Development in the manufacturing sector is advancing by leaps and bounds and concepts such as Industry 4.0 and Smart Manufacturing are increasingly present, not only in business, but also in the policies of major developed countries. In recent years, ambitious measures and targets have been adopted but these can only be achieved by putting technology at the service of industry and by making appropriate use of advances in business intelligence and data processing in the industrial manufacturing process.
In the new era of telecommunications, industrial processes must be reformulated in order to facilitate their understanding and accumulate knowledge in all areas, from sustainable design to systems integration, including advanced process control, quality and support in decision making.
In this new paradigm, automation plays a fundamental role. Thanks to the use of data extraction and mining techniques, it has become an important pillar of industry 4.0, in a process that extends from the automation of machines to the automation of information, and then on to the automation of knowledge. Through these models, different situations can be simulated for tests and experiments on how to optimise them. This is known as a digital twin.
As a consequence, the creation of value occurs as we are able to extract relevant information from the data generated in the production process and convert it into knowledge that improves operations and makes product manufacturing more efficient. If we analyse the patterns in the process data and the relationships between the different variables, it is possible to extract useful and relevant information. From this information, mathematical, statistical and intelligent models can be developed for various applications, such as monitoring different processes, diagnosing faults, grouping by different characteristics, detecting relevant variables or attributes, predicting events, etc.
The main objective of data mining, Machine Learning and Artificial Intelligence in general is to extract useful and relevant information from the data generated during processes and convert it into effective knowledge to improve understanding and support the decision-making process.
AI’s contribution to industry is undeniable, emerging as a driver of change in all sectors and defining the technological agenda of both large and small companies. Its importance is evident from the increase in productivity, to the savings in time and production costs, or the higher quality of the products which will consequently lead to an increase in demand.
Foundries are not exempt from these advances and in their day-to-day work they need tools that allow them to make decisions based on experience and to optimize their resources and processes. The modelling of the foundry process and the management of large volumes of data transformed into knowledge allows automated learning algorithms, properly trained, to predict the value of a certain variable and obtain the most efficient result for each process.