Autonomous patrols through drifts, haul ways, pump rooms and ventilation tunnels — in environments where GPS does not exist and every human entry requires a permit.
Satellite navigation ends at the portal. Fixed infrastructure (rails, cameras) only covers selected points. Legged robots with LiDAR SLAM walk the actual route — ramps, rubble, water, and all.
Methane, CO and oxygen-deficient zones make routine walkdowns a permit-controlled activity. A robot with onboard gas sensing takes the readings before anyone walks in.
Kilometres of drift and conveyor routes cannot be walked daily. Areas between scheduled inspections stay blind — conveyor hot spots (the same failure mode that drives switchyard thermal scanning), water buildup and roof changes (similar to what post-incident structural inspection looks for) go unnoticed.
| Navigation | 3D LiDAR SLAM, fully GPS-denied; re-localisation at known waypoints; autonomous re-route on blocked passage |
|---|---|
| Gas sensing | CH₄, CO, O₂ as standard; H₂S and site-specific cells configurable; continuous logging with alarm thresholds |
| Thermal camera | Radiometric module for conveyor drives, pump stations and electrical infrastructure |
| Visible camera | 4K with zoom; low-light performance for tunnel lighting conditions |
| Environmental | Dust and humidity protected build; IP67 typical; suitability for wet drift floors assessed per site |
| Runtime / charging | Platform-dependent; dock or manual swap strategies matched to your shift pattern |
| Comms | Wi-Fi / mesh backhaul per mine standard; offline autonomy for out-of-coverage segments |
| Software | ROS 2 + Python SDK; dashboard with route map, gas heat-maps, thermal library and report export |
A reference shape for a quadruped inspection rollout in mining & tunnels. Every system is re-scoped to your site — this is the typical starting point, not a fixed offer.
Priority routes, hazard zones, comms coverage and gas rules mapped with your engineering team.
Gas array, thermal thresholds and navigation behaviour tuned for your drift conditions.
Supervised rounds first, then autonomy; operators and maintenance staff trained.
LiDAR, visual, and fused SLAM each have failure modes. Here is a decision framework based on the actual environments we walk into.
Read article →When does a robot dog actually beat fixed cameras, wheeled robots, drones or human inspectors? A side-by-side guide for industrial buyers.
Read article →Certification depends on your jurisdiction and the specific classification of the workings, so we do not claim a blanket approval. We supply intrinsically safe or explosion-protected platform variants — including the Micbot MOVENEW P1/P2 — together with the technical documentation your regulator or certifying body requires, and we work to MSHA, ATEX or IECEx schemes as your site demands.
By LiDAR-based SLAM: the robot builds a map of the drift as it travels and localises against it, with wheel or leg odometry and an IMU for dead reckoning where the geometry is repetitive. It does not need tags, beacons or pre-installed infrastructure. Dust and low light degrade cameras, not LiDAR, which is why it is the primary sensor underground.
Yes — rough and wet surfaces, loose rock, debris and gradients are routine for an industrial quadruped, and the enclosure is rated against water jets and dust ingress. Water deeper than the rated wading depth, and thick slurry that packs joints, will end a round. We validate the actual routes during the site survey so limitations surface before contract, not after.
Underground the highest-value repeats are ground support and convergence checks, ventilation readings, water and sump levels, conveyor and idler condition, and gas monitoring in workings that are unsafe to enter after a blast. The robot removes the person from the exposure; it does not remove the competent person's judgement from the record.
Tell us the metres of drift, the assets on the route and your current inspection schedule. We’ll come back with a patrol plan, gas-sensor configuration and indicative pricing within one business day.