Considering that the world is awash in oil, the oil and gas industry should deploy artificial intelligence (AI) to improve efficiencies and cut costs.

The lackadaisical adoption of artificial intelligence (AI) by the oil and gas industry is surprising, says an article in the Houston Chronicle by Chris Tomlinson.

Interestingly, the oil and gas industry has pushed through impressive technological innovations in recent decades. For example, sophisticated mechanical tools can now direct a drill bit accurately through a narrow column of oil for over a mile.

Oil and gas industry: Operational sophistication

Horizontal drilling and hydraulic fracking have together made it possible to exploit massive unconventional oil and gas resources in the US. These were previously considered uneconomic to drill. Floating production storage and offloading (FPSO) and floating liquefied natural gas (FLNG) vessels are technological marvels.

The collection and analysis of seismic data are unrecognizable from the arbitrary approaches of early explorers. Geologists are now able to use seismic imaging through vibrations to visualize rock formations located thousands of feet below the ground.

Why then is the industry seemingly dragging its feet on AI?

Oil and gas industry and AI

What is perhaps holding back the industry is the risk to human lives and billions of dollars in investment. An AI experiment that backfires could, therefore, be a serious setback for the industry.

Another problem is the scarcity of adequately experienced AI companies that have the requisite skills to safely conduct the deployment of AI in oil and gas operations.

“There are dozens of companies promising huge savings from digitalization. But without a track record, executives are unwilling to spend the money,” says the article by Tomlinson.

Unfortunately, persistently low oil prices are now a harsh reality for the industry. In such an environment, oil producers must do everything in their power to cut drilling costs. Imperative also is the avoidance of any unscheduled production outages.

The oil and gas industry must, therefore, bite the AI bullet.

Recent developments are encouraging, however

Late last month, Microsoft and Baker Hughes announced an artificial intelligence (AI) partnership for the oil and gas industry. Along with AI developer C3.ai, the two giants will bring artificial intelligence onto Microsoft’s Azure cloud computing platform. Oil and gas customers can, therefore, easily adopt AI for applications in inventory, energy management, predictive maintenance, and equipment reliability.

“Companies that adopt this technology will be the next Amazon, and those that don’t adopt will be the next Sears,” C3.ai founder and CEO Tom Siebel told the Houston Chronicle.

Royal Dutch Shell processes 20 million tonnes of crude oil annually at the massive Pernis refinery in Rotterdam. About 50,000 sensors monitor equipment and operating conditions at the refinery. Since last year Shell has been using data from the sensors in a machine learning application that predicts failures in control valves. The result: advance notification periods of potential failures from 4 to 75 days. Some of these warnings would have been impossible to detect with the human eye or with conventional computer systems.

Royal Dutch Shell has 280 projects underway in AI. It is now using AI to clean up data from seismic surveys and thereby create images of rock formations on the seafloor. In turn these images help scientists locate all deposits under the ocean floor. However these images suffered from noisy data because of underwater currents and other factors that impeded their accuracy. Shell was able to innovate machine learning algorithms that filtered out this noise. The time needed to obtain clear images was cut by 80%.

Chevron Corporation has deployed AI-based software to analyze historical well performance data. It is, therefore, now able to drill in better locations and improve production by 30%.

[Related Story: Automaker Ford Using Artificial Intelligence To Optimize Daily Driving  ]