REZOOM: Autonomous electric scooter for self-driving navigation and parking

This technology significantly enhances the efficiency of electric scooters by allowing them to reposition themselves autonomously. By reducing operational costs and increasing accessibility, it addresses key logistical challenges faced by urban mobility providers.

Background

Shared electric scooters are becoming a popular transportation option in cities, but they often cause clutter when parked in inappropriate locations. This not only inconveniences pedestrians but also creates challenges for fleet operators, who must manually rebalance the scooters using vehicles, increasing costs and emissions. The autonomous repositioning technology solves these issues by allowing scooters to move themselves to high-demand areas or proper parking spots, minimizing the need for human intervention.

Technology

Researchers at the University of Maryland have created an autonomous driving framework that integrates advanced localization and path-planning algorithms into an autonomous electric scooter, allowing it to navigate urban environments between 10 – 100 meters without human intervention. This system uses a combination of GPS, inertial measurement units (IMU), and wheel encoders to accurately determine the scooter’s position, even in areas with poor GPS coverage. The scooter can autonomously move to designated locations, solving issues related to improper parking and fleet management in shared micromobility systems.

Benefits

  • Eliminates the need for manual retrieval and parking of shared scooters, reducing operational costs and downtime.
  • Ensures accurate navigation through multi-sensor fusion (dual EKF with GPS, IMU, encoders) and AprilTag initialization.
  • Optimizes travel efficiency using historical ride data indexed by Uber H3 and an A* global planning algorithm.
  • Enhances safety and obstacle avoidance via stereo‐camera detection and Time Elastic Band local path planning.
  • Improves ride comfort and stability with reinforced support wheels and suspension systems.
  • Automates repositioning and parking in designated zones, minimizing urban clutter and misplacement.
  • Leverages an open-source ROS stack on affordable single-board computers for flexible development and maintenance.

Applications

  • Shared micromobility fleet management
  • Autonomous scooter last-mile delivery
  • Automated fleet rebalancing service
  • Campus mobility shuttle services
  • Tourist urban scooter tours