Stockpile Collapse Prevention AI

Spot undercut faces, overhangs, voids over feeders, and people or machines at the toe before a pile gives way.

Try it now for freeTalk to an AI Expert
Aggregate stockpile with a loader at an undercut toe, the overhang and toe zone outlined in purple

Stockpile Collapse Prevention AI for the Toughest Quarries, Mines, Ports, and Yards

  • Deploy Anywhere, Run Everywhere

    Run stockpile collapse prevention on yard, reclaim, and drone cameras on the edge, on-prem, in your VPC, or via API.

  • One Platform, Full Adoption

    Tools every operations and safety team can adopt, from quarry managers to safety leads and loader operators, no separate ML team required.

  • Secure, Compliant, and Audit-Ready

    Data stays safe with SOC 2 Type II compliance, encrypted data, HIPAA compliance, and an uptime SLA.

  • Undercut Toes & Steep Faces
  • Overhangs & Vertical Walls
  • Voids & Bridging over Reclaim Feeders
  • People & Vehicles at the Toe or Crest
  • Slumps, Cracks & Face Change
  • Exclusion Zone Alerts Around Piles

Talk to a Vision AI engineer who's shipped in mining and aggregates.

Bring us your toughest stockpile safety problem and we'll map a working solution.

Ask us about:

  • Solution architecture that fits your stockpiles, reclaim systems, and MSHA safety program
  • A live demo on your own yard, reclaim, or drone footage
  • Deployment options: edge, on-prem, air-gapped, VPC, or on the cameras and network your site already runs
  • ROI modeling against engulfment risk, equipment damage, lost production, and citations

Over 80,000 organizations build with Roboflow.

  • Rivian
  • Pella
  • Chobani
  • USG Corporation
  • BNSF Railway
  • American Woodmark
Start where you are

Vision AI is transforming businesses

Customers across the board are solving complex challenges and driving meaningful impact.

  • Automotive customer

    $10 million

    Saved by automatically detecting defects on the production line

  • Logistics & freight company

    90%

    Less time spent manually tracking shipping inventory

  • Building materials supplier

    60%

    Lower customer return rate with improved product quality

Watch Every Pile, from the Crest to the Reclaim Feeder

Add a real-time safety layer to every stockpile with Roboflow Vision AI for stockpile collapse prevention. Built for quarries, mines, ports, and yards where a loader that undercuts the toe, a void over a reclaim feeder, or a worker at the base of a surge pile can turn into an engulfment in seconds.

The face and the toe:

  • Detect undercut toes, overhangs, and steep or vertical faces left by loading, with a model trained on your own piles and material.
  • Measure the face profile with depth estimation and flag sections steeper than your site's limit.
  • Spot slumps, tension cracks, and fresh slides on the face, and compare each pass with the last.

Voids, feeders, and reclaim:

  • Watch the surface above reclaim feeders and draw points for depressions and craters that signal a void or bridging below.
  • Flag blocked or bridged feeders early, so they are cleared with mechanical devices and not by someone standing at the toe.
  • Track pile shape and size over time alongside stockpile volume measurement.

People, machines, and exclusion zones:

  • Detect people and vehicles at the toe, on the crest, or on top of surge piles, and alert the moment they enter an exclusion zone.
  • Track loaders and dozers working the pile with the same models that power heavy equipment detection.
  • Save a clip of every undercut, slide, and zone entry for toolbox talks and incident reviews.

Bring intelligence to every stockpile today. Stop an undercut face or a hidden void from becoming an engulfment.

‍

Frequently asked questions

What is stockpile collapse prevention with Vision AI?

Stockpile collapse prevention with Vision AI uses fixed cameras and drone flights around stockpiles with a trained computer vision model (opens in a new tab) to find undercut toes, overhangs, steep faces, voids over feeders, and people or machines in danger zones, and alert operators and supervisors in real time. It works alongside stockpile volume measurement and slope stability monitoring.

What causes stockpiles to collapse?

Incident records point to the same causes again and again: piles undercut by loaders, steep or vertical faces, hidden voids and bridging over feeders and draw points, and people or machines working at the toe or on the crest. Wet material adds to the risk.

Does stockpile collapse prevention support MSHA requirements?

Stockpile collapse prevention supports MSHA 30 CFR 56.9314 (the requirement to trim stockpile and muckpile faces to prevent hazards to persons) and MSHA best practices for keeping miners out of areas where material can fall or slide, with a record of face conditions and zone entries.

Roboflow serves as the monitoring engine, and your operations and safety teams own the face limits, exclusion zones, and response.

Can it integrate with our alerting and plant systems?

Yes. Roboflow sends events, clips, and images over REST (opens in a new tab) or MQTT, or writes them to a database, so they land in safety, maintenance, and fleet systems. Alerts can go to a PLC over OPC UA to stop a feeder or sound a horn, or to Slack, email, or text.

‍