TAdviser Interview: CDO of Sibur Alexey Agapkin - about robots, IoT, analytics and other technologies on the way to digital chemistry

Alexey Agapkin: Here it is necessary to draw a line between digitalization and digital transformation. Digitalization is use in the company of digital instruments which result in economic effects, work performance improvement, maybe, even to emergence of new professions, operational improvements. And digital transformation is a change of business models: when we begin to earn in a different way money, to rebuild internal processes using digital instruments.

Difference between "digit" and automation, informatization also in determinations. Automation is a base, process to which decades. Without it digital solutions cannot just be implemented. Digitalization came when a number of key technologies at the same time showed a maturity. This logical continuation of development of technologies.

In digital transformation the big part is assigned to culture. Digital culture means dynamism, flexible methodologies. The petrochemical industry is rather conservative production environment, at the same time our corporate culture was always aimed at permanent development. On this base it is possible to implement changes which are required by innovations and digital chemistry. In Sibur we try to provide the maximum involvement of employees into changes. Pilot digital initiatives become on flexible methodologies – design session, collecting of the user stories, dense interaction of a digital-command and customers, a regular feedback.

Alexey Agapkin: There is no uniform, general determination of "the digital enterprise". I can tell about stages which we passed, and about the next plans. In 2017 we tried new technologies on different platforms, stirred interest in them in the company. Under leadership of IT together with other divisions tested a set of technologies: drones, technical sight, advanced analytics. Obtained a number of notable results and the evidence that it is possible to work with these technologies further. For example, drones fly over the Amur GPP and report how many whether earthwork it was carried out and they in the necessary direction dug a trench. The same concerns advisers in advanced analytics: we understood that we are capable to unload data from a MES system and to create the prompters facilitating work of operators on their basis.

At the same time, even to order works on development and deployment of similar solutions on the party, qualification is necessary – competences of the customer are very important. We invited the third-party companies on processing of Big Data. Received results it seems "the model works, but continued support from supplier company is necessary". Or – the model works, but for it it is necessary to purchase expensive software. Or – the model works, but mathematical dependences in it unclear in the next analysis. As a result in 2018 we began to acquire own competences as the customer, implementing own projects.

We employed enough specialists in advanced analytics who developed own products in the company and created the corporate standard. According to it our analysts work and our contractors will work.

For 2018 we developed a number of tools which are clear for the company, and held their testing. Generally at the pilot enterprises – Tobolsk and Kstovo. Now we are ready to more widely to implement them at the different enterprises.

Alexey Agapkin: In the field of production there are three technology directions and, respectively, three types of products: advanced analytics, solutions for process optimization using mobile applications, for example, TOIR, and in the field of the industry 4.0 – IoT, augmented reality. In each of the directions there are several solutions which we are going to implement within all company.

In the field of mobile applications it, for example, the application for mobile bypasses. It will help to get rid of a pile of paper magazines, need to long look for information on work of the previous shifts, allows to collect quickly data on the equipment near installations, shows a personnel safety notification, allows to communicate quickly with the shift supervisor, to announce defects using photo, video fixings.

One more application – for obtaining the digital work permit – already now has wide functionality: suggests to select parameters of works, to record schemes of an object, to tighten actions and possible risks, to select responsible, far off to approve admissions from the mobile device. Formation of admission using the application instead of 3-4 hours takes several minutes.

There are also tasks in the field of logistics which need to be solved. Here we develop a big product – the digital double of railway logistics who is brought into commercial operation. It should provide expense reduction on repairs, reduction of excess number of the rolling stock due to digitalization of management of shipments and planning of transactions, management of the available park of cars, management of repairs of the rolling stock. These are three tool kits. One of them is already brought into commercial operation, other two into stages of trial operation. And, probably, development of similar products never comes to an end, improvements are developed constantly.

Alexey Agapkin: On the way a large number of advisers in the direction advanced analytics. In this area about at us 15 projects. One of sign stories here – use of the digital solution for forecasting of the prices and quotations for our products. For example, we developed the forecast model allowing quicker and more precisely with involvement of bigger number of sources of information for two products to receive forecasts of the prices. Next year it is going to be extended to other products.

One more perspective subject – predictive hardware maintenance. For example, for an extruder on production of polypropylene we created the corresponding model which in 85% of cases guesses that the equipment will stop soon, and it is necessary to take some measures. At the expense of it we reduced extruder stops: in their 2017 was 19, after start of model – any. Now we create the platform solution based on a standard product and the remote center of diagnostics.

From the projects started recently – system deployment of RTO (Real Time Optimization) based on the vendor solution which requires big setup. It will integrate in itself three big classes of tasks. In particular – kinetic models which were always strong in petrochemistry, use of data on supply chain management, and the third – smart modeling and transfer of a control signal for the APC system. It is an auto pilot which carries out diagnostics of the equipment online, considering food prices and production volumes necessary to us. Thus, it manages valves, latches, engines of the enterprise.

It is the big project aimed at decrease in number of errors, increase in a bezlyudnost of production, labor productivity. It should begin in 2019 too.

Alexey Agapkin: Our robots are tested, for example, for a rastarka of raw materials. At the chemical companies the problem of increase in efficiency of process of a rastarka was always: products arrive in bags, their opening and oversleeping of contents took a lot of time. At the same time 5-7 years ago robots who could replace people in this work did not pay off. But over time the equipment becomes cheaper and is improved, such cases become paid back and, as a result, more interesting.

For other processes, I think, there is no solution where it would be possible to robotize all chain. However robots can be built in processes. We had laboratory experiments with robots. – the supplier of robots brought our metering equipment for testing of mechanical properties of polymers which was always serviced by people to the vendor platform. It is process where the laboratory assistant prepares material for tests, observes, monitors, makes measurements, writes data in the magazine. The robot can do all these procedures. We tried, and at us it turned out. Now there is a task to purchase production lot of these robots.

At us robots stackers of finished goods at the rubber plants are used long ago, they demand upgrade. For example, we used technical sight and now they can define what product they stack, and on one line it is possible to start not one, as usual, and two types of products.

Alexey Agapkin: It is impossible to meet the same budget. Experiments, expansion of the state, purchase of new solutions, services of integrator – all this requires money. I cannot give absolute numbers, but the budget of 2018 is not so big, though increased in comparison with last year. He turns on costs for project teams which we made for experiments.

We have about 80-100 people of developers, programmers, and about 30 people at the enterprises. Still there is a new division on management of corporate data. It collects and prepares data for use, provides computing structures. In total we have about 150 people whom we employed within a year.

Alexey Agapkin: Directly IT occupy a small part in the budget. But we the technology company, at us are the huge plants where an operational part is focus of our efforts. There are several levers at the expense of which digitalization should work. Different experts incline to assessment that in our industry 10% of EBITDA in this plan are a good digit at which it is necessary to aim. And the main part of this story is reached due to operational improvements, including due to technologies. The more tools, the there are more expenses on support. I do not think, as here expenses on IT or on "digit" at us will be reduced.

Alexey Agapkin: We always scrupulously studied and compared each other on operational indicators, studied the reasons of differences in them. Moreover, we study experiment of colleagues on the market, we participate in reference visits, using vendors. Having studied experience of the foreign companies, we understood that all move in one direction therefore, of course, we borrowed their some experience. Convergence of technologies in petrochemistry is not reached yet, but if to ask, for example, bankers who began this story 3-5 years ago they will tell that a set of solutions, a technology stack at all identical.