Unmanned Supermarket Navigation

Shenzhen Stability Support Project – robust navigation in dynamic supermarkets

Overview

This project develops an unmanned supermarket navigation system capable of operating reliably in highly dynamic, cluttered retail environments.

Role

  • Algorithm Engineer

Funding Source

  • Shenzhen Stability Support Project

Contributions

  • Implemented a Visual-Inertial Odometry (VIO) algorithm tailored to dynamic supermarket scenes with moving customers, trolleys, and changing layouts.
  • Built a multi-sensor fusion framework, combining 2D LiDAR, occupancy grids, and visual-inertial measurements to achieve robust localization and mapping.
  • Addressed challenging factors such as dynamic obstacles and lighting variations, improving navigation stability and safety in real deployments.

Techniques & Tools

  • Visual-Inertial Odometry (VIO)
  • 2D LiDAR and occupancy grid mapping
  • Multi-sensor fusion for robust navigation