Innovative Modeling Approaches for Production Systems to raise validatable efficiency

Innovative Modeling Approaches for Production Systems to raise validatable efficiency

The rise of the system complexity, the rapid changing of consumers demand require the European industry to produce more customized products with a better use of resources.

The main objective of IMPROVE is to create a virtual Factory of the Future, which provides services for user support, especially on optimization and monitoring. By monitoring anomalous behaviour will be detected before it leads to a breakdown. Thereby, anomalous behaviour is detected automatically by comparing sensor observation with an automatically generated model, learned out of observations. Learned models will be complemented with expert knowledge because models cannot learn completely. This will ensure and establish a cheap and accurate model creation instead of manual modelling.

Optimization will be performed and results will be verified through simulations. Therefore, the operator has a broad decision basis as well as a suggestion of a DSS (Decision Support System), which will improve the manufacturing system. Operator interaction will be done by a new developed HMI (Human Machine Interface) providing the huge amount of data in a reliable manner. To reach this aim, every step of the research process is covered by a minimum of two experienced consortium partners, who conclude the results of the project using four demonstrators.

The basis for IMPROVE are industrial use-cases, which are transferable to various industrial sectors. Main challenges are reducing ramp-up phases, optimizing production plants to increase the cost-efficiency, reducing time to production with condition monitoring techniques and optimise supply chains including holistic data. Consequently, the resource consumption, especially the energy consumption in manufacturing activities, can be reduced. The optimized plants and supply chains enhance the productivity of the manufacturing during different phases of production. Furthermore, the industrial competitiveness and sustainability in EU will be strengthened.

More information
    Comment: Unpredicted breakdowns of a plant are often very costly as production stops. The fault has to be identified and spare parts have to be ordered and installed before the production can continue. Condition monitoring can be used to detect wear and therefore schedule maintenance before a complete failure. In case of breakdown a root cause analysis can aid the experts in finding the error cause. Both of these approaches can decrease downtimes significantly and decrease costs.
    Comment: Data from complex production plants is acquired and used to learn models of machines and production processes which are then used for simulation, optimization and diagnosis tasks.
      Comment: Modern communication technologies like OPC-UA are used to acquire data from the controls. Data from distributed controls must be synchronized for complex plant models. Developed model learning strategies can be implemented within the controls themselves when the circumstances allow this.
      Comment: Models are learned from process data and extended with expert knowledge which are then used to detect wear, predict maintenance windows and find root causes for errors during production.
      Comment: Normal behaviour models of the plants are learned durign operation. These models can then be used to monitor the condition of the machine. New situations not covered by the originally learned model can be added through adaptive learning.
      Optimization algorithms can find sub-optimal configurations of the plant and improve them. Newly calculated system parameters are verified in a simulation before they are applied to the real plant.
      Comment: The optimization can optimize different parameters/aspects of the plant, including but not limited to energy consumptions.
    Comment: ICT solutions enable the work in IMPROVE. Big Data is collected from the production plants and modern machine learning techniques are used to learn models from data, drastically reducing the manual modelling efforts. Learned models can adapt to the modern, fast changing systems and are available for tasks like condition monitoring and root cause analysis. The results are then displayed to the user, hiding the systems complexity.
      Comment: One option is to store the data generated by sending data to a cloud based system which offers services for data analysis and model learning.

      Optimization algorithms can calculate better plant configurations. Thesting these new configurations in a real plant can be very cost intensive as production may be compromised during configuration and testing. A virtual environment which is able to simulate new parameters and verify them is a big deal as the running production is not compromised during testing and evaluation of new parameters.

      The main focus of IMPROVE are learned models. Manual modeling of system models is not suitable for the complex, fast chaning industrial plants we have today. Lots of expert knowlege is needed to manually create a model. Learned models can be created using only data and little to no expert knowledge is required depending on the technology.