SibEDGE developed the system of perspective analytics for SIBUR Holding

On October 12, 2018 SibEDGE reported that it helped to develop the system of perspective analytics for the SIBUR.

Customers set the difficult task: create the project, on the basis of operation algorithms with the large volume of data, allowing to predict most precisely and effectively breakdown of the equipment at the enterprises. The condition of equipment influences not only volumes, but also on quality of products and is, in fact, a key factor in formation of marginality of products.

Within the designated problem, specialists of SibEDGE, together with colleagues from Elesi developed the system of pro-active monitoring of the equipment and also offered options of its implementation in production process.

The offered prototype is based on Big Data technology with elements of machine learning. Using the unrolled analytics of data, a system will be able to predict much more precisely repairs and failures of technology equipment, than its previous version and also will allow to optimize work of the enterprise and it is essential to reduce costs and costs.

Two modules became principal components of the solution:

• the module of search of anomalies in operation of the equipment
• the forecasting module building the forecast of breakdown of elements.

So the application carries out the fast and objective analysis of the offered data and most precisely can define time and the nature of possible breakdown on production.

In the course of work it was necessary to consider a factor of the fact that the equipment of holding is divided into several types of criticality, therefore approaches to management of reliability, analytics should differ.

The developed application will allow to process a grid of scheduled repair, based on the analysis of a condition of equipment, having made this repair the most timely. As the output - is reduced a downtime of the equipment, risks of its use decrease. Also offered system of analytics will help to optimize business processes of the SIBUR due to reduction of sudden stops on production.

The complexity in project implementation can arise at a stage of data collection as they are stored in diverse bases, some of them are not digitized and hardly accessible. The solution of this task is found in study.