26-27 March
2019
Danilovsky Event hall
Moscow
Organizer
RU
30.01.2018
Maxim Lipatov, JSC ROTEC. Predictive analytics in industries

The Internet of Things gave the world a lot of new opportunities. One of the main opportunities was the possibility to organize and to collect continuous data from sensors and other devices. Thanks to modern "smart" solutions, it became possible to collect large amounts of data about various processes in real time.

According to Transparency Market Research, the market of predictive analytics will reach 6.5 billion dollars by 2019, while it was only 2.8 billion in 2012. The world market of systems for predictive analytics will grow by an average of 17.8% per year. As practice shows, the companies that continue to invest in technology and innovation will survive in difficult economic periods. And of course, predictive analytics is one of them.

The mode of operations for predictive analytics (Source: Forrester Research, 2013)

The predictive analytics is particularly important for industrial enterprises, where processing and understanding of a huge amount of data is necessary and there are high risks in making decisions. The process flow data is not always being used effectively, while it can be used to optimize operational processes and improve the technical - economic indicators of production. Optimization can be performed at any manufacturing process with a serious level of automation, organized collection and long-term storage of information. The intelligent systems are successfully used for this purpose that can analyze the state of technological process in real time, predict the further course of the process, determine the level of optimality and, if necessary, change the control parameters or give recommendations to an operator. The predictive mathematical model of technological process is being created to solve these tasks with the help of machine learning tools. It analyzes the input data, provides real-time predictive information of the process and suggestions for its optimization.

Maxim Lipatov, a technical director of Russian company for telecommunications, television and radio broadcasting equipment production (ROTEC), said: "The actual implementation of digital control systems in production has already sufficiently allowed to increase efficiency at production facilities. Today we are a little further, we began to apply the predictive analytics system to diagnose and identify at the earliest stages the equipment deviations. Actually, our main task that we tried to introduce is the technology implementation to reduce production and repair costs”.

Significant costs at manufacturing site come to eliminate the consequences of accidents and repair work lines. It is ideal in our time to switch from budgeting and transition from emergency repairs to real condition-based repair.

Please  fill this form to view the full version of Maxim Lipatov's speech at IoT World Summit Russia 2017.

The main advantage is the early deviation detection in technical condition. It is based on a single integral criterion. So, there is no need to evaluate different trends, to compare obvious or unobvious facts in equipment conditions. It is enough to focus on one integral criterion. An integral criterion can be reduced both for a group of objects and for one object.

Why it is important to use predictive analytics

Maxim Lipatov

Technical Director JSC “ROTEC”

In general we have been approached by insurance companies and operating organizations with a request to evaluate already accomplished event in order to see and estimate how the methods of predicative analysis give an effect in a real assessment.

Let's take one case. There was some damage in a moving part of the gas turbine, damage in the bearing supports. In this situation operating personnel spent two weeks, the gas turbine was not stopped and as a result of the turbine operation in such condition,  470 million rubles were spent on repairs. If the predictive analysis was applied in this situation, it would be possible to shorten the repair period by three times at least and the repairing costs not less than by the half.

Another interesting case is the archival data damage analysis of the flow range of turbine. In this case the need to repair the main unit was clearly visible7 days before the accident. If it was planned in time and the machine was stopped for a visual assessment of the flow range technical condition, it was possible to avoid damage of more than 1 billion rubles.

Latest opportunities in the IoT and Big Data sector, together with advanced methods of predictive analytics, are becoming an effective tool for reducing costs, improving product quality and increasing enterprise productivity. Predictive analytics has become a new trend in today’s world, which opens wide prospects for the companies in further development. Besides industrial and energy complex, the predictive platforms are successfully used in banking, insurance, retail, logistics, marketing and many other areas.

You can learn about the implementation of modern technologies for analytics, prediction and planning of energy complexes at the second Smart Energy Summit Russia, which will be held on March 27-28, 2018 in Moscow. To find out about summit details and get a full list of speakers, fill up, please this form.



Other news
right-arrow left-arrow facebook-logo youtube instagram