By Daniel Uhlhorn, Product Manager – Radar Solutions, Hexagon
Throughput is rarely lost in one dramatic event. It leaks away at crushers, conveyors, transfer points, and stockpiles, often exactly where nobody is measuring.
The pressure is on every tonne
Mining is being asked to do more with less. Research across 25 copper mines found that average ore grade fell by about 25% in 10 years, while the International Energy Agency reports a 30% decline in Chile over 15 years. As orebodies become deeper, more complex and more variable, operations must move and process more rock to produce the same amount of metal.
Public reports show that average copper mine grades have fallen by around 40% since 1991. The operational question is therefore not only how much material is moved, but how clearly teams can see its behaviour through the process.
What automating a bulk terminal taught me
My view comes from more than 15 years in bulk material handling. Early in my career, I led and commissioned projects at a fully automated import/export bulk terminal in Rotterdam, Netherlands. Working with autonomous stacker/reclaimers, train loaders, and an uncrewed shiploader taught me two lessons that still shape how I think about material flow.
First, adding state-of-the-art sensors on individual machines does not automatically improve overall performance. Increasing the capacity of individual processes, such as running a reclaimer at its nameplate capacity, does not necessarily translate into higher overall production. Knowing where a machine is and what it does is not the same as knowing what the material is doing.
Second, value sits in the interfaces. Variable frequency drives can adjust conveyor speeds to regulate material flow between assets. A train loading station can then stop the upstream reclaimer at the right moment, avoiding residual material in transit or extra material needed to top up the last wagon.
That only works with a steady material flow and when every step knows what the previous one is delivering.
How bottlenecks and variability develop
Bulk handling performance is usually reported in tonnes per hour. In practice, many processes are constrained by volume, not mass. A conveyor overflows, a chute blocks, or a bin fills up because the material takes up too much space, not because it is too heavy.
Density changes with ore type, moisture, and particle size, so a belt scale can show steady mass flow while volume varies considerably. Belt scales also drift over time. Operators therefore run below nameplate capacity because a blocked chute and unplanned stop can cost more than the tonnes surrendered as a precaution.
That safety margin is where throughput quietly disappears. Variability then travels downstream, while oversize material may be detected only after it causes a problem. Slow, unrepresentative manual sampling leaves teams reacting instead of anticipating.
Small, recurring disruptions add up. At one of our clients running a copper operation in the United States, transfer chute blockages occurred around three times a week, each costing about three hours of downtime. That is more than a full day of lost production every month, from a single transfer point.
Why the choice of sensor matters
When I assess a material flow problem, my first question is whether it can be measured reliably. Experience with encoders, inclinometers, GPS, lidar, and ultrasonic sensors has shown me where each technology reaches its limits.
Encoders and GPS locate machines but do not see material. Dust and fog can obstruct lidar measurements, while ultrasonic performance varies with atmospheric conditions, and cameras require light and a clear view. Radar works at wavelengths in the millimetre range, thousands of times longer than the light a lidar uses. It looks through the dust, steam, and fog that are part of daily life at crushers, chutes, and stockpiles. That is why I advocate radar with conviction: not because other sensors are poor, but because I have seen where they stop working.
Intelligence starts with measurement
Artificial intelligence (AI) is attracting a great deal of attention in mining, and rightly so. A leading Australian mining company reports that AI-based detection of oversize rocks and foreign objects on its Western Australian iron ore conveyors has cut crusher downtime by 20%, with around one million tonnes a year of potential uplift.
The lesson is that AI, or the machine learning in use, is only as good as the measurement underneath it. Digital transformation can stall when data remains siloed, while inferred models can drift or miss what they were not built to predict. Measured data helps close that gap.
The Hexagon perspective: from material movement to material intelligence
At Hexagon, we see material flow intelligence as one connected chain rather than a set of separate assets. Plans set the target, but operational reality spans blasting, hauling, crushing, conveying, storing, and reclaiming. At each stage, valuable information about the material can be lost.
Material flow intelligence connects data from these stages into a single view, helping operations make better decisions. Robust radar measurement at crushers, conveyors, transfer points, and stockpiles supplies live data alongside fragmentation analysis and mine planning information.
The results can be significant. At a large copper operation in Chile, radar-based visibility of the primary crusher bin increased crusher availability by 29% and material throughput by 21%. In large operations, small improvements scale quickly.
Stable flow also means less wear, less spillage, fewer manual interventions, and fewer people sent into hazardous areas. That is how precision, accuracy, and reliability in measurement turn into productivity, quality, safety, and sustainability in the operation.
Where to start
Start small and specific: identify the bottleneck limiting everything downstream, measure it reliably and connect that measurement to the control system. A single, well-chosen measurement point can deliver measurable operational benefits, demonstrating the value of better visibility and building confidence in further investment. The technology to see material flow in harsh conditions exists today. The challenge is to place it where it matters and connect the data across the value chain. When throughput, safety, and cost depend on it, the measurement has to be right.
Where in your pit-to-plant process would better visibility of material flow create the most value? I would be glad to hear your view.
About the author
Daniel Uhlhorn (MSc, Dipl.-Ing., PMP) is Product Manager for Radar Solutions at Hexagon. Based in Germany, he has more than 15 years of experience in bulk material handling and focuses on innovation, automation, and operational optimisation for safer, more productive, data-driven mining operations.