Mining Vehicle Safety and Dispatch Solutions
Cameras, radar and edge AI engineered for dust, darkness, unmarked roads — and pits where the network never reaches.
Bucket & Shovel Integrity
A bucket tooth can be gone for a full shift before anyone sees the gap.
Loss alert at the digging face
Vision watches the bucket profile through every loading cycle.
A missing-tooth event raises an in-cab alert and a dispatch notification immediately.
One system across bucket types
The detection model adapts to different bucket shapes and tooth counts.
Changing attachments means a quick re-learn on the same hardware, not a new purchase.
Downtime that is planned, not discovered
Early detection turns a surprise failure into a scheduled change-out.
Maintenance crews arrive with the right spares and a service slot already booked.
Alarms wired into your maintenance flow
Every event exports with a time stamp, site tag and video clip through an open API.
Your work-order system can raise a repair ticket the moment the alarm fires.
Crusher Feed Control
A crusher throat does not argue with an oversized rock — it simply stops.
Sizing at the feed point
A camera frames every rock as it moves toward the crusher.
The edge model compares each load against the size limit set for that site.
Alerts before the feeder commits
An oversize detection fires while there is still time to divert the rock.
Operators can stop the flow instead of clearing a jammed chamber.
Thresholds you set per pit
Size limits are configured in the SOHANIoT Cloud Platform and pushed to the site.
Different pits, crusher specs and production targets each keep their own settings.
Alarm footage you can replay
Every oversize event stores the video and time stamp for later review.
When a jam does happen, the playback shows exactly what entered the chamber.
Haul & Load Dispatch
The shortest route is rarely the fastest cycle when five trucks arrive at the same shovel.
One screen for the whole operation
Dispatchers see haul trucks, excavators and loaders moving on a live map.
Assignments are based on where each unit actually is, not where it was an hour ago.
Re-allocation that reacts to the queue
The system watches load times, travel times and waiting lines at each shovel.
When a truck idles too long, the plan adjusts and the next task is re-routed.
A production pulse for the shift boss
Daily tonnage, trips per truck and load counts update as they happen.
Slowdowns or gaps in the flow raise alerts before the shift report does.
Directions that reach the cab
The next destination and loading point can be shown on an in-cab display.
Operators spend less time on the radio and more time moving material.
Blind-Spot Guard for Unmarked Roads
On a haul road with no markings, a service pickup can hide in a mirror gap until it is too late.
Detection tuned for the mine, not the highway
The model expects rock faces, berms, dust haze and roads without lane lines.
It looks for the objects that actually appear around mining equipment.
People and light vehicles, both
Pedestrians, pickups, light trucks and inspection cars are recognized as distinct classes.
A worker on foot and a utility vehicle get different, appropriate warnings.
Profiles that follow the road
The same camera can carry separate behavior profiles for site roads and public roads.
Zone rules or a simple switch choose the profile — no hardware change needed.
Night shifts stay covered
Low-light imaging keeps people visible when the sun drops below the pit rim.
Reflective vests and work lights give the model the contrast it needs after dark.
Cabin Vigilance in Mine Conditions
Helmets, masks, sunglasses and a washboard road — attention is still readable when the model is built for it.
A vision model that grew up around PPE
Detection is trained on operators wearing helmets, masks and reflective clothing.
It keeps working through the posture shifts and vibration of real haul cycles.
Fatigue cues beyond the eyes
When sunglasses or dust block the eye view, the system leans on other signals.
Optional supplemental sensing can be added where your data policy allows it.
Alerts that do not need a signal
The edge unit drives in-cab sounders and optional seat alert hardware locally.
Nothing waits on the network — warnings fire in the deepest part of the pit.
Privacy that is set by policy, not by default
Video stays on the device unless an event occurs.
Upload rules, retention windows and access rights are configured per fleet.
360° View That Follows the Maneuver
A haul truck's blind zone is bigger than most loading docks — and it moves with the cab.
A real-time bird's-eye image
A ring of cameras is stitched into one seamless view around the truck.
The layout shows the driver what surrounds the machine at a glance.
The view changes with the maneuver
The display follows the gear, turn signal, speed and steering input.
Reversing toward a dump point shows the rear scene before the driver asks for it.
Calibration that fits a working day
A guided calibration routine completes in minutes.
It can be re-run on site after a mirror, body or camera change.
Warnings that reach outside, too
Ground crews near the wheels can receive their own audible alert.
The truck announces its blind side instead of relying on the driver's memory.
Over-the-Hill Warnings
The crest hides the oncoming truck until both drivers are already committed.
A cooperative line of sight
Vehicles and fixed points exchange position and heading over a short-range link reaching metres.
Each unit learns about the other before either one appears in a windshield.
Positioning you can tighten
An optional RTK-capable configuration sharpens relative distance and direction.
That upgrade is available where base stations are already deployed.
Alerts scaled to the moment
Warnings step up as closure time shortens, from heads-up to urgent.
Drivers respond early on open ground instead of reacting at the last second.
No cellular dependency
The mesh works directly between vehicles, in coverage holes and on remote benches.
If the site has no network at all, the warnings still travel.
Dump-Edge Rear Guard
Backing toward the edge should be a measurement, not a feeling.
Distance that updates in real time
A rear radar measures the gap to the berm, wall or drop-off — out to metres — while the truck reverses.
The driver sees the number closing instead of guessing from the mirror.
Radar that ignores the dust
Millimeter-wave sensing stays accurate when cameras are blinded by a plume.
Close-range measurement keeps working right up to the edge.
Warnings that arrive in steps
Layered alerts give the driver time to stop short of the limit.
Over-edge dumping and rollover risk drop because the stop point is known.
Vision and radar checked against each other
The rear camera and the radar must agree before a warning escalates.
In a dust storm, that means fewer false calls and more credible ones.
Forward Risk in Dust and Darkness
In a dust plume, a stopped loader can appear only seconds before impact.
Radar that sees what cameras cannot
Radar detects through dust, fog and rain, while vision adds object identity.
Sensor fusion keeps the risk picture alive when visibility collapses.
Alert logic written for the pit
The risk engine knows that stationary machines, berms and haul trucks behave differently.
It suppresses the urban-style false alarms that plague generic systems.
Profiles for the road out of the mine
When a truck runs public roads to reach another pit, the logic shifts.
Site profile and transit profile are both stored on the same unit.
Coverage across the full shift
Fusion detection is designed for night operation, not just daylight.
Low-light cameras and radar together carry the graveyard shift.
Built for the site, ready for the future
Dust gets in, darkness falls, the network drops — the hardware was designed around that reality.
Sealed and rugged by default
Cameras and the SOHANIoT AI Mobile DVR are housed for dust and vibration.
Power input covers the 12/24 V systems found on mining machines.
The edge is the primary brain
Alerts, recording and storage all run on the vehicle.
The cloud platform improves the fleet; it is not required for safety.
Autonomous-ready by design
Every device streams structured events over an open data bus.
The same position feeds, radar readings and camera streams that drive today's alerts are the inputs tomorrow's automation controller will consume.
Modules you can start small with
Deploy one scenario first — dump-edge radar, for example.
Add dispatch, V2X or driver monitoring later without replacing what is installed.
Why mining operators choose SOHANIoT
Factory-direct hardware, open integration and a size that fits the site.
Manufactured and tested in-house
The devices are built and validated in our own facility.
Certified industrial-grade hardware is the standard, not a premium option.
Edge AI plus an open cloud
Alerts run locally; data syncs to the SOHANIoT Cloud Platform when coverage returns.
APIs stay open so integrators can build on top instead of around.
Scenario-fit customization
We configure, brand and integrate to match your operation or your customer's.
Right-sized for a single pit today and scalable to a whole region tomorrow.
Start a conversation
Want the full product lineup behind this solution? Talk to a SOHANIoT engineer.
Talk to an engineerExplore the hardware that powers it.
View products