Implementation of SLAM-Based Online Mapping and Autonomous Trajectory Execution in Software and Hardware on the...

Area of Science:

• Robotics
• Autonomous Systems
• Sensor Fusion

Background:

• Agile maneuvering in confined spaces presents significant challenges for autonomous vehicles.
• Existing systems often struggle with real-time mapping and precise trajectory following in complex environments.

Purpose of the Study:

• To design and implement a SLAM-based system for online mapping and autonomous trajectory execution.
• To enable agile maneuvering of the Nimbulus-e concept vehicle in confined spaces.

Main Methods:

• Utilized a steer-by-wire system with in-wheel motors and eight joint variables for precise control.
• Integrated LiDAR for real-time environment mapping and obstacle detection.
• Employed vehicle odometry data to enhance computational efficiency and accuracy.
• Applied A* and Hybrid A* algorithms for trajectory planning and optimization.

Main Results:

• Successfully demonstrated real-time environment mapping and localization.
• Achieved precise trajectory following and smooth vehicle movement.
• Validated system effectiveness through ADAMS Car-MATLAB co-simulation and a scaled physical prototype.

Conclusions:

• The developed SLAM system effectively enables reliable and efficient navigation for agile vehicles.
• Combining advanced sensors with innovative control algorithms shows significant potential for autonomous systems.
• Future work will focus on controller robustness and expanding applications to urban and agricultural settings.