Russia's Biggest Data Lake & How Severstal Is Transforming The Steel Industry Using Machine Learning

Russia's Biggest Data Lake & How Severstal Is Transforming The Steel Industry Using Machine Learning

Severstal is among  the largest manufacturers of steel in Russia, and therefore the world. Recently it announced that as part of its digital transformation strategy it has created the country’s largest industrial data lake. Petabytes of data which was previously discarded during production of thousands of tons of steel each year will now be stored for analysis.  

Russia's Biggest Data Lake & How Severstal Is Transforming The Steel Industry Using Machine Learning

The aim is to secure a competitive advantage in the face of growing competition from other steel producing regions, including Asia and Europe. Although these international competitors may not currently generate the same volumes as Russia’s plants, many have pulled ahead in the technological race, closing the gap and prompting Severstal to launch its own advanced analytics-driven transformation.

Their head of public relations, Anastasia Mishanina, told me “We understood that – while they may have lower earnings than us - if we don’t go there, the others [international competitors] will catch us up.

“This is why we have to be constantly looking for ways to improve our responsiveness, and our competitive advantage – we started to look everywhere and ask, where can we go further? And if we are not going there now, we will be lagging behind.”

Just as in the US and China, digital transformation, the Internet of Things and Industry 4.0 are seen as critical to the future competitiveness of Russian industries – President Putin recently stated that “advanced technologies” have a key part to play in ensuring the competitiveness of the nation’s industry on the international stage.

In the Russian metals and mining sector alone, it is estimated that efforts to drive innovation through digital technology will generate growth of $320 billion in the next decade.

Severstal’s data lake, built with the help of partners including Lenovo and Microsoft, will mean a leap from capturing just 5%   of available data from production operations, to over 50% .

Uses for this data include predictive maintenance on the heavy manufacturing equipment and oversight of quality control processes – monitoring for inefficiencies or impurities introduced in the production at Severstal’s vast Cherepovets milling plant in the Volgoda region. It will soon be rolled out to other major facilities such as the mining plant at Vorkutaugol – one of Russia’s largest coal coking sites.

Online shopping

Another initiative   centre s around the launch of an “online store” for buying steel – “Which sounds bizarre,” Mishanina says, “Because buying steel is not like buying a pair of shoes! But Industry 4.0 gives us these opportunities.”

With clients able to log in and order steel in much the same way as we might order everyday items from Amazon, Severstal gains access to much of the same customer data as the online retail giant collects through its customer recommendation and segmentation operations. Currently only available to existing Russian clients, it will be rolled out to new clients and the international market in the near future.

Unstructured data

The collection and analysis of unstructured data – specifically video – to be an alys ed through machine learning algorithms, is a major focus of the initiative.

“We already have cameras all over the facilities but often they are just monitored by operators, using their eyes, trying to see or hear if something is going wrong.

“None of that information is stored and if they miss it, the opportunity to learn is gone. Now we will be storing all of that information. Not just what we can use today, but data that we will probably have a use for in the future.

“This transformation is going to be a constant and ongoing process of updating our capabilities to achieve maximum coverage of the data we have, and to make a lot of decision-making within the plants automatic,” Mishanina tells me.

So, what have they learned so far? The first lesson was that this needed to be an in-house operation – “So we hired a CDO and he built a team of data architects.”

And just as importantly, the team set about the task of securing buy-in from the business and production teams whose work would be transformed by the data they were generating.

“We tried to outsource here and there, but we re alis ed it needs constant attention, and really deep integration of the team into the plant. If there is only IT guys, or digital guys, involved, then it’s not going to work.”

Sometimes, of course, it just won’t work anyway – often this is due to a failure in the original conception of a strategy.

“Your models aren’t always going to work – you will have to reformulate some and be prepared to write off bad ideas when they don’t prove useful – not all models will work once they go into production and not all of them will be one hundred per cent   useful.”


Related Articles

 


 

Written by

Bernard Marr

Bernard Marr is a bestselling author, keynote speaker, and advisor to companies and governments. He has worked with and advised many of the world's best-known organisations. LinkedIn has recently ranked Bernard as one of the top 10 Business Influencers in the world (in fact, No 5 - just behind Bill Gates and Richard Branson). He writes on the topics of intelligent business performance for various publications including Forbes, HuffPost, and LinkedIn Pulse. His blogs and SlideShare presentation have millions of readers.

Some Of Our Customers

Connect with Bernard Marr