Mining robotics in September 2026 is advancing through autonomous drilling concepts, underground-to-surface truck automation, and stronger vehicle-safety controls. The most important caveat is that much of this progress remains at the concept, demonstration, validation, or research stage rather than broad production deployment. Mining robotics means machines and software that sense, navigate, perform work, or intervene in vehicle operation with reduced direct human control. The practical shift is toward systems that extend autonomy across more of a mine while retaining defined safety limits.
Table of Contents
- Autonomous drilling is moving toward electric, coordinated fleets
- Underground autonomy is reaching the mine boundary
- Safety systems are not the same as driverless hauling
- Research is improving underground localization and navigation
- The remaining gap is robust commercialization
Autonomous drilling is moving toward electric, coordinated fleets
Sandvik introduced Sami, a fully autonomous battery-electric surface-drilling concept. It combines robotic tool changes, autonomous navigation, and AI and digital-twin coordination for drilling fleets, according to Sandvik's September 1 release. For mine operators, the appeal is not simply an electric drill.
Robotic tool changes could reduce manual intervention, while fleet coordination could help align drilling with downstream blasting and hauling plans. The limitation matters: Sami is explicitly a concept, not a production machine. Treat it as a signal of design direction when assessing future equipment road maps, not as evidence that autonomous electric surface drilling is ready to buy at scale.
Underground autonomy is reaching the mine boundary
Epiroc expanded Deep automation so battery-electric underground trucks can autonomously transition from underground to surface. The system uses 3D LiDAR—laser-based sensing—to position the truck, map its surroundings, and detect obstacles, according to Epiroc's September 3 release. That transition is operationally significant because mines often treat underground and surface routes as separate automation problems.
A truck that can cross the boundary autonomously could reduce handoffs and simplify material movement between work areas. A June demonstration showed a truck stopping for placed obstacles. However, Epiroc did not disclose the scale of broader deployment, so buyers should ask for route constraints, recovery procedures, sensor performance in dust and weather, and evidence from their own site conditions.
Safety systems are not the same as driverless hauling
Hexagon launched a next-generation Vehicle Intervention System for open-pit mines. It can brake vehicles, inhibit propulsion, enforce speed and following rules, and place a vehicle into a fail-safe state, as described in Hexagon's September 3 announcement. This is aimed at mixed traffic, where staffed vehicles, contractors, light vehicles, and heavy equipment share operating areas.
It can be useful even where a mine is not pursuing fully autonomous haulage. Readers should separate intervention from autonomy. Hexagon presents the system as validated to technology readiness level 4, or TRL4, and not as a full driverless-hauling system. A sensible evaluation starts with the rules it can enforce, the vehicles it supports, and how it behaves when communications or sensing are degraded.
Research is improving underground localization and navigation
A 2026 *Journal of Field Robotics* study reported that DURAL sensor fusion achieved 0.167 m absolute pose error within ultra-wideband coverage and 6.456 cm mapping accuracy in real coal mines. Those results, reported in the DURAL study, point to progress on a core underground challenge: knowing where a robot is when satellite positioning is unavailable. The reported data are available only on request.
That limits independent replication, so operators should treat the figures as promising field results rather than a universal performance guarantee. A separate study in the *International Journal of Coal Science & Technology* demonstrated a two-stage vector-field planner that uses onboard sensing to find a tunnel-center path and avoid arbitrary obstacles in synthetic and real mining scenarios. Its value is clearest in narrow, low-visibility tunnels, but the paper does not report a production-mine operating deployment.
The remaining gap is robust commercialization
The ROBOMINERS environmental study reported tests of underground locomotion, mineralogical and geophysical sensing, and production tools on a small autonomous robot. Its conclusion was cautious: commercialization still needs further innovation and research over coming decades.
The MineRobot framework addresses another less visible barrier: modeling the complex movements of underground machines. It offers actuator-centered kinematic modeling and open-source simplified examples, but proprietary industrial benchmarks and commercial-robot parameters limit independent reproduction. For teams evaluating mining robotics, the practical checklist is short:.
- Ask whether the system is a concept, a demonstration, a validated subsystem, or an operating deployment.
- Test sensing, positioning, and recovery behavior in the mine's actual dust, lighting, grades, and traffic.
- Confirm which parts of the workflow remain supervised by people.
- Require evidence for the exact vehicle, route, and task under consideration.
You Might Also Like
- What Is New With Tutorials and DIY Robotics in September 2026? Latest company releases and research papers and Key Takeaways
- What Is New With Food and Beverage Robotics in September 2026? Latest company releases and research papers and Key Takeaways
- What Is New With Ethics and Policy Robotics in September 2026? Latest company releases and research papers and Key Takeaways
