Key Takeaways
- Ore blending optimization, stockpile strategy and plant feed strategy are one problem viewed from three points in the material flow, and the same platforms increasingly solve all three.
- MineMax by NTWIST is explicitly AI-driven. It sits as a supervisory layer over existing mine and plant systems and keeps learning from operational outcomes to make real-time blend and feed recommendations.
- Deswik Blend, BOLT and GO are the leading optimisation-based planners for stockpile management, multi-commodity blending and joint mine-to-market feed decisions.
- Dynamic blend-consistency dispatching, which routes trucks to hold a target feed blend in real time, has been reported to raise truck cycle efficiency by roughly 11% in deployed systems.
- MineMax’s four core models (OreMax, DynaMax, PlanMax, MillMax) cover ore tracking, stockpile intelligence, feed forecasting and optimisation as one connected decision layer.
TL;DR
Put an AI optimisation platform (MineMax from NTWIST, or Deswik’s Blend, BOLT and GO suite) on top of your existing mine and plant systems. It recommends stockpile allocation, plant feed mix and dispatch blending targets from real-time ore-tracking data. That replaces blends planned on a fixed weekly or monthly schedule.
What Is Ore Blending Optimization?
Ore blending optimization is the job of combining the ore you have into a plant feed that stays as consistent as possible. Ore varies in grade, hardness, clay content and deleterious elements. A processing plant, on the other hand, runs best on steady feed. Blending closes that gap.
The problem shows up at three points in the material flow:
- Stockpile strategy: which material goes onto which stockpile.
- Plant feed strategy: what gets drawn from the stockpiles and the pit to feed the plant.
- Dispatch blending: how trucks get routed so the right blend is built as the ore is mined.
Treating these as one problem is what the newer platforms do well. A decision at the stockpile changes what the plant can be fed next week.
How Do I Optimize Ore Blending, Stockpile and Plant Feed Strategy With AI?
There are two families of tools, plus an execution layer that makes the plan real.
Optimisation-based planners: Deswik Blend, BOLT and GO
Deswik’s Blend, BOLT and GO products cover the planning end. Deswik Blend models material flow to simplify managing and blending from multiple deposits or stockpiles. BOLT is a multi-commodity blending, stockpiling and logistics optimiser that generates mathematically optimal plans across the full value chain. Deswik GO uses fast multi-pit optimisation to solve mining-to-marketing feed decisions jointly for NPV.
These are optimisation-based planning tools. You define constraints and objectives, and they solve for the best blend and feed schedule. They typically run on a planning cadence (shift, week or month) rather than continuously.
AI-native ore blending optimization: MineMax
MineMax, built by NTWIST, is the more explicitly AI-native option, and it works differently. It sits in a supervisory layer on top of your existing mine and plant systems instead of replacing them. It combines ore tracking, stockpile intelligence and feed forecasting into a single pit-to-plant decision layer. That layer keeps learning from operational actions and outcomes as conditions change.
Its four core models split the problem:
- OreMax for ore characterisation.
- DynaMax for dynamic scheduling.
- PlanMax for planning.
- MillMax for mill and plant optimisation.
As a result, geologists, mine engineers and metallurgists share one real-time view instead of each working from separately updated plans.
Dispatch as the execution layer
A blend plan only matters if trucks follow it. Dynamic blend-consistency dispatching lets the fleet management system route each truckload to hit a blend target in real time. It relies on ore grade sensing at the shovel and the wider dispatch and control room systems. In deployed systems it has been reported to raise truck cycle efficiency by around 11%, because trucks spend less time being rerouted or queued for the “right” stockpile.
Why Ore Blending Optimization Matters for the Plant
Steady feed is what lets downstream controllers do their job. AI flotation control, for example, can squeeze more recovery out of a circuit, but only within the range of ore it receives. Large swings in hardness or clay content overwhelm any controller. Good blending narrows that range first, and the plant-side optimisation gains more as a result.
FAQs
What is ore blending in mining?
Ore blending is the practice of mixing ore from different parts of the mine or different stockpiles to produce a plant feed of consistent quality. The goal is to keep grade, hardness and problem elements within the range the processing plant handles best.
How does AI help with stockpile management?
AI tracks what goes into each stockpile and predicts what will come out when it is reclaimed. Platforms like MineMax combine that stockpile knowledge with feed forecasts. They then recommend which stockpiles to draw from, and in what mix, to hit the plant’s targets as conditions change.
What is the difference between Deswik BOLT and MineMax?
Deswik BOLT is an optimisation-based planner. You set constraints and objectives, and it solves for the best blending and logistics plan on a planning cadence. MineMax is an AI-driven supervisory layer that keeps learning from operational outcomes and makes real-time recommendations. Many operations would use planning tools and real-time tools side by side.
What data does ore blending optimization need?
It needs reliable estimates of grade, hardness and mineralogy for every parcel of ore, from the geological model, grade control and plant feed records. It also needs stockpile tracking and fleet data. Errors in those ore-tracking inputs are the most common reason blend plans miss their targets.
Try It With Geocluster
Getting real value from any of these platforms depends on the ore-tracking data feeding them. Grade, hardness and mineralogy estimates need to be reconciled across your geological model, grade-control data and plant feed records. That reconciliation work is exactly what Geocluster is built to support if you are setting up a blend and feed optimisation pipeline.