Drilling in a straight line sounds simple. It isn't.
The moment a drill bit enters the ground, it enters a world that doesn't cooperate. Rock formations shift. Unexpected fault zones appear. Hard and soft material alternate in ways no surface survey fully predicted. The drill wants to follow the path of least resistance—and the path of least resistance is rarely the path to high-grade ore.
This is the core challenge of surgical mining: you're not just drilling a hole, you're trying to hit a narrow target zone hundreds of meters underground, and the geology between here and there is actively working against you.
Why Geology Makes Course Correction So Hard
A few hundred meters of overburden can hide a lot of surprises.
Interbedded formations—alternating layers of hard and soft rock—create differential forces on the drill string that gradually push the bit sideways. Fault zones introduce sudden changes in rock strength, sometimes mid-run, causing the drill to deflect sharply before an operator even notices. Natural fractures and voids offer zero resistance, inviting the bit to wander. And the deeper you go, the slower the feedback loop. By the time a deviation shows up in surface telemetry, you may already be meters off your intended path.
In conventional drilling, the response is reactive: pull out, assess, re-enter. Every re-entry is time. Every lost meter is money. And if you miss the target zone entirely, you've drilled a hole with essentially no return.
Surgical mining changes the economics entirely—the ore targets are precise, the entry points are engineered, and there's almost no room for deviation tolerance. That precision is the whole point. But it also means the margin for error collapses.
What Propel Built
Propel designed an AI-powered course-correction engine specifically for surgical mining drill systems—built to keep the bit on the optimal path in real time, even as geology pushes back.
Real-Time Deviation Detection
The system continuously monitors telemetry from the drill string, comparing actual trajectory against the planned path. Rather than waiting for a threshold breach to trigger an alert, the system detects subtle drift patterns early—before they compound into a costly deviation. Think of it as the difference between catching a car drifting toward the shoulder and catching it after it's already left the road.
Steering Recommendation Algorithms
When deviation is detected, the system doesn't just flag it—it tells the operator what to do about it. Steering recommendations are generated based on the current deviation vector, the remaining path to the target zone, and the mechanical constraints of the drilling assembly. The algorithm accounts for the fact that overcorrection is its own problem: a sharp correction in soft ground can create new issues. The recommendations are graduated, precise, and designed to bring the drill back to the optimal path without introducing secondary error.
Live Operator Dashboards
Underground drilling is a high-pressure environment. Operators don't have time to interpret raw data streams. Propel built intuitive dashboards that surface the right information at the right moment—current trajectory, deviation status, recommended corrections, and distance to the target zone—all in a format that supports fast, confident decision-making without cognitive overload.
Hardware and Telemetry Integration
The system integrates directly with existing drilling hardware and telemetry infrastructure. No rip-and-replace. The AI layer sits on top of what's already there, pulling data from existing sensors and feeding recommendations back through familiar interfaces.
The Results
The impact showed up where it matters most.
Deviation-related costs dropped significantly. Less time spent pulling out and re-entering. Fewer meters drilled outside the target zone. Less ore left in the ground because the drill missed its mark.
Time spent in the high-grade zone increased—which is ultimately the only metric that matters in surgical mining. The ore is the point. Everything else is overhead.
Operational outcomes became more consistent. Fewer surprises, fewer emergency decisions, fewer runs that end in a write-off. And perhaps most importantly, this work represents a critical step toward semi-autonomous drilling—a future where the system doesn't just recommend corrections but executes them, continuously and without interruption, while the operator monitors rather than steers.
Why This Matters for the Future of Mining
Surgical mining is premised on precision. The entire value proposition—minimal surface disturbance, selective extraction, smaller environmental footprint—depends on actually hitting the target. AI-guided course correction is what makes that precision durable in the face of real-world geology.
The ground doesn't cooperate. The AI corrects for that. And the further this technology develops, the closer the industry gets to drilling systems that navigate complex geology the way modern aircraft navigate turbulence: automatically, continuously, and with a level of consistency no human operator can match alone.