Industry is a special branch of human activity, due to the dynamics of changes in the environment in which it functions, as well as what it generates. No wonder that companies wanting to survive in the market support a variety of IT systems. One of the most popular solutions of this type is Business Intelligence systems, which for many years now have constituted the core of IT systems related to business decision-making, especially within large organizations, based on data that the company already possesses. It is often said that we are now living in the era of the IT revolution, of which data is the key ingredient. This means that smaller organizations also need to learn how to use this data.
In today’s world, organizations cannot afford delay or downtime, which presents a big challenge for highly complex IT systems. Business analytics is no exception. Not only do BI systems have to deliver the right data in the right format, but such data has to be delivered almost immediately. Hence the concept of real-time analytics, or reporting for analytical needs in real time. Modern BI systems enable reporting from transaction systems (such as ERP, CRM or MES systems) without interfering with their work and without the need to create copies of data for analytical purposes.
In traditional BI solutions, data must first be processed and adapted using the ETL process (i.e. the process of extraction, transformation and loading) before it is included in the report. This process can take anywhere between a few minutes and several hours. For business processes that require an almost immediate response, this will not be satisfactory. The solution to this problem may be the implementation of the operational analytics mechanism, meaning special indexing structures which operate directly within the database and interact with the structures existing within, which are responsible for operational activities. These indexes are an interesting solution because they can rapidly aggregate huge amounts of data. The diagram below shows how they operate:

OLTP workload is the basic operations of a given system, e.g. accepting orders, saving data from the technological process, etc. Innovation is based on the fact that there is a special column index which operates on the same structures created as part of the basic activity of the system, which allows for the generation of various types of analytical statements, even in the case of very large volumes of data. So this technology allows data to be analyzed without the necessity of loading it into a central data repository, i.e. a data warehouse. From the user’s point of view, it is nothing more than an immediate view of the production data, without the need to wait a long time for reports.
The concept of Intelligent Enterprise and Industry 4.0 is growing ever more popular. In this area we may find the specific application of the concept of the Internet of Things. In this model of enterprise operation, data is everywhere and is continuously generated by even the smallest devices. Such data takes the form of a data stream that flows in a continuous manner, and so analyzing it with the help of traditional relational databases can be very difficult or even impossible. Therefore it is necessary to apply technologies that enable data capture, analysis and drawing conclusions. Both the cloud – Big Data – as well as AI and machine learning technologies are used for this purpose. They allow users not only to describe the current position of the organization in an intelligent way, but also to capture non-standard behavior, and even to predict the future based on historical data and predictive analysis.
An example of the implementation of these technologies is the case of Schneider Electric, a giant in the oil and gas industry. Thanks to a combination of IoT technology and machine learning, the company has achieved previously unattainable capabilities in monitoring machine operation. Thanks to the implementation of machine learning mechanisms, it is now possible to react proactively to non-standard events which signify the danger of machine failure. In other words, the fault can be removed long before it causes a machine failure. The company used Microsoft cloud components called Azure Machine Learning and Azure IoT Edge for this purpose. The diagram below shows how the system works:

Azure Machine Learning, as used in this case, builds predictive models and informs users of potential threats depending on the situation. The benefits of the above solution are obvious – proactive monitoring not only allows for savings related to the avoidance of failures that have not yet occurred, but also allows users to ensure the safety of those operating the machine, which is very important in industries such as the mining industry.
In addition to the challenge of efficient data storage and analysis, visualization is also very important. Static data transfers are not as attractive as they used to be. Users expect interaction, i.e. the ability to work with the report independently. Also, in this case, the choice of the right tools is quite wide-ranging, and the opportunities they offer are increasing. A great example here is Power BI, which offers a whole range of reporting options, from the simplest tables to the most sophisticated visualizations of maps and charts. This tool is so comprehensive that it enables the development of hybrid solutions based both on cloud technology and on the servers that the company already has.
There are numerous examples of the successful implementation of this type of tool, for instance KPMG, which uses Power BI to analyze business processes. The following are examples of reports that the company creates for this purpose:


KPMG used transaction data from ERP systems. The purpose of implementation was to detect processes that may have a negative influence on the performance of other processes, among others. In addition, the reports were used to visualize certain kinds of deviation and non-standard actions which were not detectable outside of this report, and thus were not visible to the decision-makers. Thanks to the Power BI solution, KPMG was able to significantly improve or discover links between processes that had not previously been noticed and continue to improve them, monitoring progress and reducing costs.
The advantage of each of the above implementations is the fact that they operate on the same data platform (the Microsoft platform), which means they integrate well and offer the opportunity to build solutions for various recipients – from small implementations for medium-sized enterprises, to huge investments by international corporations. Is Business Intelligence attractive to industry and other branches of the economy? I think so – especially since reporting and building data structures is an inseparable element of contemporary business. In addition, if we already have orderly and proven reporting methods in the organization, it’s possible to develop them by adding further elements, such as machine learning or IoT. There are many advantages of implementing this type of solution: from the most obvious, such as savings and the ability to make the right decisions, to strategic ones, such as achieving a competitive advantage, which is so difficult in today’s business environment.
Would you like to learn more about the possibilities of cooperation? Do you have a question? Write to us!
Comment