What problem does it solve?
Building manipulator models in MATLAB often leads to silent failures: missing collision meshes, misaligned end-effector frames, unconverged IK solutions, and grippers attached in the wrong orientation. This Skill guides the correct construction, configuration, and validation of rigidBodyTree robot models so downstream simulation, planning, and pick-and-place tasks behave as expected.
Core Features & Use Cases
- Model Loading and Construction: Choose between loadrobot, importrobot (URDF/Xacro/SDF), or building from DH parameters, with guidance on collision mesh generation via VHACD decomposition.
- Gripper Attachment: Select and attach Robotiq parallel-jaw or vacuum grippers with correct frame alignment, including the KUKA iiwa tool0 correction and contact-frame IK targeting pattern.
- Kinematics Solving and Validation: Decide between inverseKinematics and generalizedInverseKinematics, then mandatorily validate exit flags, position error, constraint violations, and joint continuity across waypoints.
- Use Case: A user asks to simulate a UR5e picking up a box. The Skill loads the robot with loadrobot, attaches a robotiq2F85 gripper with ReplaceBase=false, adds a grasp-center contact frame, solves IK to each waypoint, and validates every solution before visualization.
Quick Start
Ask your AI agent to load a UR5e robot model in MATLAB, attach a Robotiq gripper, and solve validated inverse kinematics for a target pick-and-place pose.