According to Drilling Contractor, the official publication of the International Association of Drilling Contractors, one of the central challenges facing the drilling industry’s AI push isn’t the technology itself. It’s figuring out how to keep people meaningfully involved once AI enters the workflow.
David Reid, Chief Technology Officer and Chief Marketing Officer at NOV, made that point during a panel session held March 16 at a workshop co-hosted by SPE’s Drilling Systems Advancement Technical Section and the IADC Advanced Rig Technology Committee. Reid’s core argument: before companies bolt AI onto existing operations, they need to get their underlying systems working as efficiently as possible. AI applied to a broken process doesn’t fix the process. It accelerates the dysfunction.
Background
Reid’s comments came during a period of rapid and sometimes chaotic AI adoption across the energy sector. Operators and service companies alike are deploying machine learning tools for everything from predictive maintenance to real-time drilling optimization, often faster than their organizations can absorb the change.
The panel where Reid spoke was part of a broader industry conversation happening at the IADC and SPE level about how to standardize and responsibly scale AI in drilling environments. Interoperability, specifically the ability of AI tools from different vendors to work together on a rig, has emerged as a recurring theme at these events. A separate IADC virtual panel earlier this year tackled that issue directly.
Reid pointed to aviation and commercial shipping as industries that have navigated the human-automation integration challenge more deliberately than oil and gas. Both sectors have decades of hard-won experience designing systems where human judgment and automated processes hand off responsibility in predictable, well-defined ways. Crashes and maritime disasters have, unfortunately, provided those industries with painful data on what happens when that handoff goes wrong.
Analysis
Reid’s framing around “systems-based thinking” carries practical weight that goes beyond executive-level strategy talk. What he’s describing is a sequencing problem. Companies that chase AI deployment without first cleaning up their operational processes are setting themselves up for expensive failures, and potentially dangerous ones on the rig floor.
The aviation analogy is particularly instructive for the drilling sector. Autopilot didn’t replace pilots. It redefined what pilots do, shifting their role toward monitoring, exception handling, and decision-making in high-stakes moments. The drilling industry is heading toward a similar inflection point, where the driller’s job may shift from executing procedures to supervising automated systems and intervening when those systems reach their limits.
That shift has significant implications for training, crew composition, and liability. If an AI-assisted drilling system makes a decision that leads to a wellbore integrity issue or a safety incident, determining accountability becomes complicated fast. Operators and contractors are only beginning to work through those questions at a contractual and regulatory level.
The “human in the loop” concept Reid emphasized is also relevant beyond the rig itself. Data interpretation, anomaly detection, and remote monitoring functions are increasingly being pushed to centralized operations centers, many of them staffed by third-party service providers. That means the humans in the loop aren’t always NOV’s customers directly. They’re subcontractors and service company personnel sitting in monitoring centers in Midland, Houston, or Calgary.
For AI integration to actually work at scale, those service providers need to be part of the system design conversation, not just handed a new interface and told to adapt.
What It Means for Subcontractors
-
Automation doesn’t mean fewer service calls, it means different ones. As AI handles routine drilling optimization tasks, the demand shifts toward technicians who can troubleshoot AI systems, validate outputs, and respond to edge cases the algorithm wasn’t trained on. That’s a skills opportunity for service companies willing to invest in training.
-
Wellsite monitoring and remote operations contractors are directly in the crosshairs. If operators move toward AI-assisted, remotely supervised drilling, third-party monitoring service providers become a critical link in the human-in-the-loop chain. Companies already operating in that space should be positioning themselves as AI-ready, not just data-capable.
-
System design is now a service offering. Consultants and engineering service firms that help operators map and clean up their workflows before AI implementation will find growing demand. Reid’s point about systems-based thinking implies operators need outside help doing that groundwork.
-
Interoperability creates integration work. When AI tools from multiple vendors have to work together on a single rig, someone has to do the integration and ongoing maintenance. Specialty technology service contractors with experience across multiple OEM platforms are well-positioned for that role.
-
Don’t wait for OSHA to define the rules. Regulatory guidance on AI accountability in safety-critical operations is still thin. Subcontractors operating in high-risk environments should be developing their own internal standards now, before a federal or state regulator does it for them under less favorable terms.
