Blog /
1 NOVEMBER 2017

Some Insights from Our Joint Conference with Gazprom Neft

What do process manufacturers think about AI today?

Following our successful Artificial Intelligence (AI) in Process Industries conference in St. Petersburg, co-organised with Gazprom Neft, one of Russia’s leading vertically integrated oil companies, our COO Alexander Khaytin shares his key takeaways from the day and vision of the future of industrial production.

1. Large uncertainties = big opportunities

Many process industries are competing on a tough, demanding market but AI is now being recognised as the new source for optimisation in process industries.

This is all thanks to AI’s ability to incorporate “uncertainty” within its algorithm and handle it alongside the other thousands of factors it processes automatically. This is especially important in industries full of uncertainties; from fluctuations in the composition of raw materials and reservoir conditions, to measurements that cannot be correctly taken in real time.

AI therefore presents a big opportunity for optimisation and improved efficiency that cannot be overlooked.

2. Evolving beyond an Industry 4.0 talking point

Just a year ago, AI was no more than a discussion around the Industry 4.0 trend. But today, as companies begin to see it has a very concrete and practical way of increasing operational efficiency for different processes, AI has come into its own and become a standalone trend.

AI pilots discussed by leading industry experts from oil and gas, petrochemical, metallurgy, food processing and consumer goods during the conference, provided clear evidence of AI’s ability to impact a number of sectors. From optimising gas fractionation unit performance and increasing drilling speed and automatic selection of drilling modes, to optimising moisture content levels in food processing and preventing failures in electric submersible pumps.

With the number and type of AI applications rising rapidly, it won’t be long until we start to see tangible results. Indeed, we are already seeing a change from asking how applicable AI is to process industries to discussing the practical issues of applying it – for example, “Who is in charge of the decisions made by AI?” or “How are the management practices getting changed to reflect this shift?”

As this shift occurs, addressing the challenges of real AI implementation will only become more important.

3. AI should not be seen as a genius

The industry today is far more pragmatic and realistic in its approach to AI and what it can deliver.

AI should not be viewed as a super-intelligence but rather, a simple worker, proficient in a specific area. Just like any other employee, AI should be trained, tested, and controlled. However, AI’s success will quickly remove any doubts and concerns. Therefore, when successful, AI is able to solve tasks automatically, faster and more accurately than a human. Taking away the human factor, AI is able to account for an almost infinite number of factors and find correlations within these to allow an increase in both quality and speed.

4. The inevitability of AI means companies have become more comfortable with the idea of a “black box”

Accepting the use of a “black box” has often been a barrier to AI implementation, but this is changing. When AI shows measurable results, it becomes less important for it to be explainable – especially when these results equal monetary profit.

Nowadays, other issues and questions such as “How can the safety of decisions AI makes be secured?” “How can we move from recommendations on the spot to the automation of the whole process?” arise instead and become more important. Indeed, we are now seeing the industry envision the future of its businesses with AI and asking what they should await from the 2050’s factories.

5. AI will bring global changes

With the advent of AI, the main focus is, and has been, in the very specific use cases and their effect. However, the industry is now becoming more concerned with the global changes this new technology will bring.

I expect to see a shift from local optimisation to the optimisation of larger segments and entire production processes. While enhanced efficiency will present new opportunities for the industry, it will also present new challenges.

AI’s task will only become greater as technology and innovation grows, which will see an overall expansion of AI implementation to the production process.

The “Artificial Intelligence in Process Industries” conference was held on September 13, 2017 in St. Petersburg. It was co-organised by Yandex Data Factory and Gazprom Neft and gathered leading industry experts from oil and gas, petrochemical industry, metallurgy, food processing and consumer goods.

To find out more about AI developments in process industries, sign up to our newsletter below.

Alexander Khaytin
Alexander Khaytin
Alexander Khaytin is the Chief Operating Officer of Yandex Data Factory. He oversees all operational activities, while lending his expertise to YDF’s partnership, sales and technology strategies.
ALL PUBLICATIONS