Pick-and-Place Robotic System

(Intro to Robotics Final Project)

What:

• Collaborated in a team to develop a robotic system for a pick-and-place challenge using a Franka Emika Panda robotic arm.

• Task involved acquiring and stacking static and dynamic blocks in both simulated and hardware environments.

How:

• Designed efficient algorithms for static block stacking and dynamic block tracking, incorporating inverse kinematics for precise manipulations.

• Created predictive models to anticipate block movement on a rotating platform, enabling accurate and reliable grasping.

• Used Gazebo and ROS for simulation, transitioning the system to real-world hardware while addressing challenges such as camera glare, z-axis offsets, and real-time calibration.

• Performed iterative testing to fine-tune gripper control and optimize system performance.

Results:

• Achieved 100% accuracy for static block pickups and 82.75% accuracy for dynamic blocks.

• Successfully handled dynamic block movement with an angular velocity of π/60 radians/second.

• Completed key tasks with average timings of 3.78 seconds for moving to hover position, 5.23 seconds for static block pickup, and 5.61 seconds for dynamic block hover positioning.

• Optimized gripper actions, achieving average times of 0.28 seconds for static blocks and 1.49 seconds for dynamic blocks.

• Delivered a robust system capable of bridging the gap between simulation and hardware, demonstrating efficiency in real-world automation tasks.