+49 231 97 00 340 contact@divis-gmbh.com
We will be happy to advise you concerning the details of a thorough process analysis including all aspects from sensor technology to the optimized product or process. We start with a consultation meeting. Our final goal is the implementation of a customized system into your internal IT-environment. Of course, the subsequent support by our specialists is self-evident.
The following services are part of the project realization:
Upon request, we develop interfaces to your data bases. That ensures an intelligent data storage for analyses, modeling and optimizations:
1. Generate Loader
2. Configuration of data base server
3. Server installation
4. Network connection
5. Install clients
Industry 4.0 is a term for the fourth industrial revolution, facilitated by the combination of process-wide data collection and data driven machine learning methods. Digitalization and the intelligent connection of machines and process states allow for many new opportunities. The aim is not to optimize just a single component of the plant, but the whole value chain. Thus, many different aspects have to be taken into account.
Machine Learning is a class of methods which are able to derive relations and patterns from data and observations. These relations can be made available as predictive models or explicit knowledge about process relations. The aim of machine learning is to identify such relations and derive predictions.
Automatic ML is an advanced method where the learning algorithms for each data set and task are chosen and optimized automatically. Our ClearVu Analytics product family provides these functionalities.
By means of incremental learning algorithms the models are able to recognize and self-learn process variations (changes of machines or raw materials) and product adjustments.
Machine Learning is a powerful tool for the automatic analysis of data and for a continuous production control.
The technology can be applied in any business process, including R&D, logistics, production, sales, purchasing and production planning.
Predictive maintenance helps to predict the remaining life time of machine components. Thus, the replacement is only necessary when it becomes evident that a defect will become inevitable. The alarm reporting the imminent defect is signaled early enough to adjust the maintenance schedule according to the plant’s operating schedule so that downtimes and delivery dates can be taken into account. It also becomes possible to optimize the maintenance schedule based on remaining useful lifetime predictions.
Opposed to this, there are the most common maintenance strategies—the reactive and the preventive maintenance. In case of reactive maintenance, the components are not replaced until a defect occurs. However, the consequence could be a spontaneous and long downtime at an unfavorable moment. In case of preventive maintenance, the components are replaced in predetermined intervals, no matter whether there is a defect or not. The advantage is that you can schedule the maintenance precisely, however, at the cost of wasting money because the whole life time of a component is not exploited.
Our methods for data analytics facilitate a continuous quality optimization. That means that at any time all process steps can be analyzed based on data. This approach guarantees a smooth production process. By applying predictive models in the production process, machine learning predicts the quality of the final product early on, including whether all objectives will be reached, or deviations will occur. By combining data-driven machine learning and optimization algorithms, process parameters can be optimized dynamically for minimizing process deviations.
To receive relevant process data it is necessary to install suitable sensors at the right places of the process chain to record data and transfer them to a data warehouse. There, all relevant data are collected to conduct the needed analytics and optimizations. Interfaces to your internal IT and process environment can be implemented. The results of data analytics and optimization are then available at all important process stages.
Based on your process data we conduct targeted optimizations of products and processes. The optimization criteria can be very different, e.g.:
– Quality characteristics of the end product (color, dimensions, surface properties)
– Energy efficiency or yield of the process
If conflicting objectives need to be optimized, e.g. yield maximization and energy efficiency, the best compromise that considers both targets will be found by optimization.
With our powerful software ClearVu Analytics you can analyze process data as well as other types of data. You can generate models which identify the most important process parameters and their effect on the end product. Moreover, the models help you predict future process conditions. If the predicted conditions do not meet your requirements, the process and product parameters can be optimized according to the target values. This is also possible if the target values are in conflict with each other, for example, low costs and high quality. In such a case, ClearVu is able to find the best compromise. Since ClearVu generates all models automatically, and also identifies the best solution, the user does not need to have any knowledge of statistical modelling.
We develop customized systems according to your individual requirements. This is done by adapting the modules to company specific requirements and designing the user interface according to your needs. Furthermore, we connect our software to your internal data base. The final system will be integrated into your internal IT-environment and the existing production process. Additionally, we provide a 24/7 support, to help you in case of questions and any other needs.
Contact details
divis intelligent solutions GmbH
Joseph-von-Fraunhofer-Str. 20
44227 Dortmund, Germany
+49 231 97 00 340
contact(at)divis-gmbh.de
Newsletter
Our free newsletter informs you about news and dates.
Here you can subscribe to our newsletter: