Modern production facilities are equipped with large numbers of different technical installations, that are connected to each through the production processes. Usually, these installations all have their own internal control systems, that ensure optimized operations of the equipment. When looking at supply facilities in particular, they are controlled and managed without taking changing environments or the behavior of third-party installations into account. This often leads to the system as a whole to be run rather inefficiently.
A very simple example can be found in the interaction of an AC and heating system. Independently each system is managed to operate efficiently. However, in spring and autumn times, both systems may still be running and in fact counteract each other. Their operational controllers may each do a perfect job but fail to “talk to each other” and will both only react to external influences. In a production facility similar behavior can be seen with supply systems for compressed air, water, etc.
The process specialists involved may already have a “gut feeling” that their equipment is not running as it should be. In many cases they are unable to support their feeling with reliable data, that could justify any sort of intervention.
This is where WiriTec can offer a solution, by using mathematical and AI-powered applications to analyze the situation. In close cooperation with the customer all relevant plant, environmental and process parameters are identified at the start. All data correlating to the processes identified are then capture and stored in WirTtec.
In the case of our above example of compressed air, the relevant parameters for both the supply- and the demand side are recorded. This includes data such as power consumption, pressure levels, operational status of individual compressors, temperatures, etc. These data are complemented with data concerning the consumers of compressed air. These could be operational status of production facilities, individual air consumption as well as predictions of future demands, to name a few. Needless to say, that the choice of values strongly depends on the process at hand.
All the historical data are analyzed by regression, mathematical algorithms, or machine learning methods. These comprehensive analyses will yield information on the mathematical correlations between the installations – in simple terms, the cause and effect.
This detailed knowledge on the correlations between processes allows for identifying and avoiding problems and abnormalities at the different operational stages. Continuous monitoring processes can also benefit from the information on the correlations. Based on the mathematical dependencies, consumption as well as expected behaviors may be forecasted and compared to the actual data. Alarm messages can be triggered if the actual values vary too much from the forecast. This approach allows for an automated monitoring of all processes and ensures that they run smoothly.
If you have some technical installations in mind, where you have a “gut feeling”, please do get in touch. We are more than happy to analyze your processes to ensure optimal operations.