Pick and place robots are being increasingly used in applications such as manufacturing, warehouse automation and e-commerce order fulfillment. They speed up the process of picking up parts or items and placing them in other locations. By handling repetitive tasks, pick and place robots free up human workers to focus on more complex work and thus increase the overall system productivity.

The goal for this project is to move objects between containers. Specifically, the forward task is to move objects from the middle container to containers on the side, and the reset task is to move objects back from the side containers to the middle one. The robot should switch to the other task when the current container is empty.

We implemented the state-machine for both forward and resetting task. The major approaches for this project is listed below:

• ResNet-50-based binary classifier: Trained the pre-trained ResNet-50 model to check if the container is empty or not using the RGB image from the wrist camera on the robot.
• GQCNN 2.0 optimal pose predictor: Used the pre-trained GQCNN 2.0 model to predict the optimal grasping pose using the depth image from the wrist camera on the robot.
• RRT trajectory planner: Used the RRT algorithm to plan the trajectory for robot to move towards the target pose and grasp the objects.