Upstream opportunities and challenges

AI+ upstream, how to break the game

Artificial intelligence (AI) technology is known as the engine of the "fourth" industrial revolution, which will greatly promote the digital transformation and intelligent development of the oil and gas industry, and produce significant social and economic benefits. According to Tres, more than 3,000 oil and gas companies spent about $1 trillion on well and related infrastructure operations in 2019, which could be reduced by about 10% if automation and digitization were accelerated. PWC predicts that by 2025, the upstream operations of oil and gas companies could save $100 billion to $1 trillion in capital and operating expenses through the application of AI technologies.

In recent years, the oil and gas industry faced with the problem of low oil prices, many international oil companies and oilfield service companies have joined hands with IT giants to achieve cross-border cooperation, increase the construction of artificial intelligence platforms, promote the rapid digital transformation of exploration and development business, and effectively improve the efficiency of the potential space. In 2017, Schlumberger and Google launched the DELFI cloud platform, which deeply integrates big data, cognitive computing and other technologies with oil and gas exploration and development, and builds a digital, automated and intelligent professional application environment for the whole process of exploration and development, supporting the innovation and development of enterprises, and making the platform development enter the era from "N" to "1".

Since 2020, major oil companies have stepped up efforts to build smart oil and gas fields in order to achieve cost reduction and efficiency improvement. Shell proposed to build Smart Oilfield, which aims to increase production by 10%, increase recovery by 5% to 10%, reduce operating expenses by 20% and shorten field development cycle by 50%.

At present, the construction level of intelligent oil fields in China is in the transition stage from digital to intelligent, and a few oil areas have basically built the prototype of intelligent oil fields, which have the self-diagnosis, early warning and alarm of oil Wells, and can recommend optimized decision-making schemes.

The challenges of AI development in the oil and gas industry

In the past 20 years, China's petroleum enterprises have created the "oil and gas field Internet of Things construction mode" and "intelligent oil and gas field construction mode" in the construction of intelligent oil and gas fields. The integrity, scale and management of intelligent oil and gas fields have been in the forefront of foreign oil and gas fields. However, the development of artificial intelligence in the oil and gas industry still faces a series of challenges.

First, challenges such as data silos in exploration and development. Upstream exploration and development in the past few decades of information construction, the existence of multiple data entry, standards are not unified, repeated development of functions, information and business integration is not close and other problems have gradually emerged, resulting in more databases, more platforms, more isolated applications and other phenomena are becoming increasingly prominent. The difficulty of data sharing and business collaboration brings challenges to the landing application of artificial intelligence in the oil and gas industry, and the landing application of high-quality artificial intelligence technology requires high-quality big data as the premise and basis. Due to the limited collection technology means, the data that can represent the characteristics of the problem does not have the diversity characteristics, and the single feature big data is not the real big data, which also brings challenges to the research in the complex field of oil and gas exploration and development. Data is the soul of the development of artificial intelligence technology, big data, data quality and governance determine the future of intelligent development, digital transformation to seize the data and data governance is equivalent to seize the future of the development of artificial intelligence technology.

Second, the barrier challenge between artificial intelligence algorithm engineers and business personnel. Usually, business personnel do not understand the artificial intelligence algorithm, and algorithm engineers do not understand the professional expertise of business personnel, resulting in the phenomenon of "not understanding, not being able to say, not getting along" between algorithm engineers and business engineers, which brings obstacles to the landing of artificial intelligence in the business field. The application of artificial intelligence in the oil and gas industry is different from other industries. The objects of oil and gas exploration and development are underground, which is an invisible black box system that uses artificial intelligence technology to solve problems. Unlike the artificial intelligence AlphaGozero in the human-machine game, the Go board it takes is regular and visible. Most of the problems to be solved and dealt with in the oil and gas industry are not the object of any rules and problems to be solved, with super uncertainty. To solve the application problems of artificial intelligence in the oil and gas industry, professional knowledge and industry experience are very important. It is necessary to solve the challenges of barriers between algorithm engineers and business personnel. In order to promote the application of artificial intelligence technology products or scenes.