Grocery Store Robot: Arm Manipulation

Background

Grocery stores and warehouses still rely heavily on manual labor for grasping items from shelves. This process is repetitive, slow, and prone to human error. As robotic automation advances, there is a clear need for a reliable, low-cost robotic arm capable of performing collision-free grasping in cluttered retail environments.
Our project addresses this challenge by developing an autonomous robotic arm integrated with Team 1's mobile base to create a complete grocery-grasping system.

Goal and Objectives

Project Goal:
Design, simulate, and implement a robotic arm capable of fully autonomous, collision-free grasping of shelf items using ROS 2, Gazebo, and planning algorithms.

Objectives:

• Select arm hardware; develop two-link kinematic model
• Build ROS 2 control package and hardware-in-the-loop interface
• Construct Gazebo simulation environment and grasping workflow
• Implement MoveIt/OMPL collision-free motion planning
• Integrate with Team 1's mobile base and achieve full-system demonstration

More Information

This project integrates perception, kinematics, simulation, and motion planning. Major technical components include:

• Forward and inverse kinematics for a two-link robotic arm
• MoveIt-based collision-free planning and execution
• YOLO-based object detection and shelf-item localization
• MATLAB–ROS 2 integration for controller development
• Gazebo simulation of shelves, items, and robot workspace
• Gripper force analysis and servo trade studies
• Arm–base coordination for safe picking in narrow aisles

Link to Documentation

Preliminary Design Review:

https://docs.google.com/presentation/d/1BevuY3QyifiUK-B4by5CUpGox8OD5u6oIELpYGJYa-8/edit?usp=sharing

Team Contacts

• Pengfei (Perry) Xia: pengfx1@uci.edu
• Ashwath Palavalli: palavala@uci.edu
• Tian (Kris) Xia: xiat14@uci.edu
• Cai Chen: caic11@uci.edu
• Mikel Vicente: mikelv1@uci.edu

Sponsor / Advisor

Faculty Sponsor: Prof. Solmaz Kia (solmaz@uci.edu)
Faculty Mentors: Prof. Mark Walter, Prof. David Copp
TA: Mohamed Shorbagy

Team Identity

Our team name, ArmY, reflects both our technical focus on robotic arm manipulation and the collaborative strength of our group. Each member contributes expertise in kinematics, simulation, ROS 2 development, perception, or system integration. Together, we aim to build a robust, efficient, and safe platform for real-world autonomous item retrieval.