A geologist spends months building a block model. The geology team turns it into dig blocks and material classifications and load it into the FMS. The Dispatch team send it out to the digger and activate it. Then a 250-tonne truck rolls up to the shovel or excavator, and the FMS decides: ROM pad, stockpile, or waste dump, and the associated grades. That decision happens hundreds of times per shift, and getting it wrong in either direction costs money. Ore to the waste dump is lost revenue. Waste to the mill is wasted processing capacity.
The theory (cut-off grades, block models, dig blocks) is covered in Grade Control Fundamentals. What follows is the mechanics: how the FMS turns geological knowledge into truck routing decisions at the dig face.
How the FMS Knows Where the Bucket Is
Every routing decision starts with a position fix, but not just the machine’s position. The FMS needs to know where the bucket is digging, and on a large excavator or shovel the bucket can be 15+ metres from the GPS antenna on the machine body.
Two systems work together to solve this:
GPS/GNSS fixes the machine’s base position on the bench. With RTK corrections, this is typically accurate to 10 to 20 cm. That tells the FMS where the machine is sitting, but not where it’s digging.
Arm geometry fills in the rest. Inclinometers (attitude sensors) or cylinder stroke sensors mounted on the boom, stick, and bucket measure the angle of each joint. Because the FMS knows the physical dimensions of each arm component, it calculates the 3D position of the bucket teeth relative to the machine body. Combine that with the GPS base position and you get the precise mine coordinates of the digging point.
Systems like Cat Terrain for Loading and Komatsu ProVision use this GPS-plus-arm-geometry approach, some achieving bucket-tip accuracy within centimetres. Without arm geometry, the FMS would only know where the machine is parked, not where the bucket is cutting into the face.
Dig block management and block model integration are covered in dedicated articles.The Grade Control Workflow
With the bucket position established, every load decision flows through the same sequence:
- The system locates the bucket. GPS fixes the machine position; arm geometry sensors calculate the bucket tip location in mine coordinates.
- It identifies the dig block. Based on those coordinates, the FMS determines which dig block the material is coming from.
- It retrieves the predicted grade. The dig block carries an assigned grade from the block model. The FMS uses this to classify the load.
- The truck gets a destination. Classification rules map the material to a destination: high-grade to ROM, medium-grade to stockpile, waste to dump. The FMS tells the truck where to go.
- The truck hauls and confirms. The FMS tracks the truck to verify it arrived at the assigned destination. If a truck dumps at the wrong location, the system logs the discrepancy.
Automated vs Manual Dig Block Assignment
Not all FMS systems handle step 2 the same way. The level of automation varies:
Fully automated: the FMS uses GPS and arm geometry to identify the dig block without operator input. The system looks up the grade, classifies the material, and assigns a destination. This is the approach used by systems like Cat Terrain and Komatsu ProVision integrated with DISPATCH.
Operator-selected: some systems display the available dig blocks on the operator’s touchscreen (or mobile data terminal) and require the operator to select or confirm the active dig block and material type. Wenco’s FMS, for example, allows operators to change dig block, material, and dump selection via foot pedal or touchscreen, keeping hands on the controls while making rapid selections.
Hybrid: many sites use automated assignment with operator override. The FMS suggests a dig block based on the bucket position, but the operator can manually change it if they know the automated assignment is wrong. This is common near dig block boundaries or when the geology team has flagged a discrepancy.
On properties running multiple dig units, dump and stockpile locations, this workflow fires hundreds of times per shift. The alternative, supervisors making routing calls by radio based on memory and visual cues, doesn’t scale.
What Makes It Work (or Not)
The FMS can only be as good as its inputs. Four things determine whether grade control actually works:
Positioning accuracy is rarely the weak link. GPS with RTK corrections fixes the machine to within centimetres, and arm geometry extends that precision to the bucket teeth. Combined, this means the digging point is known within a margin well inside the size of a dig block. Problems arise when RTK corrections drop out and GPS accuracy degrades to 10+ metres, or when arm geometry sensors fall out of calibration. Either can cause the system to assign the wrong dig block.
Dig block currency is where most mines get caught. If the dig blocks haven’t been updated to reflect current mining progress, every truck gets routed based on stale information. The operation mined through that boundary two benches ago, but the FMS still thinks it’s ore. This is the most common cause of misrouting, and the easiest to fix.
Block model quality sets the ceiling. Insufficient drilling, poor geostatistics, or an outdated resource model means the predicted grades won’t match reality regardless of how well the FMS performs. You can’t route your way out of a bad model.
Classification rules must match the plant. If the cut-off grades and material categories don’t reflect the processing plant’s actual capabilities, you’re sending material to the wrong place with high precision. Most mines review these when commodity prices shift or processing conditions change.
Destinations
Every material category maps to a physical destination. The FMS enforces these assignments load by load. Material destination control is covered in a dedicated article.
ROM Pad
The ROM pad (run of mine) feeds the primary crusher. High-grade ore goes here for direct processing. This is the destination that generates revenue. Everything else is either stored or discarded.
Stockpiles
Most mines maintain multiple stockpiles to buffer between mining and processing and to segregate material by grade:
- High-grade stockpile: above current mill target, held for blending down
- Medium-grade stockpile: processable but below target, held for blending up
- Low-grade stockpile: below current cut-off but potentially viable if commodity prices rise or processing costs fall
Getting stockpile management right is its own discipline. The FMS tracks what goes onto each stockpile, but reclaiming and blending from stockpiles is a separate operational challenge. Stockpile management and grade blending are covered in dedicated articles.
Waste Dumps
Material below cut-off grade goes to waste dumps. The FMS tracks waste movements for reconciliation and environmental compliance, even though the material has no processing value. Knowing exactly where waste went matters when regulators ask.
Heap Leach Pads
Some operations use heap leaching for low-grade material that’s too expensive to mill but contains enough metal to extract through chemical leaching. The FMS routes eligible material to the leach pad based on grade thresholds. This is common in gold and copper operations.
Key Takeaways
- The FMS needs both GPS and arm geometry to know where the bucket is digging. GPS alone only locates the machine. Arm sensors extend that precision to the bucket teeth.
- Understand whether your FMS assigns dig blocks automatically, relies on operator selection, or uses a hybrid approach. Each has trade-offs in speed, accuracy, and operator workload.
- Keep dig blocks current. Stale definitions are the single most common cause of misrouting, and the cheapest to fix. If geology updates lag behind mining progress, the FMS routes on fiction.
- Match your material categories to your actual processing options. Three well-defined categories that operators understand beat eight categories that create confusion at the dig face.
- Don’t blame the FMS when material ends up in the wrong place. Work backwards: classification rules → dig block definitions → block model quality → positioning accuracy. The problem is almost always upstream.