Design and Implementation of HD Mapping, Vehicle Control, and V2I Communication for Robo-Taxi Services

Area of Science:

• Robotics
• Computer Science
• Transportation Engineering

Background:

• Autonomous driving systems require robust software stacks for real-world applications like robo-taxi services.
• Key challenges include high-definition mapping, precise vehicle control, and effective communication systems.
• Vehicle-to-infrastructure (V2I) communication offers potential for enhanced situational awareness and safety.

Purpose of the Study:

• To present an autonomous driving (AD) software stack developed for a robo-taxi service contest.
• To focus on the critical components of high-definition mapping, vehicle control, and V2I communication.
• To demonstrate the application of V2I communication for improved traffic light recognition and robo-taxi operations.

Main Methods:

• Development of an integrated autonomous driving software stack.
• Implementation of high-definition mapping and advanced vehicle control algorithms.
• Integration of Vehicle-to-Infrastructure (V2I) communication for data exchange with roadside units.

Main Results:

• A V2I communication-enabled traffic light recognizer achieved a nearly perfect recognition rate.
• A robo-taxi service handler was successfully developed and implemented.
• V2I communication facilitated timely, non-line-of-sight-aware traffic information for AD applications.

Conclusions:

• The developed AD software stack, leveraging V2I communication, is effective for robo-taxi services.
• V2I communication significantly enhances traffic light recognition accuracy.
• The approach demonstrates a viable solution for autonomous passenger transportation on public roads.