How do you apply artificial intelligence to your oil and gas needs?
Artificial Intelligence (AI) refers to the body of science, algorithms, and machines able to perform some version of learning and independent problem solving, relying on sufficiently advanced software and hardware components. Within the AI field of study are other sub-branches of computer science, including machine learning, natural language processing, neural networks, and deep learning, among others. But, what does this mean for the oil & gas industry? We spoke to Dr. Russell Roundtree, Vice President, Upstream Data Analytics at IHS Markit to get his take on the technology.
What are key ways oil and gas companies are using artificial intelligence?
Historically, the most successful applications in the oil and gas industry have been around operational data. Improving real-time data feeds off very expensive equipment, whether it be compressors or ESP pumps, where the cost of an equipment failure breakdown is very high. The business impact is large, the data is in large volume and of high quality, and operational decisions can be made that can impact the bottom line quickly.
Can you share any potential risks for the energy industry embracing artificial intelligence?
Interestingly, in my career, AI was a big issue already, even when I came out of school years ago. It got dramatically oversold, and then it stopped. AI ended up being hidden from our industry for a long time because it had been so oversold. So now that AI is back, and is much better than it was back in the day, I think the key issue is managing expectations properly. Find an impactful, small and focused business problem, solve that, demonstrate success, and move on to the next one. But it is so important to ease your way in. Don't promise the world and underdeliver, because then we will be back in a credibility gap like we saw 35 years ago.
What can we expect to see in artificial intelligence for oil and gas in the next 6 months?
I think one of the biggest issues around success and deployment of artificial intelligence in our industry is not so much the artificial intelligence. We can do AI in a lot of places and in a lot of ways, but there is absolutely no question that AI is dramatically more efficient if implemented on the data that resides in the Cloud. So really, the issue around rapid take-up and success in AI is very much tied to the decision about a willingness to move data and work in the Cloud.
Dr. Russell Roundtree, Vice President, Upstream Data Analytics at IHS Markit
Posted 21 December 2018
- Underdog to Champion: Unlock the presalt potential
- The Vaca Muerta Play: Gaining competitive economics through operational efficiencies
- Hydrogen as an Energy Carrier: Can China accelerate the path down the learning curve?
- Southeast Asia/far east jackup rig market update
- India’s city gas distribution (CGD) sector making headway: Brings tempo to downstream gas infrastructure business
- Ahead of its first upstream bid round, has Senegal tightened terms too far?
- Cambodia’s upcoming revision of its power development plan may include LNG-fired power plants
- Finding feedstocks in a quickening energy transition scenario