The Influence of Micro-Hexapod Walking-Induced Pose Changes on LiDAR-SLAM Mapping Performance

Area of Science:

• Robotics
• Artificial Intelligence
• Sensor Fusion

Background:

• Micro-hexapods offer advantages for exploration in confined or disaster areas.
• Simultaneous localization and mapping (SLAM) is crucial for autonomous robot navigation.
• Existing SLAM research inadequately addresses body axis variations in legged robots.

Purpose of the Study:

• To evaluate the impact of walking locomotion on the SLAM performance of hexapod robots.
• To compare the SLAM capabilities of hexapod robots with a crawler robot.
• To identify challenges in SLAM for hexapods and propose future solutions.

Main Methods:

• Implemented a consistent SLAM system and LiDAR sensor for both hexapod and crawler robots.
• Conducted comparative analysis of map accuracy and LiDAR point cloud data using pattern matching.
• Assessed the effect of hexapod body axis fluctuations on sensor data.

Main Results:

• Hexapod robots exhibited significant fluctuations in LiDAR data due to body axis changes.
• These fluctuations led to a measurable decrease in overall map accuracy compared to the crawler robot.
• The study highlights the detrimental effect of dynamic body posture on hexapod SLAM.

Conclusions:

• Body axis variations pose a significant challenge to the accuracy of SLAM in hexapod robots.
• Future SLAM systems for multi-legged robots must account for dynamic body movements.
• Integration of inertial measurement units (IMUs) is proposed to compensate for body axis changes during locomotion.