Manipulator Motion Planning

Path planning using RRT and rigid body trees

Manipulator motion planning involves planning paths in high-dimensional space based on the degree-of-freedom (DOF) of your robot and the kinematic constraints of the robot model. Kinematic constraints for the robot model are specified as a rigidBodyTree object. Use the manipulatorRRT to plan paths in the joint space using the rapidly-exploring random tree (RRT) algorithm.

Objects

`chompCollisionOptions`	Collision options for CHOMP trajectories
`chompSmoothnessOptions`	Smoothness options for CHOMP trajectories
`chompSolverOptions`	Solver options for CHOMP motion planner
`manipulatorCHOMP`	Covariant Hamiltonian optimizer for rigid body tree motion planning
`manipulatorRRT`	Plan motion for rigid body tree using bidirectional RRT
`manipulatorStateSpace`	State space for rigid body tree robot models
`manipulatorCollisionBodyValidator`	Validate states for collision bodies of rigid body tree
`workspaceGoalRegion`	Define workspace region of end-effector goal poses

Functions

expand all

RRT

`plan`	Plan path using RRT for manipulators
`interpolate`	Interpolate states along path from RRT
`shorten`	Trim edges to shorten path from RRT

State Space and Validator

`isStateValid`	Check if state is valid
`isMotionValid`	Check if path between states is valid

Workspace Goal Region

`sample`	Sample end-effector poses in world frame
`show`	Visualize workspace bounds, reference frame, and offset frame

Topics

Pick and Place Using RRT for Manipulators
Using manipulators to pick and place objects in an environment may require path planning algorithms like the rapidly-exploring random tree planner.
Pick-and-Place Workflow Using RRT Planner and Stateflow for MATLAB
This example shows how to setup an end-to-end pick-and-place workflow for a robotic manipulator like the KINOVA® Gen3.
Pick-and-Place Workflow in Gazebo Using Point-Cloud Processing and RRT Path Planning
Set up an end-to-end, pick-and-place workflow for a robotic manipulator like the KINOVA® Gen3.
Plan Paths With End-Effector Constraints Using State Spaces For Manipulators
Plan a manipulator robot path using sampling-based planners like the rapidly-exploring random trees (RRT) algorithm.

Featured Examples

Pick and Place Using RRT for Manipulators

Using manipulators to pick and place objects in an environment may require path planning algorithms like the rapidly-exploring random tree planner. The planner explores in the joint-configuration space and searches for a collision-free path between different robot configurations. This example shows how to use the manipulatorRRT object to tune the planner parameters and plan a path between two joint configurations based on a rigidBodyTree robot model of the Franka Emika™ Panda robot. After tuning the planner parameters, the robot manipulator plans a path to move a can from one place to another.

Open Live Script

Pick-And-Place Workflow Using CHOMP for Manipulators

Use CHOMP to plan smooth and collision-free trajectories for a pick-and-place workflow

Open Live Script

Pick-and-Place Workflow Using RRT Planner and Stateflow for MATLAB

Setup an end-to-end pick-and-place workflow for a robotic manipulator like the KINOVA® Gen3.

Open Live Script

Pick-and-Place Workflow in Gazebo Using Point-Cloud Processing and RRT Path Planning

Set up an end-to-end, pick-and-place workflow for a robotic manipulator like the KINOVA® Gen3.

Open Live Script

Motion Planning for Backhoe Using RRT

Plan a path for a backhoe, in an environment that contains obstacles, by using a motion planner. In this example, you model the backhoe in simplified form as a kinematic tree, then use a sampling-based motion planner to determine a viable path for the backhoe between two poses in the presence of obstacles. Simulate the outcome in a Simscape™ Multibody™ model.

Open Live Script