Campus Autonomy - Autonomous Indoor-Outdoor Delivery Vehicle

Development of an autonomous delivery vehicle capable of navigating both indoor and outdoor environments within a campus setting

Project Overview

The “Campus Autonomy” project focuses on developing an autonomous delivery vehicle capable of navigating both indoor and outdoor environments within a campus setting. By assembling an Agile X HUNTER SE Ackermann model drive vehicle equipped with advanced sensors like LiDAR and panoramic camera, the project aims to address the complex challenges of autonomous localization, path planning, and navigation.

Left: Agile X HUNTER SE vehicle chassis; Right: Completed autonomous delivery robot with full sensor suite

Key Technologies

Hardware Platform

  • Vehicle Platform: Agile X HUNTER SE Ackermann drive vehicle
  • LiDAR: Hesai PandarQT64 for high-precision environment perception
  • Vision System: Insta360 Air panoramic camera for 360° visual input
  • Navigation Sensors: Odometer and IMU for precise movement tracking

Software Stack

  • Framework: ROS2 with Navigation2 package
  • SLAM: Cartographer for simultaneous localization and mapping
  • Path Planning: Global and local planners for optimal route generation
  • Obstacle Avoidance: Real-time dynamic obstacle detection and avoidance

System Architecture

The system integrates sophisticated hardware and software components:

ROS2 Navigation2 system architecture showing the integration of perception, planning, and control modules

Key Features

1. Dual Environment Navigation

  • Indoor Navigation: Precise localization in structured environments
  • Outdoor Navigation: GPS-aided navigation with obstacle avoidance
  • Seamless Transition: Automatic switching between indoor and outdoor modes

2. Real-time Obstacle Avoidance

  • Dynamic obstacle detection using LiDAR point clouds
  • Adaptive path re-planning for moving obstacles
  • Safety-first approach with emergency stop capabilities

3. Modular Design

  • Scalable architecture for future sensor additions
  • Plugin-based navigation components
  • Easy configuration and parameter tuning

Technology Stack

  • Programming Languages: C++17, Python 3.8+
  • Robotics Framework: ROS2 Humble
  • Navigation: Navigation2 (Nav2) stack
  • SLAM Algorithm: Google Cartographer
  • Point Cloud Processing: PCL (Point Cloud Library)
  • Computer Vision: OpenCV, ROS2 Image Pipeline
  • Hardware Interface: ROS2 device drivers
  • Simulation: Gazebo Classic

Experimental Results

  • Localization Accuracy: ±0.2m in indoor environments
  • Path Planning Efficiency: 95% success rate in reaching destinations
  • Obstacle Avoidance: 100% collision-free navigation in test scenarios

System Metrics

  • Real-time Performance: 20Hz sensor processing
  • Battery Life: 4+ hours continuous operation
  • Payload Capacity: Up to 10kg delivery capacity

Implementation Highlights

Left: Field testing of the autonomous delivery robot; Right: Real-time visualization in RViz2

Future Work

Future development will focus on:

  1. osmAG Map Integration: Introducing osmAG map format into the Navigation2 stack
  2. Custom Global Planner: Replacing default global planner with osmAG Planner plugin
  3. Advanced Localization: Replacing AMCL with osmAG Localizer for improved accuracy
  4. Multi-robot Coordination: Enabling fleet management capabilities
  5. Weather Adaptation: Robust operation in various weather conditions

Project Team

  • Lead Developer: Jiajie Zhang (zhangjj2023@shanghaitech.edu.cn)
  • Co-developer: Yongqi Zhang (zhangyq12023@shanghaitech.edu.cn)
  • Advisor: Professor Sören Schwertfeger

This project demonstrates the practical application of autonomous navigation technologies in real-world campus environments, contributing to the advancement of service robotics and autonomous delivery systems.