1Demo
End-to-end autonomous shelf-stocking in the Webots simulation.
2Overview
This course project explored how a mobile robot could perform autonomous shelf-stocking tasks in a structured retail environment.
The robot was simulated using Webots and designed as a small mobile manipulator equipped with a gripper, camera-based object recognition, distance sensing, GPS-based localization, and compass-based heading estimation. The task was to let the robot identify products in a central goods area, pick them up, determine an appropriate shelf location, navigate through the supermarket layout, and place each item into the selected shelf cell.
Compared with a simple pre-programmed pick-and-place routine, this project emphasized a more complete task pipeline: product search, visual detection, grasping, shelf selection, autonomous navigation, and placement on both lower and upper shelf levels.
3Robot Platform
The simulated robot was a compact mobile manipulator with an onboard gripper and multiple sensing modules.
Mobile base
A small differential-drive base was used to move around the supermarket aisles and position the robot in front of target shelves.
Manipulator and gripper
The robot used a gripper mechanism to pick up products and place them into shelf cells. The arm configuration was adjusted to support both low-level and high-level placement.
Vision sensing
Camera recognition was used to detect target products and help align the robot before grasping.
Localization sensing
GPS and compass information were used to simplify localization and heading control inside the simulated supermarket.
4Simulation Environment
The supermarket scene was built in Webots with a central goods area and surrounding shelf regions.
The supermarket scene in Webots: a central goods area surrounded by shelf regions.
5Task Pipeline
The project combined perception, task planning, navigation, and manipulation into a complete shelf-stocking workflow.
TECHNICAL SKILLS: Webots simulation · Mobile manipulation · Camera recognition · GPS/compass localization · Task planning · Autonomous navigation · Pick-and-place control