Offroad Navigation for Autonomous Haul Trucks in Open Pit Mine

Since R2024a

Hauling material in an open pit mine requires a planning stack capable of both navigating at the global scale and avoiding obstacles during local path following. This example series uses Robotics System Toolbox™ Offroad Autonomy Library and Navigation Toolbox™ to do the planning, navigation, and visualization in an Unreal Engine® pit mine environment. To download the Offroad Autonomy Library support package, see Install Robotics System Toolbox Offroad Autonomy Library Support Package.

This figure shows a visualization of the haul truck in the pit mine environment.

The solution proposed by this example series approaches the offroad navigation problem in four steps described in these four examples:

1. Create Route Planner for Offroad Navigation Using Digital Elevation Data

This example shows how to create a road network from a point cloud.

2. Create Onramp and Terrain-Aware Global Planners for Offroad Navigation

This example uses global planners to create kinematically feasible paths that merge from the start position onto the road network, follow the network, and exit toward the final goal.

3. Navigate Global Path Through Offroad Terrain Using Local Planner

This example uses a local planner to follow the reference path obtained from the previous example while respecting haul-truck kinematics and avoiding obstacles.

4. Create Path Following Model Predictive Controller

This example uses an MPC to generate smooth velocity commands from the local planner’s output.

5. Model and Control Autonomous Vehicle in Offroad Scenario

The last example in the series shows how to use Simulink® to integrate the road network and planners into an autonomous navigation stack with logic controlled by Stateflow®.

This figure shows the top level of the autonomous navigation stack.

This example is also available on file exchange using a project format. See Offroad Navigation for Autonomous Haul Trucks in Open Pit Mine in File Exchange for more information.