Remote Driving Contactless Logistics Platform for Epidemic Scenarios
Shanghai Jiao Tong University — Team Project
Role: Team Lead & System Integration • Tools: ROS, C++, Python, OpenCV
Introduction / Social Motivation
During severe epidemic outbreaks, logistics workers who deliver medical supplies, food, and equipment to high-risk areas face significant exposure. Our project proposed a contactless logistics platform in which an operator remotely drives a small electric vehicle through hospital corridors or quarantine zones, reducing unnecessary person-to-person contact while maintaining reliable delivery service.
System Design & Methods
- Built a ROS-based teleoperation architecture connecting on-vehicle sensors (cameras, LiDAR) with a remote operator station over a wireless network.
- Implemented a visual assistance interface that fused multi-view camera feeds with predicted trajectories and key overlays (e.g., distance to obstacles, corridor boundaries).
- Mapped joystick inputs from the operator to low-level motor and steering commands, including smooth acceleration/braking profiles.
- Added safety layers such as soft geofencing, maximum speed enforcement, and a software emergency stop.
Results / Evaluation
We evaluated the platform in a mock hospital corridor with narrow passages and multiple turns. The system maintained end-to-end control latency within approximately 150 ms while streaming live video and sensor data. Remote operators were able to complete delivery routes with centimeter-level tracking accuracy and no collisions, even under limited visibility.
The project was recognized at a national-level innovation competition for its potential to support safer logistics workflows in future public health emergencies.
My Contribution
- Led the overall system architecture and task planning as team leader.
- Implemented ROS nodes for trajectory prediction visualization and sensor fusion on the operator interface.
- Integrated the onboard embedded controller with the remote control station and tuned the control command mapping for smooth teleoperation.
- Designed and executed test scenarios, collected performance metrics, and prepared the technical documentation and presentation materials.
Artifacts
- ROS package for teleoperation and visual assistance.
- Demo video showing remote deliveries in a corridor environment.
- Poster and written report summarizing the system and evaluation results.