Automating Drilling Processes Using Machine Learning: A Three-Step Checklist


By Philippe Herve and Marla Rosner

There’s been a great deal of enthusiastic discussion around the idea of automating drilling processes in the oil and gas sector. Research suggests automation will be one of the key drivers in the oil and gas industry between now and 2025. Automation improves efficiency, visibility, and reliability of operations, and it’s estimated to have an ROI of anywhere from 15 to 30x for oil and gas[1].

As exciting as all of this is, however, automation can’t take place overnight. Before automating drilling processes, you need to ensure that your rigs are safe, your assets are in good condition, and your operations are optimized—otherwise, you’re simply automating poor or suboptimal procedures.

Machine learning can address these needs by following these three steps:

  1. Using anomaly detection to guard against cyberthreats
  2. Monitoring and optimizing operations using prescriptive maintenance
  3. Sensing and adapting to downhole conditions in real time

1. Cyberthreat detection

The energy industry has been cited as the most vulnerable sector of global business. Between April 2013 and 2014, cyberattacks hit 53% of energy companies, and there are about 540 cyberattacks on oil and gas companies every day [2].

The best approach to cybersecurity for oil and gas is anomaly detection. With the advent of IoT, preventing all threats from entering a network is far less feasible than simply detecting the ones that have already made it in. Anomaly detection software is designed to monitor the behavior of endpoint devices within the network and flag any unusual behaviors or abnormal signals being sent out.

Machine learning accomplishes this by generating hypotheses using multiple data sets, which allows it to catch subtle attacks involving unusual behaviors that still fall within the normal rules of operation, but are statistically unusual. Anomaly detection software powered by machine learning monitors the behavior of endpoint devices and flags unusual behaviors in four steps:

  1. Trains on log data to build a feature set of IP addresses in normal inter-IP communication
  2. Creates a profile of normal traffic patterns
  3. Creates a profile of suspicious or malicious behavior
  4. Detects and blocks anomalous activity

2. Prescriptive maintenance

According to Gartner, prescriptive analytics is “the final frontier of analytic capabilities.” Using machine learning algorithms, prescriptive analytics uses historical and current data to not only predict what will happen, but to pinpoint why it will happen, recommend potential plans of action, analyze the ramifications of each possible choice, and even help set a chosen plan into motion.

Prescriptive analytics needs to be powered by robust machine learning software, which allows it to adapt to changing conditions, understand unstructured data with natural language processing, and scale to any size of operation.

3. Drilling optimization

Downhole conditions can be unpredictable, and drilling dysfunctions can massively disrupt the cost to drill and complete a well. Machine learning algorithms use sensor data to analyze downhole conditions and immediately alert personnel when an issue arises, increasing speed and accuracy of detection.

There are a number of such issues, each of which can present major threats to drilling operations: well kicks, stuck pipe, excessive downhole vibration (which can strain the drill string, causing it to wear out too quickly), and more.

Machine learning applications address these threats by using and analyzing sensor data. Potential problems can thus be flagged with a reliable digital indication in a timely manner. This, in turn, can help to optimize drilling by allowing companies to reach the next casing point in both the shortest possible time and the safest possible manner.

You can’t build a sturdy house on shaky bedrock. Similarly, there’s no point in automating drilling processes unless you’ve ensured those processes are optimized in every way possible. By applying machine learning to your drilling processes, you can make certain that your operations are the best they can be, and enable yourself to move forward into the future that is automation.

[1] Montague, Jim. “It’s Time to Improve Automation in Gas and Oil.” Control Global. 23 June 2015. Web.

[2] Haidar, Tim. “Cyber 9/11: Is the Oil & Gas Industry Sleepwalking into a Nightmare?” Oil and Gas IQ. 2015. Web.

Also, check out our webinar on the uses of AI in the Oil and Gas industries to learn more.


Latest blogs

Abstract depiction of our Generative AI Platform created by DALL-E
Campbell LeFlore

What’s Inside our Generative AI Platform?

Perfected over ten years of real-world engagements serving many of the world’s largest brands in energy, manufacturing, transportation, utilities, financial services, and other industries, SparkCognition’s

Read More
Campbell LeFlore

The Top Challenges HSE Managers Face Today

HSE managers are the behind-the-scenes heroes enabling today’s high-throughput industrial environments to operate at peak performance. Their determined efforts ensure the well-being and protection of

Read More
SparkCognition is committed to compliance with applicable privacy laws, including GDPR, and we provide related assurances in our contractual commitments. Click here to review our Cookie & Privacy Policy.