Investigating the Improvement of Autonomous Vehicle Performance through the Integration of Multi-Sensor Dynamic...

Area of Science:

• Robotics and Artificial Intelligence
• Transportation Engineering
• Computer Vision

Background:

• Autonomous vehicles (AVs) represent a paradigm shift in mobility, moving beyond mere transportation to potential mobile offices or leisure spaces.
• Current AV technology faces commercialization challenges due to limitations in accuracy and stability.
• Ensuring the safety and reliability of AVs is paramount, necessitating advancements in driving technology.

Purpose of the Study:

• To propose a novel method for building precision maps tailored for multi-sensor-based autonomous driving.
• To enhance the accuracy and stability of autonomous vehicle technology through improved environmental perception.
• To address the limitations of current autonomous driving systems.

Main Methods:

• Development of a precision mapping method utilizing dynamic high-definition maps.
• Integration of multiple sensors, including cameras, LIDAR, and RADAR, for comprehensive environmental data acquisition.
• Leveraging multi-sensor data to improve object recognition and path planning for autonomous vehicles.

Main Results:

• The proposed method significantly enhances the recognition rates of objects surrounding the vehicle.
• Improved autonomous driving path recognition capabilities were demonstrated.
• The precision mapping approach contributes to greater accuracy and stability in autonomous driving maneuvers.

Conclusions:

• The developed precision mapping technique is crucial for advancing the reliability of autonomous driving systems.
• Multi-sensor integration with dynamic HD maps offers a viable solution to current AV technology limitations.
• This research paves the way for more robust and commercially viable autonomous vehicles.