Dynamic Validation of Calibration Accuracy and Structural Robustness of a Multi-Sensor Mobile Robot

Area of Science:

• Robotics
• Sensor Systems
• Calibration and Validation

Background:

• High-precision calibration and structural robustness are critical for multi-sensor mobile robot reliability.
• Existing validation methods are insufficient for dynamic traveling scenarios.

Purpose of the Study:

• To present a novel validation method for multi-sensor mobile robot calibration accuracy and structural robustness.
• To enable dynamic evaluation during mobile robot testing phases.

Main Methods:

• A "ground surface simulation field (GSSF)-mobile robot -photoelectric transmitter station (PTS)" mechanism was developed.
• Static GSSF established with true north datum; PTS for real-time mobile vehicle pose measurement.
• Dynamic sensor pose measurement and evaluation using epipolar line alignment and point cloud projection/superposition.

Main Results:

• The method dynamically evaluates calibration accuracy and structural robustness of various sensors (Navcam, Hazcam, multispectral, TOF, LiDAR).
• Epipolar line alignment assessed relative orientation calibration for binocular cameras.
• Point cloud projection and superposition evaluated absolute calibration and robustness against the vehicle body.

Conclusions:

• The proposed method provides a reliable means for dynamic validation of multi-sensor mobile robots.
• This contributes to ensuring the healthy and robust operation of mobile robots in complex environments.