The mining industry has a dirty secret that no one likes to talk about at shareholder meetings: we are running out of the good stuff. The high-grade, easy-to-process deposits that fueled the last century are mostly gone, replaced by massive, low-grade monsters that are increasingly expensive to wake up.
By March 2026, the industry has reached a breaking point. Energy costs are volatile, water scarcity is a localized nightmare, and carbon taxes are no longer a theoretical threat: they’re a line item on the balance sheet. Digging up millions of tons of rock just to crush and grind most of it into useless tailings is a recipe for bankruptcy.
Traditional mining is essentially a logistics business that moves waste. Sensor-based ore sorting (SBS) is turning that model on its head. Instead of processing everything and hoping for the best, we’re finally getting smart enough to throw away the junk before it even hits the mill.
The Grade Crisis is Real
We aren’t finding many 2% copper deposits anymore. Today, if you find something north of 0.5%, you’re popping champagne. But the energy required to extract that metal doesn’t care about the grade. It takes just as much power to crush a ton of barren rock as it does a ton of high-grade ore.
In 2026, the margin for error has evaporated. For many operators, the difference between a profitable quarter and a massive write-down is the “sub-grade” pile. Historically, this material was treated as waste or stockpiled for a “someday” that never came. Sensor-based sorting is making “someday” happen right now.
How the Tech Actually Works (No Fluff)
Sensor-based sorting isn’t magic. It’s high-speed data processing paired with brutal physical execution. The process involves a conveyor belt or a gravity-fed stream of rocks: anywhere from 10mm to 300mm in size: passing under a suite of sensors. These sensors make a decision in milliseconds: keep or toss.
The heavy hitters in 2026 are X-ray Transmission (XRT) and Near-Infrared (NIR) spectroscopy.
X-ray Transmission (XRT)
This is the gold standard for many operations. XRT doesn’t just look at the surface; it looks through the rock. It detects differences in atomic density. If you’re looking for high-density sulfides in a sea of low-density quartz, XRT sees it instantly. The sensor identifies the valuable chunks, and a row of high-pressure air nozzles blasts them into a separate chute.
Near-Infrared (NIR)
NIR is the go-to for industrial minerals and certain base metals. It identifies minerals based on how they reflect light in the infrared spectrum. It’s particularly effective at spotting specific alteration minerals or deleterious materials like clays that can gum up a downstream copper processing circuit.
The result? You’re not feeding the mill a 0.4% copper ore. You’re feeding it a 1.2% pre-concentrate. The mill works less, uses less water, and produces more metal.
The Economic Calculus: Less is More
Let’s look at the brutal numbers.
A typical processing plant spends roughly 50% to 70% of its energy on comminution: crushing and grinding. If you can use SBS to reject 40% of your feed as barren waste before it hits the primary mill, your energy bill drops almost proportionally.
But it’s not just about saving power. It’s about throughput. If your mill capacity is the bottleneck, removing waste upfront allows you to process more “actual” ore. You’re essentially expanding your mine’s capacity without building a single new flotation tank or buying another ball mill.
Per ton. That’s the metric that kills projects. By the time we hit 2026, the sensor-based ore sorting market has hit a valuation of USD 1.8 billion. It’s projected to double by 2032. That’s not a rounding error. That’s a fundamental shift in how we value assets.
AI and the “New Set of Wings”
The hardware is impressive, but the software is what’s driving the 2026 surge. Early sorting systems were binary: they were either on or off, and they struggled with complex mineralogy.
New AI-powered platforms like OBTAIN™ have effectively doubled sorting capacity without increasing the footprint of the machines. These systems use deep learning to recognize mineral patterns that traditional algorithms missed. Then there’s CONTAIN™, which is a game-changer for “inclusion-type” ores.
In minerals like tin, tungsten, and nickel, the valuable stuff is often buried in microscopic inclusions. Old sensors would miss these, sending them straight to the waste pile. AI-enhanced sensors can now detect these signatures at scale, revealing value where we previously saw nothing but grey rock. This is particularly relevant for projects in the Vicuña District or other emerging polymetallic frontiers.
Case in Point: Bolivia and Beyond
Look at Eloro Resources and their Iska Iska project in Bolivia. They’ve been using Tomra’s XRT technology to test their polymetallic ore. The results? They’ve managed to reject huge amounts of sub-cut-off-grade waste.
For a project like that, SBS isn’t a “nice-to-have” feature. It’s the difference between a massive, expensive CAPEX project and a lean, high-margin operation. By rejecting the waste early, they’ve drastically lowered the capital required for downstream processing.
The same logic applies to critical minerals. As China tightens export controls, the rest of the world is scrambling to bring domestic projects online. Many of these are lower grade or more complex than the overseas deposits they’re replacing. SBS is the only way to make the economics work for a rare earth project in Texas or a lithium plant in the Pilbara.
The Environmental “Bonus”
Let’s be honest: most miners don’t invest in new tech primarily to be “green.” They do it to be profitable. But in 2026, the two are inextricably linked.
Sensor-based sorting can reduce water consumption and greenhouse gas emissions by up to 50% per unit of metal produced. Because you aren’t grinding the waste into fine tailings, you also reduce the volume of tailings you have to manage. This lowers the risk of dam failures and reduces the long-term liability of the site.
Regulators are noticed. In jurisdictions like Chile and Australia, proving you’ve optimized your process for water and energy efficiency is becoming a prerequisite for permitting.
Why Everyone Isn’t Doing It (Yet)
If the benefits are so obvious, why hasn’t every mine on the planet installed a sorter?
The strategic calculus here isn’t subtle: it’s about ore characterization. Sorting doesn’t work for every deposit. If your mineralization is finely disseminated: meaning the gold or copper is evenly spread through every single pebble like sugar in a cake: you can’t sort it. You need “heterogeneity.” You need the valuable stuff to be concentrated in some rocks and absent in others.
There’s also the “fear of the new.” Mining is a notoriously conservative industry. Engineers would rather use a 50-year-old process that they know will work (even if it’s inefficient) than risk a multi-million dollar CAPEX on a “shiny” new sensor system.
But that hesitation is dying out. The 13.2% CAGR in the sorting market tells the story. The early adopters have already proven the ROI. The rest of the pack is now playing catch-up.
2026: The Inflection Point
We are past the trial phase. Sensor-based sorting is no longer a “special project” for the R&D department. It’s becoming a standard component of the modern flowsheets.
From the bauxite mines of Ghana to the copper pits of the Andes, the objective is the same: stop wasting money on rock that doesn’t pay.
The companies that thrive in the late 2020s won’t necessarily be the ones with the biggest deposits. They’ll be the ones with the smartest sensors. They’ll be the ones who realize that in a world of declining grades and rising costs, the most valuable thing you can do is throw away the waste.
It’s about squeezing every possible cent of profit from material that our predecessors would have walked away from. And in 2026, that’s not just innovation. It’s survival.
The clock is already ticking. As we look at the mine plans for 2027 and 2028, the integration of real-time SBS is the primary differentiator between the tier-one assets and the stranded ones. If you aren’t sorting, you’re likely just a very expensive logistics company moving a lot of worthless dirt. That’s a tough sell to a market that’s finally starting to demand real efficiency.
