Text size

Line spacing

Contrast

Placeholder Placeholder
Background image
For visually impaired

International Self-Driving Shuttle Project

International Self-Driving Culture Shuttle Project

From June to October 2024, Tartu hosted a groundbreaking self-driving culture shuttle pilot project, combining future technologies, public space, and mobility. The project brought together technology companies from three countries: Yazaki Innovations, a subsidiary of Japan's Yazaki Group; Avanti R&D, a U.S. tech firm; Auve Tech OÜ, an Estonian developer of autonomous vehicles; and the City of Tartu.

In November 2024, the project was showcased at the world’s largest smart city event, Smart City Expo Barcelona.

Connecting Cultural Landmarks

The self-driving shuttle, an eight-seat self-driving vehicle, operated five days a week on a set route along Roosi Street. It connected the Tartu City Museum and the Estonian National Museum from June 21 to October 31, transporting nearly 2,000 passengers in total.

Project Goals

The initiative aimed to develop passenger-related technologies and test connectivity between the self-driving shuttle and road infrastructure. During the pilot, data was gathered to demonstrate how city-installed cameras and sensors provided real-time information about other vehicles, pedestrians, and cyclists at complex intersections, ensuring smooth navigation for the self-driving shuttle. The project also tested technologies for offering passengers additional information during the ride while maintaining safety.

What we piloted 

We integrated advanced Artificial Intelligence computer vision and audio analytic technologies through 5 key components:

  • Sit / stand detection utilizing pose estimation algorithms to ensure passengers are safely seated
  • Demographic-targeted media for personalized, multilingual information
  • VRU (Vulnerable Road User) collision avoidance utilizing real-time data from edge computing roadside sensors, which informs the shuttle of nearby cyclists and pedestrians at blind intersections.

Key Outcomes

  • Passengers spoke 15 different languages during the trial.
  • Smart sensors issued nearly 5,000 warning alerts about approaching pedestrians and cyclists at low-visibility intersections.
  • A total of 1,902 people rode the self-driving shuttle, which traveled 3,426 kilometers, 86% of which were driven autonomously.

This innovative collaboration highlights Tartu’s commitment to exploring cutting-edge mobility solutions and promoting sustainable urban development.

For more information, visit:
www.auve.tech/case-studies/tartu-autonomous-public-transportation

Last changed 29.11.2024