Jay Menon

Robotics @ ASU

This project presents a dual-mode approach for implementing a palletizing task using a Universal Robots UR5 arm. The robot was tasked with picking die-like cubes from a linear conveyor and placing them onto a pallet in a structured 3x2 grid. The implementation was carried out using two methodologies: (1) direct programming via the UR Teach Pendant interface, and (2) simulation and control via ROS2 and MoveIt. This comparison highlights trade-offs in flexibility, scalability, and integration complexity.

Video Demo


The initial implementation was conducted directly on the UR5 hardware using the Teach Pendant’s block-based programming interface. This method leverages Universal Robots’ Polyscope environment, enabling users to develop sequences with drag-and-drop programming blocks and real-time waypoint teaching.

Implementation Details

  • Palletizing Wizard: The built-in Palletizing Wizard was employed to define the grid pattern. By specifying the first and last positions of a row, the wizard interpolated intermediate positions, streamlining the programming process.
  • Waypoint Definition: Waypoints for pick and place positions were taught by physically guiding the robot arm while holding the Freedrive button, allowing for precise positioning without manual coordinate entry.
  • Gripper Control: Gripper operations were integrated using I/O commands within the program, ensuring synchronized opening and closing actions during pick-and-place cycles.
  • Looping: A simple loop structure was defined to repeat the pick-place cycle for all six cubes.

ROS2 Implementation

To overcome the limitations of the Teach Pendant approach, the task was re-implemented in a ROS2 simulation environment using Gazebo and MoveIt. This allowed for greater flexibility, scalability, and integration with other systems.

  • The MoveIt Task Constructor was employed to define a sequence of stages for picking and placing, including approach, grasp, lift, transport, and release.
  • Gripper actions were controlled via ROS2 services.