Creating the Factory of the Future with 4.0 Solutions
Facing increased global competition, the manufacturing industry depends on high-level solutions to ensure excellent machine functionality. Current analyses estimate that system downtimes and component breakdowns lead to an energy waste of 33% in the production and a significant loss of profits. At the same time, the complexity of production plants is steadily rising due to increasing product variances, product complexity, and pressure for production efficiency. Production systems must therefore evolve rapidly and operate optimally, which creates challenges for larger industries and especially for small and medium-sized enterprises (SMEs).
To meet these challenges, the European research and innovation project IMPROVE joins forces from 13 leading players in the field of academia, industry, and software development from Europe and beyond. Together, they have developed novel data-based solutions to enhance machine reliability and efficiency. Innovative tools in the fields of simulation & optimization, condition monitoring, alarm management, and quality prediction provide manufacturers with a human machine interface (HMI) and decision support system (DSS) to ensure best possible user support.
IMPROVE’s solutions can be standardised, commercialised, made accessible and applicable for European SMEs by tackling the problem of user support functions in terms of self-diagnosis and self-optimisation. Alternatively to relying on human expertise and engineering to formulate necessary knowledge, data-driven models are used for self-diagnosis and optimisation of production plants. New self-learning tools extend capacities of manual creation, making it possible to learn accurate virtual factory models of complex, large, and distributed plants through use of real time analytics.
By ensuring an efficient and reliable manufacturing process, IMPROVE contributes to reducing energy consumption and making manufacturing more environmentally friendly. The project is funded by the European Union’s Horizon 2020 programme with an amount of €4.2 million.
In a nutshell
The following figure illustrates the IMPROVE process to enhance the manufacturing process.
Working structure to create the factory of the future
The following figure depicts the IMPROVE project strategy and interactions of elements for creation of the virtual factory of the future (vFOF). Click to learn more about the individual work packages.
Synchronized Data Acquisition and Management
Information from physical systems such as factory machine sensor data, will need to be captured and integrated. A defined method based mainly on existing standards, will be performed for distributed and time-synchronized data acquisition, data pre-processing and data management. Security and privacy aspects will also be considered. From these inputs, corresponding software features in controllers and other computation platforms will be developed.
Existing machine learning methods will be applied, extended, developed and verified to subsequently formulate capabilities of the virtual factory of the future (vFOF). As a result, verified suitable machine learning algorithms are published and corresponding software components are available for commercialization and standardization.
Learned models of normal behaviour generate the baseline framework necessary for further capability development. These can be used for example in comparative analysis of model simulations with physical system observations.
Causality models are integral, as they are required to identify the root cause once a system anomaly is detected. Such causality models cannot be learned completely based on machine data but require additional inputs such as expert knowledge.
Learned models are extended with expert knowledge such as required for development of causality models. This knowledge will be collected from novel Human Machine Interfaces (HMIs) and from earlier engineering phases.
Intelligent optimization algorithms can determine optimal plant parameters by simulating and evaluating different parameter configurations before the configuration is tried in the real plant. Heuristic search algorithms are used since brute force methods are not suitable for large distributed automation systems. Self-optimization can be implemented which identifies and amends criteria such as resource consumption, attribution or product throughput.
Condition monitoring of the manufacturing system uses simulations of the learned behavior models in comparison to the physical factory. Requirements are defined for typical fault situations, allowing anomaly detection, predictive maintenance and typical suboptimal condition monitoring such as energy consumption. These components are devised for use in distributed factory systems and knowledge collected from operators will generate results suitable for industries.
Deviations identified in condition monitoring are classified as critical or non-critical. Non-critical anomalies such as wear invoke the predictive maintenance algorithm to forecast the further behaviour to decide the point of needed maintenance. For critical anomalies, causal models are used to compute the root cause behind the anomalies and support the operator in the repair process.
The Human Machine Interfaces (HMI) with its underlying decision support system (DSS) will present relevant information to users in a suitable, context and role-aware way. A challenge for the HMI is to capture user inputs using methods suited for a production scenario. The close interaction between the user and the DSS assures human centred automation. As a result, a verified HMI concept will be published which can be commercialized and standardized.
The virtual factory of the future (vFOF) concept will be implemented in demonstrators and prototypes. In the early project phases, lab demonstrators will be used to verify results as preliminary to industrial plants. An important step for all IMPROVE partners is the quantitative verification of results. These are important attributes for exploitation and help ensure market acceptance.